BOUGHT WITH THE INCOME 
 FROM THE 
 
 SAGE ENDOWMENT FUND 
 
 THE GIFT OF 
 
 Hetirg W. Sage 
 
 iSqx 
 
 v4. ^OV^2l(i>4 /3/7//7<?6 
 
 5901 
 
 i£H£ 
 
Cornell University Ubrary 
 BT78 .M17 
 
 Some dogmas of religion, by John McTagge 
 
 oljn 
 
 3 1924 029 370 644 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924029370644 
 
SOME DOGMAS OF RELIGION 
 
SOME 
 DOGMAS OF RELIGION 
 
 BY 
 
 JOHN McTAGGART ELLIS McTAGGART 
 
 DOCTOR IN LETTERS 
 
 FELLOW AND LECTURER OF TRINITY COLLEGE IN CAMBRIDGE 
 
 AUTHOR OF 
 
 ' STUDIES IN THE HEGELIAN DIALECTIC ' AND * STUDIES IN HEGELIAN COSMOLOGY ' 
 
 LONDON 
 EDWARD ARNOLD 
 
 41 & 43 MADDOX STREET^ BOND STREET, W. 
 
 1906 
 
 [All riffhts reserved] 
 
TO MY MOTHEE 
 
 AND 
 
 MY WIFE 
 
PREFACE 
 
 The first chapter of this book is based upon an 
 article entitled ' The Necessity of Dogma ' which 
 appeared in the International Journal of Ethics (Jan., 
 1895). The third and fourth chapters appeared in 
 the same magazine (Jan., 1903, and Oct., 1904), but 
 have been rewritten and considerably altered. Part 
 of the second chapter appeared in the Hihiert Journal 
 for Oct., 1905, but this also has been rewritten. 
 
 I am much indebted to Mr. G. L. Dickinson of 
 King's College, to Mr. G. E. Moore of Trinity College, 
 and to my wife, for their kindness in reading this 
 book in MS. and suggesting many valuable altera- 
 tions. To Mr. Moore, in particular, I owe the deepest 
 gratitude for criticisms which revealed — and, I trust, 
 enabled me to correct — many lurking errors. 
 
 Trinity College, Cambridge, 
 
 January, 1906. 
 
TABLE OF CONTENTS 
 
 CHAPTER I 
 
 THE IMPORTANCE OF DOGMA 
 
 A 
 SECT. PAGE 
 
 1. Definition of Dogma 1 
 
 2. This is not inapplicable to those dogmas which assert historical 
 
 events . 2 
 
 3. Definitions of RelifijiouB Dogma, and of Religion .... 3 
 
 4. This definition of Religion is not too narrow .... 4 
 
 5. And even Magic appears by this definition to be a form of 
 
 Religion, though a rudimentary form 5 
 
 6. This definition will include Buddhism 6 
 
 7. The proposals to define Religion as identical with Morality. Its 
 
 inconvenience 6 
 
 8. The utility of such a definition in the past was due to causes 
 
 which are ceasing to act 7 
 
 9. The proposal to define Religion as * Morality touched by 
 
 emotion '. Its inconvenience 8 
 
 10. Mr. Dickinson's definition of Religion 10 
 
 11. Religion in the present stage of thought requires that the uni- 
 
 verse should be judged to be good on the whole . . ,11 
 
 12. Religion requires Dogma 13 
 
 13. Nothing taught us by Science can take the place of Dogmas . 13 
 
 14. And, if Philosophy should convince us that the true Dogmas 
 
 were identical in content with the fundamental propositions 
 
 of Science, there could be no basis for Religion a priori . . 14 
 
 15. Nor can there be an empirical basis, either in the condition of the 
 
 present 16 
 
 16. Or in our hopes of the future 18 
 
 17. The suggestion that Science can give some basis for Religion, 
 
 though not so good a basis as can be found elsewhere . . 19 
 
 18. The same continued 21 
 
 19. Nor can Religion be based on virtue 22 
 
 B 
 
 20. The theory that to the Christian Religion, in particular, nothing 
 
 is essential but morality. But Jesus taught Dogmas . . 24 
 
X TABLE OF CONTENTS 
 
 SECT. PAGE 
 
 21. And regarded them as essential. And the Church has followed 
 
 him in this 26 
 
 22. Christianity may be defined to exclude Dogma, but it will not be 
 
 the Christianity of the past or the present. And it will not 
 be a Keligion 27 
 
 23. Christianity is sometimes maintained to be undogmatic by 
 
 persons who define Dogma as consisting of propositions which 
 are false, or unproved 28 
 
 24. Or who define it as consisting of propositions which are com- 
 
 plicated and difficult ........ 29 
 
 25. Or who define it as consisting of propositions on which Chris- 
 
 tians do not agree 30 
 
 C 
 
 26. However we define Religion, the importance of Dogma for our 
 
 happiness is very great 82 
 
 27. Metaphysics are in reality the least abstract branch of knowledge, 
 
 and are among the most practical 38 
 
 28. The dependence of Religion on Dogma makes religious questions 
 
 controversial and doubtful ....... 38 
 
 29. But if it is eliminated it t^,kes with it the valuable element in 
 
 Religion 35 
 
 30. This is not completely satisfactory, but it is useless to refuse to 
 
 face it 36 
 
 CHAPTER II 
 
 THE ESTABLISHMENT OF DOaMA 
 
 31. Introductory 38 
 
 A 
 
 32. Dogmas are sometimes supported by the assertion that they do 
 
 not rest on arguments, and cannot be shaken by them . . 38 
 
 33. Such an assertion can give no reason for any person to believe 
 
 a dogma who does not believe it already . . . .39 
 
 34. The analogy of a blind man, and the analogy of a man in 
 
 delirium 40 
 
 35. These analogies further considered 42 
 
 36. It is possible that a man may be mistaken in asserting that his 
 
 belief in a particular dogma does not rest on arguments and 
 cannot be shaken by them 43 
 
 37. All knowledge must rest on some immediate convictions, but 
 
 this gives no reason for supposing that any religious dogmas 
 are immediately certain 45 
 
TABLE OF CONTENTS xi 
 
 B 
 
 SECT. PAGE 
 
 88. It is sometimes asserted that a particular dogma must be true 
 because it is believed by all people, or certain people. The 
 argument that it is believed by every one . . . .46 
 
 39. The argument that it is believed by all unsophisticated people . 46 
 
 40. The argument that it is believed by most people . . .47 
 
 41. The argument that it is believed by workers of miracles. The 
 
 evidence for miracles ........ 48 
 
 42. Miracles do not prove the existence of a God . . . .50 
 
 43. Nor is it certain that God could avoid deceiving us . , ,51 
 
 C 
 
 44. Again, it is asserted that a particular dogma must be true, 
 
 because, if it were false, the universe would be intolerably bad 52 
 
 45. But we have no right to assume that the universe is not in- 
 
 tolerably bad 53 
 
 46. Even if we had proved the universe to be good on the whole, we 
 
 could not safely argue to the non-existence of a particular 
 evil 54 
 
 47. The reality of our aspirations and desires gives us no ground to 
 
 hope they will be gratified ....... 56 
 
 48. Nor does their moral value, nor the fact that disbelief in them 
 
 would paralyse our activity ....... 58 
 
 49. Nor is there any reason to suppose that disbelief in any dogma 
 
 would make action either absurd or impossible . . .60 
 
 50. The argument that practice is supreme over theory . . .62 
 
 51. What consequences would result from the validity of the argu- 
 
 ment from consequences ?....... 63 
 
 52. The same subject continued ....... 65 
 
 D 
 
 53. The appeal to faith in matters of dogma on account of the 
 
 limitation of our knowledge 66 
 
 54. The objections to this 67 
 
 55. The appeal to faith in God's character. The limitations of this 
 
 appeal 69 
 
 56. But we have no reason to trust in God's goodness without 
 
 a demonstration which removes the matter from the sphere 
 
 of faith 70 
 
 57. The same continued 72 
 
 58. Can faith in God be based on love of God ? .... 72 
 
 59. Conclusion. The legitimate use of faith in relation to belief . 74 
 
xii TABLE OF CONTENTS 
 
 CHAPTER III 
 
 HUMAN IMMORTALITY 
 
 SECT. PAGE 
 
 60. Introductory 77 
 
 61. Reasons for doubting Immortality 78 
 
 62. Is the Self an activity of its body? The facts could be explained 
 
 on a Dualistic hypothesis 79 
 
 63. But there is a tendency to adopt a Monistic explanation, and to 
 
 prefer Materialism to Idealism 80 
 
 64. Reasons for the preference of Materialism 81 
 
 65. These reasons entirely depend on the possibility of Matter existing 
 
 independently of Spirit 83 
 
 66. My belief in the independent existence of Matter does not arise 
 
 as an inference from my sensations. But it requires justification 84 
 
 67. And it does not seem possible to justify it except by an inference 
 
 from my sensations , , . 86 
 
 68. It may be admitted that my sensations require part-causes which 
 
 exist independently of me 88 
 
 69. But there is no reason to hold that the causes possess the Primary 
 
 Qualities, since a cause need not resemble its effect , . 89 
 
 70. And, on the theory that our sensations are caused by indepen- 
 
 dently-existing Matter, the Matter is admitted not to resemble 
 all its effects 89 
 
 71. The inconsistency of this would be diminished if the Secondary 
 
 Qualities were attributed to Matter. But this would make 
 the independent existence of Matter less plausible . . .91 
 
 72. And it would not entirely remove the inconsistency . , .92 
 
 73. Thus the independent existence of Matter is, at the best, a per- 
 
 fectly gratuitous hypothesis 94 
 
 74. If we reject it, we do not find ourselves in conflict with everyday 
 
 experience 95 
 
 75. Nor with Science 96 
 
 76. Nor need our rejection of it lead us into Solipsism or Agnosticism 98 
 
 77. The Self, then, cannot be one of the activities of its own body , 99 
 
 78. And we have thus disproved a theory whose truth would have 
 
 been fatal to Immortality 101 
 
 B 
 
 79. But is the Self of such a nature that it cannot exist without 
 
 a body? Reasons for this view 103 
 
 80. But a Self which had left one body might pass to another , . 104 
 
 81. And the reasons for supposing a body to be indispensable to 
 
 a Soul are not conclusive 104 
 
TABLE OF CONTENTS xiii 
 
 SECT. PAGE 
 
 82. The bearing of apparitions of the dead on the question of 
 
 Immortality 106 
 
 C 
 
 83. Does the transitory nature of mateiial objects give any ground 
 
 for doubting our Immortality ? But what is transitory in those 
 objects is merely combinations 107 
 
 84. And the Self is not a combination 108 
 
 85. The arguments of this chapter fall far short of justifying a 
 positive belief in Immortality 109 
 
 CHAPTER IV 
 
 HUMAN FEE-EXISTENCE 
 
 86. Introductory 112 
 
 A 
 
 87. If Immortality is true, it is probable that Pre-existence is true. 
 
 Any directly metaphysical argument which proves one would 
 probably prove the other 113 
 
 88. While arguments that Immortality must be true because of its 
 
 goodness cannot be accepted as valid 114 
 
 B 
 
 89. If Pre-existence is true, it renders the doctrine of a Plurality of 
 
 Lives more probable 116 
 
 90. This would, indeed, be in any case the most probable form of the 
 
 doctrine of Immortality, whether our life after death remained 
 
 on the same level as at present, 116 
 
 91. Or whether it rose or sank 117 
 
 92. Especially when we consider the fragmentary character of our 
 
 present lives 119 
 
 93. But Pre-existence would increase the probability of such a 
 
 plurality 119 
 
 94. The theory of Pre-existence would offer an explanation of cex-tain 
 
 personal relations 120 
 
 95. And of certain innate tendencies and qualities .... 121 
 
 96. These tendencies and qualities might perhaps be explained, 
 
 though scarcely as well, by Heredity 122 
 
 97. The inheritance of ancestral qualities is not inconsistent with 
 
 the doctrine of Pre-existence 124 
 
 98. The same continued 124 
 
 99. Any difficulties which arise from this explanation would arise 
 
 equally if we rejected Pre-existence 126 
 
xiv TABLE OF CONTENTS 
 
 c 
 
 SECT. PAGE 
 
 100. Is Pre-existence desirable? It lias been asserted that Immor- 
 
 tality cannot be said to exist if memory ceases at death . 127 
 
 101. This does not seem justifiable 128 
 
 102. But would Immortality in this case have any value ? The loss 
 
 of memory need not prevent other lives from being as good 
 as this 130 
 
 103. But will it prevent progress ? What is the value of memory 
 
 for progress ?.....,.,., 131 
 
 104. Memory is not essential to ensure that we should profit by the 
 
 past in the matter of wisdom 132 
 
 105. Nor in the matter of virtue 133 
 
 106. There remains love 133 
 
 107. May we expect to meet in other lives the people whom we have 
 
 loved in this? 134 
 
 108. If certain conditions are true, we may. (And, if those conditions 
 
 were not true, we should be no better off if memory persisted) 135 
 
 109. And in that case the efi'ect of love in one life may be preserved 
 
 in another life 137 
 
 110. The prospect of loss of memory will always appear to be a loss 
 
 — but less so in proportion as we understand life better . 137 
 
 111. The prospect of a Plurality of Lives in the future ofi'ers certain 
 
 attractions 138 
 
 CHAPTEK V 
 
 FREE WILL 
 A 
 
 112. Freedom of Self-Determination 140 
 
 113. Freedom of Self-Direction 141 
 
 114. Freedom of Self-realization 141 
 
 115. Freedom of Indetermination, commonly called Free Will. , 142 
 
 116. The position of the Indeterminist towards the Laws of Causality 
 
 and of the Uniformity of Nature 143 
 
 117. The arguments used by Indeterminists. The assertion of an 
 
 immediate conviction that the Will is Free .... 145 
 
 B 
 
 118. The argument from our feeling of Freedom , . , . 146 
 
 119. This feeling can be accounted for by our possession of Freedom 
 
 of Self-Direction 147 
 
 120. And nothing is proved by our failure to discover all the deter- 
 
 minants in any particular case 148 
 
 121 . Nor by our power of acting in a way which does not give us the 
 
 greatest pleasure 149 
 
TABLE OF CONTENTS xv 
 
 c 
 
 SECT. PAGE 
 
 122. The argument from judgements of Moral Obligation . . 149 
 
 123. This is an argument from the validity of these judgements to 
 
 the validity of Indeterminism. The validity of judgements 
 
 of Moral Obligation may be admitted 150 
 
 124. They presuppose that something is such that its existence is 
 
 good or bad, and that some person can exercise by his will 
 some influence on the existence of that thing . . . 152 
 
 125. Determinism would not be inconsistent with either of these 
 
 propositions 153 
 
 126. Neither would Determinism be incompatible with the approval 
 
 or disapproval of the agent or his actions . . . .154 
 
 127. But would the approval or disapproval be such as to constitute 
 
 a judgement of Obligation ? 155 
 
 128. Determinism would not prevent us from placing right volition 
 
 above all other excellences, 156 
 
 129. Nor does it seem that we always do this ..... 157 
 
 130. Is Determinism incompatible with Responsibility? . . . 159 
 
 131. (a) Responsibility to our fellow men. Determinism does not 
 
 make rewards and punishments unreasonable . . . 160 
 
 132. And so does not destroy this responsibility .... 161 
 
 133. The Determinist attitude towards Punishment . . . .162 
 
 134. iV) Responsibility to God. If God is omnipotent we should 
 
 have no responsibility to him on the Determinist theory . 164 
 
 135. But we should also have none on the Indeterminist theoiy . 165 
 
 136. If God is not omnipotent we could be responsible to him on 
 
 the Determinist theory, and not on the Indeterminist . . 166 
 
 137. (c) Responsibility to self. This is not incompatible with Deter- 
 
 minism . . .168 
 
 D 
 
 138. The argument that Determinism would make all choice absurd 169 
 
 139. This confuses Determinism with Fatalism 170 
 
 140. The argument from the disastrous consequences of Determinism. 
 
 But Determinism does not have any tendency to diminish 
 virtue ........... 171 
 
 141. It may make us more pitiful towards sinners, but that is not 
 
 a disaster 172 
 
 142. Nor is the outlook to the future more gloomy for the Deter- 
 
 minist than for the Indeterminist . - . . . .173 
 
 143. Nor is it the case that Determinism would reduce us to the level 
 
 of machines .......... 175 
 
 144. And if some Determinist theologies have been horrible, it was 
 
 not their Determinism which made them so . . . . 176 
 
xvi TABLE OF CONTENTS 
 
 E 
 
 SECT. PAGE 
 
 145. We now pass to arguments against Indeterminism. The main 
 
 argument here would be the establishment of a universal Law 
 of Causality 177 
 
 146. Indeterminism, if true, would make morality invalid . .177 
 
 147. For we could not then logically speak of persons as good or bad, 
 
 except when actually doing good or bad deeds . , .178 
 
 148. And attempts to avoid this lead the Indeterminist into fresh 
 
 inconsistencies ......... 179 
 
 149. Indeterminism, if true, would make all predictions about human 
 
 actions absolutely unjustifiable 182 
 
 150. And so makes all non-vindictive punishment absolutely unjustifi- 
 
 able 184 
 
 151. Nor can even vindictive punishment be justified for the 
 
 Indeterminist unless the sin and the punishment are 
 simultaneous 184 
 
 CHAPTER VI 
 
 GOD AS OMNIPOTENT 
 
 A 
 
 152. Definition of God 186 
 
 153. Disadvantages of the Spinozistic and Hegelian use of the word 187 
 
 154. Different conceptions of God 188 
 
 155. Reasons for believing in an Omnipotent God. Several reasons 
 
 frequently given have already been discussed in Chapter II . 189 
 
 156. The argument for the necessity of a first cause. If valid at all, 
 
 it would not prove a first cause who was God . . . 190 
 
 157. But does it prove a first cause at all ? It is said that an infinite 
 
 regress of causes is impossible . . , . . .191 
 
 158. Granting this, yet we are not entitled to accept the idea of a 
 
 first cause, either existing in time, 192 
 
 159. Or of a timeless nature 194 
 
 160. If, therefore, an infinite regress of causes is impossible we must 
 
 find some third alternative 196 
 
 161. The argument from design, as stated by Kant .... 196 
 
 162. This argument takes much of the universe not to be good as 
 
 an end, but only as a means 198 
 
 163. Kant points out that this argument cannot prove a creative God 199 
 
 164. We may add that, if it is valid at all, it would actually disprove 
 
 the Omnipotence of God 201 
 
 165. The argument that an omniscient being is necessary could not 
 
 prove that he was also Omnipotent 202 
 
TABLE OF CONTENTS xvii 
 
 B 
 
 SECT. PAGE 
 
 166. Is Omnipotence compatible witli Personality? In the first 
 
 place, the conception of an Omnipotent will seems contra- 
 dictory 202 
 
 167. Then, we know of no person who exists without an Other . 204 
 
 168. An Omnipotent person, however, must be capable of existing 
 
 without an Other, 205 
 
 169. Even an Other which was not existent 206 
 
 170. This, however, does not amount to an absolute disproof of the 
 
 existence of an Omnipotent person . , • . . 208 
 
 171. Is Omnipotence compatible with Goodness? Some evil certainly 
 
 exists 208 
 
 172. An Omnipotent being must have permitted this evil . . . 210 
 
 173. To cause or permit evil is often justifiable in a being of limited 
 
 power 210 
 
 174. But the justification cannot apply to an Omnipotent being , 211 
 
 175. How, then, can he be called good ? The theory of Pascal and 
 
 Mansel 212 
 
 176. This theory considered 213 
 
 177. The same subject continued 215 
 
 178. The theory that God, though Omnipotent, cannot do certain 
 
 things 216 
 
 179. The disadvantages of calling such a being Omnipotent . . 218 
 
 180. The theory that God is the sole reality 220 
 
 CHAPTER VII 
 
 A NON-OMNIPOTENT GOD 
 
 181. Introductory 221 
 
 A 
 
 182. A Creative God. He may or may not be free to abstain from 
 
 creating 223 
 
 183. Can a Creative God, who has created this universe, be good ? . 224 
 
 184. God's Antecedent and Consequent Volition .... 225 
 
 185. Does the evil in the universe thwart God's Antecedent Volition ? 
 
 If not, he cannot be good 226 
 
 186. It is possible for men to will Consequently what is repugnant 
 
 to their Antecedent Volition 227 
 
 187. But the impotence of a man to do any action always requires 
 
 for its explanation something beyond his own nature . . 228 
 
 188. Now when a creative God creates, nothing but himself exists. 
 
 If there is anything he cannot do it must be explicable 
 entirely from his own nature. Is it possible that his will 
 should be thwarted? 229 
 
 b 
 
xviii TABLE OF CONTENTS 
 
 SECT. PAGE 
 
 189. It is not legitimate to answer that God's will is thwarted by the 
 
 law of his nature, 231 
 
 190. Or by the limitation of his power 232 
 
 191. These considerations do not render a creator impossible, but 
 
 they render it impossible that he should be a God , . . 233 
 
 192. The theory that a non-omnipotent God is the sole reality . . 234 
 
 B 
 
 193. The theory of a non-creative God 235 
 
 194. Its superioiity to the theory of an omnipotent God or of a 
 
 creative non-omnipotent God ....... 235 
 
 195. But can the existence of such a God be proved? . . . 237 
 
 196. The argument from design 238 
 
 197. If Matter exists, can the traces of order in it come by Chance ? 239 
 
 198. What is meant by Chance 240 
 
 199. The improbability of Chance producing certain results . . 241 
 
 200. Chance is improbable as an explanation of traces of order in 
 
 the material world, though not so improbable as has been 
 supposed, 242 
 
 201. Especially considering the possibility of only a very small part 
 
 of the universe showing such traces of order .... 243 
 
 202. If Spirit is the only reality, we may conceive the universe (a) as 
 
 a unity in which selves are united by laws of a mechanical 
 nature 245 
 
 203. In this case there would be still some difficulty in dispensing 
 
 with the idea of a directing mind, though not so much as if 
 the existence of Matter was admitted 246 
 
 204. But if we conceive the universe (b) as a unity which possesses 
 
 spiritual significance and value, there is no need of a directing 
 mind to account for the traces of order in it . . , . 247 
 
 205. The existence of such a unity would then be a fundamental fact 
 
 of the universe ......... 248 
 
 206. Such a unity— if it were proved to exist — would not disprove 
 
 the existence of a directing God, but would deprive the argu- 
 ment from design of all force 249 
 
 207. Thus Idealism does not necessarily work in favour of Theism . 250 
 
 208. The argument for the necessity of an omniscient being . . 250 
 
 209. But this argument is only valid as against one form of Idealism 251 
 
 210. If the existence of a Director of the universe was proved, he 
 
 could not be perfectly good 253 
 
 211. And he might be very bad, 254 
 
 212. Or, if not very bad, no better than the average man . . 256 
 
 213. Or there might be a plurality of such directors .... 257 
 
 214. And there are certain possibilities, pointed out by Hume, which 
 
 we cannot refute 258 
 
 215. Conclusion 259 
 
TABLE OF CONTENTS xix 
 
 CHAPTER VIII 
 
 THEISM AND HAPPINESS 
 SECT. PAGE 
 
 216. Introductory 261 
 
 A 
 
 217. Our interest in God's works. We should gain nothing here 
 
 from the existence of an omnipotent God .... 263 
 
 218. Beliefs that we should do so are erroneous .... 264 
 
 219. Nor from the existence of a non-omnipotent Director of the 
 
 universe . 265 
 
 220. Nor from the existence of a non-omnipotent Creator . . 267 
 
 221. The only chance lies in determining how far the Nature of the 
 
 universe permits a non-omnipotent God to be successful in 
 his efforts 268 
 
 222. And such a determination of the nature of the universe might 
 
 give us optimism without the belief in a God .... 269 
 
 223. It is often held that Theism by itself is an adequate basis of 
 
 optimism, and that the falsity of Theism would disprove 
 optimism. The latter proposition would be ti*ue except on an 
 Idealistic theory of the nature of reality, .... 270 
 
 224. Which is not usually held 272 
 
 225. The existence of God would not prove human immortality . 274 
 
 226. The same continued 275 
 
 227. The existence of a Creator would render human immortality less 
 
 probable 277 
 
 228. Nor have belief in God and belief in immortality been always 
 
 connected historically 278 
 
 B 
 
 229. We pass to our interest in God's person —first as an ideal which 
 
 may guide us 279 
 
 230. God could not, at present, be perfect, but he would be better 
 
 than any other person 280 
 
 231. A personal ideal, though not indispensable to morality, is often 
 
 useful 280 
 
 232. But such an ideal may be found among men .... 281 
 
 233. They will be less exalted, but their example will be more 
 
 applicable, for they will be better knovm, .... 282 
 
 234. And their circumstances will be more like ours, . . . 284 
 
 235. Even if God is conceived as incarnate 284 
 
 236. The same continued 286 
 
 237. A man may lose moral stimulus by losing the belief in God . 288 
 
 238. Or comfort in difficulties 289 
 
 239. And he will lose the power to worship, though not to reverence 289 
 
 240. Love 290 
 
XX TABLE OF CONTENTS 
 
 CONCLUSION 
 
 SECT. PAGE 
 
 241. The result of our inquiries has been almost entirely negative. 
 
 This is to be expected 291 
 
 242. No religion is justifiable unless it is based on metaphysics. 
 
 This involves that religion ib justified for few people . . 292 
 
 243. But this would not prevent the good in which religion believes 
 
 being true 293 
 
 244. And some results of the loss of religion would not be undesirable 294 
 
 245. And the extent to which it would be lost may be exaggerated . 295 
 
 246. Still, the result is tragical 297 
 
 247. But that does not disprove it 297 
 
SOME DOGMAS OF EELIGION 
 
 CHAPTEE I 
 
 THE IMPORTANCE OF DOGMA 
 
 1. By metaphysics I mean the systematic study of 
 the ultimate nature of reaHty, and by dogma I mean 
 any proposition which has a metaphysical significance. 
 Tliis may seem at first sight a paradoxical definition. 
 For dogmas are held, and disputed, by many people to 
 whom metaphysics are absolutely unknown. 
 
 EREATUM 
 
 Page 185, line 13, for determinist read, indeterminist 
 
 significance. For it decides, for him, a problem which 
 is unquestionably of a metaphysical nature — one of 
 those problems which must be dealt with in any 
 systematic study of the ultimate nature of reahty. 
 
 I believe that my definition of dogma is the one 
 which accords best with the ordinary use of the word. 
 Not all propositions about matters of fact are called 
 dogmas. We do not give the name to the law of 
 gravitation, nor to the statement that Waterloo was 
 fought in 1815.1 g^j; g^H smgIi statements as are 
 
 ^ The law of gravitation might he accepted by a Materialist or 
 a Dualist as an expression of the ultimate nature of reality. In 
 
XX TABLE OF CONTENTS 
 
 CONCLUSION 
 
 SECT. PAGE 
 
 241. The result of our inquiries has been almost entirely negative. 
 
 This is to be expected 291 
 
 242. No religion is justifiable unless it is based on metaphysics. 
 
 This involves that religion is justified for few people . . 292 
 
 243. But this would not prevent the good in which religion believes 
 
 being true 293 
 
 244. And some results of the loss of religion would not be undesirable 294 
 
 245. And the extent to which it would be lost may be esaggrerated . 295 
 
 246. Still, the result is tragical 297 
 
 247. But that does not disprove it 297 
 
SOME DOGMAS OF RELIGION 
 
 CHAPTEE I 
 
 THE IMPORTANCE OF DOGMA 
 
 1. By metaphysics I mean the systematic study of 
 the ultimate nature of reality, and by dogma I mean 
 any proposition which has a metaphysical significance. 
 Tliis may seem at first sight a paradoxical definition. 
 For dogmas are held, and disputed, by many people to 
 whom metaphysics are absolutely unknown. 
 
 But we must remember that a proposition which 
 has metaphysical significance may be held indepen- 
 dently of metaphysical considerations. If a man 
 asserts the existence of God because he accepts the 
 argument from design, then his belief in God's exis- 
 tence rests on a metaphysical basis. If he asserts the 
 existence of God because a priest has told him that 
 God does exist, then his beUef does not rest on a meta- 
 physical basis, but it nevertheless has a metaphysical 
 significance. For it decides, for him, a problem which 
 is unquestionably of a metaphysical nature — one of 
 those problems which must be dealt with in any 
 systematic study of the ultimate nature of reality. 
 
 I believe that my definition of dogma is the one 
 which accords best with the ordinary use of the word. 
 Not all propositions about matters of fact are called 
 dogmas. "We do not give the name to the law of 
 gravitation, nor to the statement that Waterloo was 
 fought in 1815.1 g^^ all g^^li statements as are 
 
 ^ The law of gravitation might be accepted by a Materialist or 
 a Dualist as an expression of the ultimate nature of reality. In 
 
 B 
 
2 SOME DOGMAS OF EELIGION 
 
 found in the Christian creeds are called dogmas. And 
 the name would be extended to statements of the 
 same nature found in other creeds, and to the denials 
 of these statements. ' Jesus is the Son of God' and 
 ' Mohammed is the Prophet of God ' are both dogmas. 
 And so are ' there is a God ' and * there is no God '. 
 
 2. It may be objected that many creeds contain 
 statements which would be called dogmas, but which 
 are assertions of events as historical, and are not meta- 
 physical at all. Such are, for example, the crucifixion 
 and resurrection of Jesus, in the Christian creeds. 
 Now it is quite true that, in asserting the death of 
 Jesus, we are asserting a historical event, as much as 
 if we were asserting the death of Bruno. But if it 
 were regarded as merely a historical event, it would 
 not be called a dogma. It has its place as a dogma in 
 virtue of the belief that this death had some unique 
 influence on the relation between God and man, and 
 for any person who did not believe this there could 
 be no reason for calling the assertion of the event 
 a dogma. Any proposition as to the relation between 
 God and man is clearly of metaphysical significance, 
 and if it is asserted that the relation was in any way 
 determined by a particular death, then the occurrence 
 of that death has a metaphysical significance. 
 
 Again, a proposition is usually called a dogma irre- 
 spective of the reason for which it is believed. The 
 assertion of God's existence is equally a dogma 
 whether the believer has arrived at it by argument, or 
 accepted it from tradition, or feels an instinctive and 
 irresistible conviction to believe it. 
 
 that case it would be a metaphysical proposition, and might, 
 I think, naturally be called a dogma. But the name would be 
 refused to the law of gravitation as it is held by science, for that 
 involves no metaphysical assertion and is accepted by Berkeleians 
 and Hegelians as well as by Materialists and Dualists. 
 
THE IMPOETANCE OF DOGMA 3 
 
 Now, since no proposition without metaphysical 
 significance is called a dogma, and since so many 
 which have that significance are commonly called by 
 the name, it seems to me that it is desirable to give 
 the name to all propositions with metaphysical signi- 
 ficance. It will then, no doubt, include many proposi- 
 tions which are not usually called dogmas, since the 
 word is generally used only in relation to religion, 
 and many metaphysical propositions have but little 
 bearing on rehgion. But it would be unsatisfactory 
 to confine the use of the word to those metaphysical 
 propositions which bear on religion, because the in- 
 fluence of metaphysical propositions on rehgion has 
 many degrees of intensity, and it would be very 
 difficult to draw the line between what was dogma 
 and what was not, 
 
 3. I propose, therefore, to class as dogmas all pro- 
 positions which have any metaphysical significance, 
 and to define as religious dogmas those whose accept- 
 ance or rejection by any person would alter his 
 religious position. Thus the existence of a personal 
 God is a religious dogma, for the religion of a man 
 who believes it cannot resemble very closely the 
 religion of a man who disbelieves it. One may be 
 as religious as the other, but it will be in a different 
 form. On the other hand, the existence of matter is 
 a dogma which does not seem to me to be religious. 
 I do not see that the acceptance or rejection of it need 
 involve any change in religious attitude. 
 
 How then shall we define religion? Religion is 
 clearly a state of mind. It is also clear that it is not 
 exclusively the acceptance of certain propositions as 
 true. It seems to me that it may best be described 
 as an emotion resting on a conviction of a harmony 
 between ourselves and the universe at large. 
 
 B 2 
 
4 SOME DOGMAS OF EELIGION 
 
 Any definition less wide than this would be too 
 narrow. The word religion is habitually used of the 
 traditional national systems, such as those of the Hin- 
 doos, Greeks, and Eomans ; it is used of the revealed 
 systems of religion ^ ; and it is used of the attitude 
 of various people who do not accept any of them. 
 Plato, Spinoza, and Hegel would all, I suppose, be 
 called religious men. 
 
 If the word is to be used in this way, it is clear that 
 no one dogma can be regarded as essential to rehgion. 
 For example, the cases of Buddhism and of Spinoza 
 would prevent us from regarding behef in a personal 
 God or in personal immortality as essential to it. 
 
 4. Our definition, on the other hand, is not too 
 narrow, for it covers all the cases. In the case of the 
 early traditional religions, indeed, the harmony is only 
 rudimentary. In the early stages of these religions 
 men had scarcely formed the conception of the uni- 
 verse as a whole, and did not therefore ask the question 
 whether they were in harmony with it as a whole. 
 But their rehgion was based on theories which enabled 
 them to regard with more or less approval the part of 
 the universe in which they were interested. The gods 
 were regarded as having great power. They were also 
 conceived as being, in the main, Hke men, as having 
 the same ideals, sympathies, and standards as men. 
 In so far as they differed from men they were held to 
 be morally their superiors. (The actions reported of 
 them would, indeed, have often been judged very 
 wicked in men, but, when attributed to gods, they 
 were not thought wicked. This distinction is often 
 required in dealing with more developed religions.) 
 
 ^ For the sake of brevity, I shall use the term 'revealed 
 religions' for all those systems which are held, by those who 
 believe in them, to have been revealed. 
 
THE IMPOETANCE OF DOGMA 5 
 
 On the whole, then, the dogmas of these creeds led 
 to a belief that the world was to a large extent good, 
 and better than it would have been judged to be by 
 the same men if they had not held these dogmas. 
 These religions, therefore, were accompanied by a con- 
 viction of a certain harmony between the individual 
 and the universe. 
 
 We must also remember that it is just those reli- 
 gions in which the righteousness and power of the 
 gods is only rudimentary, which put most weight on 
 the possibility of propitiating the gods and influencing 
 their decisions by prayers and offerings. And it is 
 clear that the belief that we can induce the rulers of 
 the universe to comply with our desires will produce 
 a conviction of some harmony between ourselves and 
 the universe — though perhaps the harmony would not 
 be of the highest kind. 
 
 5. The position of magic, which appears to be the 
 historical predecessor of the reUgions we have men- 
 tioned, seems different. Magic rests on the belief that 
 it is possible by charms and rites both to compel men 
 and non-human spirits to do my will in certain ways, 
 and also to influence matter — either directly or through 
 those spirits — so as to cause it to conform to my 
 desires. The spirits affected by magic are not pro- 
 pitiated or persuaded — as in religion proper — but 
 compelled by the efficacy of the magic. 
 
 Here there is no recognition of the government of 
 the universe by beings better than om'selves — or at 
 least not worse. We do not get that harmony be- 
 tween ourselves and the universe which comes from 
 moral approval of the universe. But magic does assert 
 a possible harmony — though of the lowest kind — 
 between myself and the universe. For it asserts that 
 the nature of the universe is such that I can — if I 
 
6 SOME DOGMAS OF EELIGION 
 
 know the right trick — bend it more or less to my will. 
 And thus I can bring the universe more or less into 
 harmony Avith my desires.^ 
 
 6. It might be objected that Buddhism does not assert 
 any harmony between man and the universe, since it 
 teaches that all existence is evil. But it teaches, also, 
 that the constitution of the universe provides a way by 
 which it is possible for each of us to escape from the 
 tyranny of existence. With the wise and good man 
 the universe is in harmony. For he strives towards 
 the goal of Nirvana, and the universe is such that, if 
 he is wise enough and good enough, he will attain it. 
 
 7. Our definition, then, is broad enough for what- 
 ever is usually called religion. But attempts are 
 sometimes made to use the word in a much wider 
 sense, and it will be well to examine three of the most 
 important of these. 
 
 In the first place, it is suggested that religion 
 should be defined as identical with morality. The 
 best-known statement of this position is to be found 
 in the Epistle of St. James (i. 27), ' Pure religion and 
 undefiled before our God and Father is this, to visit 
 the fatherless and widows in their affliction, and to 
 keep himself unspotted from the world.' 
 
 This would, I think, be admitted not to be the most 
 usual meaning of rehgion. It is generally advocated, 
 indeed, as a change of meaning. It seems to me to be 
 a bad definition, because, with two things and two 
 words before us, it proposes to apply both words to 
 one thing, and to leave the other thing nameless. 
 
 To do what I think right is obviously something 
 very different from the possession of an emotion 
 which arises from a conviction of my harmony with 
 
 ^ Cp. Hegel's Philosophy of Beligion (Part II, Div. I, chap, i) 
 and Mr. Frazer's Golden Bough (chap, i, p. 63, ed. 1900). 
 
THE IMPOETANCE OF DOGMA 7 
 
 the universe. Nor are they always found together. 
 Hume and Spinoza were both men who did what they 
 thought right; but, while Spinoza was absolutely 
 dominated by such an emotion, Hume, so far as we 
 know, was absolutely devoid of it. 
 
 It is highly desirable that each of these separate 
 things should have a separate name. Now the first 
 has a name which is universally applied to it — 
 morality. And if we take rehgion in the more usual 
 sense, which distinguishes it from morality, then the 
 second will also have a separate name. But if rehgion 
 is to be taken as identical with morality, then the 
 first will have two equivalent names, and the second 
 will have none at all. This is surely wasteful. 
 
 8. The definition of rehgion as nothing but morahty 
 has had supporters worthy of much respect, but I think 
 we can see that its value — for it undoubtedly had 
 value — is due to a cause which is rapidly passing 
 away. This cause is the belief, which has been very 
 prevalent in the past, that it is a sin not to hold the 
 true rehgion. (The phrase ' a true rehgion ' is scarcely 
 accurate, since religion is not a system of propositions 
 but an emotion. But it is a usual and convenient 
 phrase for the religion which normally comes from 
 the acceptance of true dogmas, provided, of course, 
 that the true dogmas are such as to permit a religion 
 to be based on them.) 
 
 The extreme form of this belief — and it is not yet 
 altogether extinct — is that any one who fails to attain 
 the true rehgion before the death of his present body, 
 will be tormented through unending time. This, how- 
 ever, is not so common as formerly. But the milder 
 idea still lingers that it is in some degree sinful to fail 
 in attaining the true religion. If this were true, we 
 should escape serious difficulties if rehgion was iden- 
 
8 SOME DOGMAS OF RELIGION 
 
 tical with morality. It is, of course, sinful to be 
 immoral. But it is absurd to represent as sinful all 
 failures to understand the nature of the universe. 
 Some such failures may be sinful, if they are due to 
 indolence or to cowardice. But others clearly are not. 
 They are due to intellectual errors. And to fail in 
 matters of such difficulty — matters in which either 
 Berkeley or Hume, either Aquinas or Scotus, must 
 have failed — is httle reproach to a man's intellect, 
 and none to his virtue. 
 
 But the idea that religion is something, the failure 
 to attain which is sinful, is d3dng out rapidly. To 
 attain the true rehgion would be regarded by most 
 people as good, and by many people as the highest 
 good. But it would now be generally recognized — 
 and certainly by all thinkers worthy of our serious 
 consideration — that it is not a duty to attain the true 
 religion, although, of course, it may be, and often is, 
 a duty to seek it. And then there is no reason why 
 we should not define rehgion to mean something 
 different from morality, while, as I have pointed out, 
 there are two reasons why we should do so — that it 
 does less violence to the ordinary usage, and that 
 it finds names for two things of great importance, 
 instead of for only one of them. 
 
 9, The second proposed definition of religion which 
 we must consider is not so wide as the first. Arnold 
 speaks of religion as ' morality touched by emotion '. 
 And the author of Natural Religion defines it as ' the 
 influence which draws men's thoughts away from their 
 personal interests, making them intensely aware of 
 other existences, to which it binds them by strong ties, 
 sometimes of admiration, sometimes of awe, sometimes 
 of duty, sonietimes of love '.^ 
 
 ^ Natural Beligion, Book II, chap, vi, p. 227 (ed. 1891). 
 
THE IMPORTANCE OF DOGMA 9 
 
 This definition seems to me better than that which 
 identifies rehgion with morality. It does not waste 
 two names on one thing, since morality touched by 
 emotion is something different from morality pure and 
 simple. Nor does it disturb the ordinary application 
 of the word so much, since it does not compel us to 
 class all moral men as religious. 
 
 Still it does disturb the ordinary application of the 
 word more than can be considered justifiable. Eeligion 
 has always, I think, implied a belief in some funda- 
 mental harmony, some sort of reconciliation between 
 the claims of our own nature and the facts of the 
 universe. Now an enthusiasm about morality, or 
 an enthusiasm about duty or love, does not by itself 
 ensure harmony or reconciliation. Enthusiasm for 
 any worthy ideal, whether fulfilled or unfulfilled, is 
 doubtless good. But unless the ideal is fulfilled, or 
 we believe it is going to be fulfilled, it does not bring 
 peace but a sword. The more we long for an ideal, 
 the less in harmony shall we be with a universe which 
 refuses to realize it. And therefore we shall not be 
 making the usual application of the word religion, if 
 we apply it to cases of an enthusiasm for an ideal 
 when behef in the realization of the ideal is absent. 
 
 This violation of usage may be productive of de- 
 ception. For the usage is old and persistent, and with 
 the word religion we have come to associate that 
 particular happiness which comes from the belief that 
 we are in harmony with the universe. And the 
 association is not easily disturbed, even when a change 
 in the meaning of the word has deprived it of all 
 justification. I think that those people who use the 
 word to mean nothing more than an enthusiasm for 
 virtue, often unconsciously deceive themselves by 
 transferring with the word the mystery and the rest- 
 
10 SOME DOGMAS OF EELIGION 
 
 fulness which only clung about it in respect of its older 
 meaning, and thus believe themselves to have kept 
 what, in fact, they have lost. 
 
 10. A third definition of religion has been lately 
 proposed by Mr. Lowes Dickinson.^ According to 
 his view, religion consists in the attitude of the spirit 
 towards the general situation in which we find our- 
 selves. So far I am able to agree with him. But he 
 goes on to say that any attitude towards the universe 
 is rehgious 'if it be greatly and imaginatively con- 
 ceived' — even if it should be an attitude which 
 includes condemnation and defiance of a universe 
 recognized as evil. 
 
 For such a religion, dogma, in the sense in which 
 I have used the word, would be necessary. "We must 
 know what the general situation is, before we can take 
 up any attitude towards it. But, on this definition, 
 religion can be based on any theory of the nature of 
 the universe. All that is required is that we should 
 take up a worthy and a dignified attitude towards the 
 universe. And that is always possible, however bad 
 or trivial the universe may be. 
 
 There is much to be said for this definition of religion, 
 but on the whole it seems to me to be more convenient 
 to keep to the common usage by which religion is 
 conceived as something which brings with it rest and 
 peace and happiness. ^ Now an attitude of defiance to 
 a universe recognized as evil is worthy and dignified, 
 but it is not restful, or peaceful, or felicific. And thus it 
 might lead to confusion if such an attitude were called 
 religion. The question, however, is of no importance for 
 
 ' Beligion, a Criticism and a Forecast, by G. Lowes Dickinson, 
 chap, iii, p. 60. 
 
 ^ I cannot consider Buddhism as an exception to this, for the 
 reasons given above (Section 6). 
 
THE IMPOETANCE OF DOGMA 11 
 
 our present object, since the adoption of Mr. Dickinson's 
 definition would leave dogma essential to religion — 
 though not dogma of any particular nature. 
 
 11. If religion is to be defined as resting on a convic- 
 tion of the harmony of ourselves with the universe, the 
 question arises of how much harmony is necessary. 
 If complete harmony were necessary, no one could be 
 called religious except the few mystics who deny the 
 existence of any evil, since any evil must involve some 
 want of harmony. We must say, I think, that religion 
 is a matter of degree. The more complete the harmony 
 it asserts, the more completely religious will it be. Of 
 course the more complete religion is not necessarily 
 the more true, for it may assert a harmony which 
 does not exist. 
 
 What then, we must now ask, is the minimum of 
 harmony between ourselves and the universe that 
 can be the basis of religion? Some minimum there 
 must surely be. For the conviction that the universe 
 was not quite as bad as it could possibly have been 
 would involve a belief in some harmony between our- 
 selves and the universe. But this is not sufficient for 
 rehgion. It would be an inadequate foundation for 
 a religion that every man is not always hungry. And 
 yet this shows some harmony between ourselves and 
 the universe. 
 
 It seems to me that the minimum harmony required 
 to give us an emotion which could now be called 
 religion, is that the universe should be judged to be 
 good on the whole. That is to say, the harmony must 
 be a harmony with what is judged by us to be the 
 highest part of our own natures, and not with those 
 of our desires which, even while we have them, we 
 recognize as comparatively worthless or as wrong. 
 And the harmony must be sufficiently complete to 
 
12 SOME DOGMAS OF EELIGION 
 
 admit of the universe ^s a whole being approved 
 rather than condemned. 
 
 I have said that this is necessary for an emotion 
 which could now be called religion. In the more 
 rudimentary stages of religion, no doubt, it is different. 
 A god is worshipped if he is believed to make things 
 rather better, or to abstain from making them worse, 
 for the worshipper and his neighbours. No question 
 is raised as to the universe as a whole, or as to the 
 balance of good and bad. And in a still more rudi- 
 mentary stage the question is only as to the satisfac- 
 tion of the desires of the individual, regardless of their 
 moral quality. 
 
 But religion has developed into a more exacting 
 form. It was impossible, as the conception of an 
 objective good developed, to find sufficient harmony 
 with our nature in anything which we condemned 
 as evil, even if it were personally convenient to our- 
 selves. And when the universe came to be regarded 
 as a whole — which was inevitable for monotheists — it 
 was impossible to find sufficient harmony in anything 
 less than the universe as a whole. When once the 
 objective and the universal had been realized, no 
 subjective or partial harmony could be adequate. 
 
 Thus the declaration of Mr. Kipling's dying girl, 
 ' I bear witness that there is no God but thee. Be- 
 loved', is not religion. It is rather its expHcit rejec- 
 tion in favour of something which is not religion — 
 though it may well be infinitely better. But the 
 belief that the love of man for man is God — which we 
 find in some mystical philosophers, and possibly in 
 the First Epistle of St. John — might be the basis of 
 a religion, for it is an assertion about the ultimate 
 nature of reality which might involve the consequence 
 of our harmony with the universe. 
 
THE IMPOETANCE OF DOGMA 13 
 
 12. If religion is to be defined in this way, or in any 
 other way which bases it on a behef in fundamental 
 harmony and goodness, I maintain that dogma is 
 necessary to religion. Dogma is not religion, any 
 more than the skeleton is the living body. But we 
 can no more be religious without dogma than our 
 bodies could live without their skeletons. 
 
 If we were to reject dogma, on what could we hope 
 to base rehgion ? It is said that we should still 
 have science, on the one hand, and, on the other, 
 virtue, with all the activities and pursuits the exercise 
 of which are considered virtuous. Will either of these 
 be sufficient to assure us that we are in a harmony 
 with the universe sufficiently fundamental to form 
 a basis for religion ? Let us take science first. 
 
 13. If we assert that science is capable of assuring 
 us of such a harmony, we assert that science can give 
 us information of the ultimate nature of reality. For 
 it is clear that science cannot assure us of our harmony 
 with the universe unless it is able to tell us what is 
 the nature of the universe. But if we assert this, then 
 we have not got rid of dogma. The fundamental pro- 
 positions of science have become metaphysical dogmas. 
 In themselves they are not so, for they make no claim 
 to express the ultimate nature of reality. Tliis is shown 
 by the fact that Materialists, Duahsts, Berkeleians, and 
 HegeHans, who differ so profoundly as to the ultimate 
 nature of reality, all accept the same system of scien- 
 tific propositions. But when these propositions are 
 taken as the ultimate truth about reality, they are 
 metaphysical. 
 
 'But', it may be replied, 'although these proposi- 
 tions may fall within the definition of dogma, they are 
 free from the defects which have rendered us un- 
 wiUing to declare dogma essential to religion. They 
 
U SOME DOGMAS OF EELIGION 
 
 have not the obscurity or the uncertainty of meta- 
 physics. They are given us by science, and what 
 science gives us is plain and certain.' 
 
 This, however, is a confusion. Science may give us 
 these propositions as scientific truths, but it cannot 
 assure us that they express the ultimate nature of 
 reality, for that, as was said above, does not concern 
 science at all. If they are taken as expressing that 
 ultimate nature, then the statement that they do ex- 
 press it must either be proved, or simply assumed. To 
 prove it we must go to metaphysics, for — to mention 
 one point only — it is impossible to refute the alterna- 
 tive theories as to the nature of reality except by 
 metaphysical arguments. And then our religion would 
 rest, not on science, but on a metaphysical system of 
 dogmas, which was what we had professed to avoid. 
 But if we do not prove that science expresses the 
 ultimate nature of reality, We have no right to beheve 
 that it does. For this is frequently denied, and it 
 cannot be said to be a seK-evident truth. 
 
 14. It is worth while, however, to consider another 
 point. If the result of our metaphysical consideration 
 was to convince us that the fundamental propositions 
 of science express the ultimate nature of reality, they 
 would be dogmas, but would they be such dogmas as 
 would serve for a basis for rehgion ? I do not think 
 that they would be, because they would not give us any 
 ground for beheving that the universe was good. 
 
 A certain part of the universe is what is called, in 
 ordinary language, matter. Now about this part of 
 the universe science tells U3 that it is governed by 
 certain laws which are of the type usually called 
 mechanical. And there science leaves it. It does 
 not inquire whether these laws are manifestations of 
 deeper and more fundamental laws, or whether they 
 
THE IMPORTANCE OF DOGMA 15 
 
 are due to the will of a creative God. It does not 
 deny that this is so ; it does not assert it. It has 
 nothing to do with the question. 
 
 But if the results of science are to be given meta- 
 physical vaUdity, then it would follow that these 
 mechanical laws express the ultimate nature of that 
 part of the universe of which science asserts them. 
 For we are to have no metaphysics now but the 
 results of science, and this is all that science tells us 
 about them. 
 
 The behaviour of matter is of great importance for 
 the question of the goodness of the universe. We 
 may hold, indeed, that the only thing which can be 
 intrinsically good or bad is conscious spirit, and that 
 matter, which is unconscious, can only be good or bad 
 as a means of producing good or bad results in 
 spirits. But, in that part of the universe which we 
 know empirically, matter is in close connexion with 
 spirit — so close that the good or bad state of spirit 
 must always largely depend on its relations with 
 matter. Pain, for example, may not be the only evil, 
 but it cannot be denied to be evil. And our relation 
 to matter is capable of causing us great pain — some- 
 times almost continuous for a long period. And there 
 is scarcely any quality, the possession of which has 
 ever been held to be good for men, the development 
 of which may not be prevented, or at least greatly 
 thwarted, by the condition of the body or its environ- 
 ment. Science gives us no reason to suppose that 
 this state of things will not be permanent. 
 
 Now have we any reason to beheve that the action 
 of matter, guided solely by mechanical laws, would be 
 such as to make the universe good ? It does not seem 
 to me that we have, in the first place, the slightest 
 right to believe this a priori. For there is nothing in 
 
16 SOME DOGMAS OF RELIGION 
 
 the general nature of such mechanical laws which 
 would tend to make them produce good more readily 
 than evil, or would show that they would produce 
 a universe good as a whole. It is entirely indifferent 
 to them which they produce. It is, of course, possible 
 that the action of those laws would produce such 
 a universe. But it is no more probable than the 
 contrary alternative, and the possibility must be at 
 least probable before we can base a religion on it. 
 
 Eevealed rehgions, as a rule, escape this difficulty 
 by the hypothesis of a beneficent deity, who arranges 
 and controls matter in such a way as to render it com- 
 patible with, and subservient to, the realization of 
 goodness in the universe. Some systems of meta- 
 physics escape it by maintaining that matter is essen- 
 tially of the nature of spirit, and will be found in the 
 long run harmonious to the demands of spirit. But 
 science knows nothing of such theories as these, and 
 if we are to make our metaphysics exclusively out of 
 science, such theories are not available for us. 
 
 15. Let us pass from a priori to empirical considera- 
 tions. There is no a priori reason why unloaded dice 
 should turn up double sixes rather than any other 
 number. And yet in a particular case it might be 
 possible to know empirically that they had turned up 
 double sixes. Is it possible to know empirically that 
 the universe is good as a whole ? 
 
 In the first place it does not seem possible to 
 know empirically anything about the universe as 
 a whole. All we can base such a judgement on is our 
 very imperfect knowledge of what has happened on 
 one planet for a few thousand years. To make an 
 empirical inference from so little to so much would be 
 as wild as to argue that all Chinese were evil because 
 the first one we met turned out to be so. 
 
THE IMPOETANCE OF DOGMA 17 
 
 And then, if we could legitimately make such an 
 inference, would the inference be that the universe as 
 a whole is good ? Is that part of the universe which 
 we know empirically, taken as a whole, good or bad ? 
 It certainly contains a great deal that is good. It also 
 contams a great deal that is bad. There is much 
 virtue and happiness. There is much sin and misery. 
 Which overbalances the other it would be impossible 
 to assert with any approach to acciu'acy. The judge- 
 ment of the wisest man would be unduly affected by 
 his own disposition, his own good fortune, and the 
 character of that small part of the world of which he 
 had personal experience. And we find that the judge- 
 ments of wise men on this question differ as widely 
 as possible. 
 
 Even if it were certain that the amount of good 
 experienced by different people in the world empiri- 
 cally known to us was greater than the evil, would it 
 follow that that world was to be considered as good 
 on the whole, and to be approved rather than con- 
 demned ? I do not think it would necessarily follow. 
 For good and evil are very unequally distributed, 
 and, whatever the aggregate of each may be, it cannot 
 be denied that there are some people whose lives are 
 much more evil than good. Now it seems to me that 
 we must recognize each individual as an end in him- 
 self; and as having claims to possess the good — claims 
 which cannot altogether be cancelled by any amount 
 of good possessed by other people. A universe in 
 which three people out of every five were in heaven 
 and two in every five in hell might have a greater 
 amount of virtue and happiness in it than of sin and 
 misery. But I do not think we should be prepared 
 to accept it as a good imiverse on that account. 
 
 If we take into account then, firstly, the great 
 
18 SOME DOGMAS OF KELIGION 
 
 amount of evil known to us, and, secondly, the great 
 excess of evil over good in the lives of many people, 
 as those lives are open to our empirical observation, 
 we must, I think, come to the conclusion that, if the 
 universe as a whole resembled what we can observe, 
 we should have no adequate basis for a religion. 
 
 Some metaphysicians cut this knot by denying the 
 reality of all evil, and by asserting that, in spite of 
 appearance, the imiverse is completely good. Whether 
 this is right or wrong, it is clear that it finds no sup- 
 port in science, and cannot form part of a meta- 
 physical system consisting of transplanted scientific 
 propositions. 
 
 16. There remains the view which, at any rate in 
 the Western world, forms the most usual basis of 
 rehgion — the view that things are improving, that the 
 evil in the future will be less than the evil of the 
 present, and (it is sometimes added) will eventually 
 vanish altogether. 
 
 Now can science offer us this comfort ? I do not 
 see how it can. It cannot say that the individual will 
 be happier in the future, for from the point of view of 
 science there is not the slightest ground for supposing 
 that we survive the deaths of our present bodies. 
 And, if there were, it would be no ground for hoping 
 that things would improve. The mere fact that life 
 goes on is no reason for supposing that it improves. 
 Some people are no better at sixty than at twenty. 
 Some people are worse. Any confidence that the pro- 
 longation of life beyond death would involve improve- 
 ment must be based on independent metaphysical 
 considerations. It would remain utterly ahen to 
 science, even if science had admitted that life was 
 prolonged after death at all. 
 
 So of the race. Even if science could give us 
 
THE IMPOETANCE OF DOGMA 19 
 
 ground to believe that the human race would con- 
 tinue to exist indefinitely, there would be no guarantee 
 that it would improve — indeed, no reason for sup- 
 posing that it was more likely to improve than to 
 deteriorate. Under certain circumstances races of 
 living beings do degenerate, and these circumstances 
 are just as likely to occur as more favourable ones. 
 
 But science gives us no ground to beheve that the 
 human race will continue to exist indefinitely. On 
 the contrary, I imagine that it insists that the tem- 
 perature compatible with human life is only a very 
 transitory episode in the history of a planet. Neither 
 of the individual nor of the race can it assert per- 
 manence. It reserves it for matter or for energy. 
 And the permanence of these is scarcely adapted to 
 afford us consolation. It is rather cold comfort — as 
 MiU is reported to have replied to an enthusiastic 
 Positivist — to look forward to our whole civilization 
 being eventually transmuted into an infinitesimal 
 augmentation of the temperature of space. 
 
 Nor could science comfort us by pointing out that 
 each man may deUver himself from unhappiness by 
 ceasing to exist. A metaphysic of scientific proposi- 
 tions would find no diflSculty in believing that every 
 man not only may, but does, cease to exist. And 
 this, if life is evil, would doubtless be a consolation, 
 for it would limit the amount of it which must be 
 endured by each man, and thereby make it more 
 tolerable. But while it would lessen the evil, the fact 
 would still remain that the universe contains more 
 evU than good, and we should not have an adequate 
 basis for rehgion since such a universe would scarcely 
 be called good as a whole. (See note, p. 37.) 
 
 17. It has sometimes been said that science, while 
 it cannot give us so good a rehgion as can be given us 
 
 C 2 
 
20 SOME DOGMAS OF KELIGION 
 
 by metaphysics or revelation, can, nevertheless, give 
 us some sort of religion, which is better than none at 
 all. The God revealed in science ' is also the God of 
 Christians. That the God of Christians is something 
 more does not affect this fact '.^ 
 
 In the first place, science does not reveal any God 
 at all. It is possible that the facts discovered by 
 science may give us a legitimate ground for inferring 
 the existence of a God, But the inference is not 
 science. It is as much metaphysics as an3rthing in 
 Kant or Hegel. 
 
 Let us pass by this objection, however, and let us 
 admit, for the purpose of the argument, that a God is 
 revealed by science. Such a God, I suppose, is to be 
 taken as the ultimate source of all the facts discovered 
 by science. It has been occasionally maintained that 
 God is only the part-cause of these facts, or only the 
 cause of some of them. But any discrimination of 
 this sort is obviously completely metaphysical. It 
 involves some knowledge of God derived from other 
 sources than the observations of scientific facts. If 
 science can be said to tell us anything about God at 
 all, it can only be as the being on whom all that is 
 studied by science depends. 
 
 Now such a God as this cannot, I think, be fairly 
 said to be the God of Christians. The God of 
 Christians is held by them — rightly or wrongly — to 
 embody the highest perfection which we can conceive. 
 But a God of whom we only know that he caused 
 those facts of which science tells us falls very short of 
 such perfection. It may be said that, since the idea 
 of a perfect God implies complete symmetry and 
 order, and since science informs us of a certain incom- 
 
 ^ Natural Eeligion, Book I, chap, i, p. 22 (ed. 1891). 
 
THE IMPORTANCE OF DOGMA 21 
 
 plete symmetry and order, the difference is from one 
 point of view quantitative. But this difference may 
 become quahtative in its effect upon reHgion. The 
 difference between a greater and a lesser amount of 
 food is quantitative, but it may involve a quahtative 
 difference of life and death. So the difference between 
 a being who fulfils our ideal of good and a being who 
 falls short of it is from one point of view merely 
 quantitative ; but it may be all the difference between 
 a God and a devil, and that, from the point of view of 
 religion, may be considered qualitative. 
 
 18. It is clear that, so far as science can tell us, the 
 good is often thwarted. In that case there are only 
 two alternatives as to the God revealed in science. 
 He may be utterly indifferent to good and evil. In 
 that case he is not good. His control of the universe 
 gives us no reason to suppose that the universe is 
 good on the whole. Neither does he deserve the 
 name of God — if that is to imply that he is a fit object 
 of worship and reverence. For what quaUty would 
 such a being have which would deserve worship ? He 
 would be, no doubt, stronger than we are. And if we 
 suppose him to possess a Hvely, and not very dehcate, 
 sense of personal vanity, it might doubtless be pru- 
 dent to worship him. But prudence seems the only 
 merit that such worship would possess. 
 
 Or, on the other hand, the God of science may have 
 regard for good and evil. In that case he ordained 
 the evil, knowing that it was evil. And such a God 
 as this has no hmitations, which might have compelled 
 him to ordain the evil as the only means to avoid 
 worse evils. He was defined as the source of all that 
 is known to science, so that he cannot be limited in 
 any way known to science, and, as we have excluded 
 all means of knowledge except science, we can have 
 
22 SOME DOGMAS OF EELIGION 
 
 no reason to suppose him limited by anything else. 
 We must suppose, therefore, that he put the evil 
 there because he Hked it. In which case the God of 
 science seems to bear a close resemblance to the devil 
 of theology. 
 
 He bears, indeed, a still closer relation to the God of 
 some theological systems. But, as we shall see in 
 Chapter VII, it is possible to construct a theology 
 with a better God than this, while it seems impossible 
 that the God of science should be better than this. 
 Moreover, the God of theology is not the only alterna- 
 tive to the God of science, since it is quite possible to 
 have a religion without a God at all. 
 
 19. We cannot, then, base religion on science. Can 
 we base it on virtue ? We have already seen reasons 
 for deciding that neither virtue nor an enthusiasm for 
 virtue is religion. But it remains possible that reli- 
 gion may be based on virtue. If a virtuous man 
 could derive from his virtue, independently of any- 
 thing else, a feeling of harmony with the universe as 
 a whole, then virtue would be a sufficient basis for 
 religion. 
 
 But can we get any such feeling from virtue alone ? 
 I think not. MoraUty, no doubt, still remains binding 
 on us whether the universe is good or bad. The idea 
 of the good is vahd for me. If it is not vahd for the 
 universe, so much the worse for the universe. No 
 doubt, too, however much we believed that the stars 
 in their courses were fighting against us, we might 
 feel not only enthusiasm but a certain amount of plea- 
 sure in carrying out the demands of our own con- 
 sciences, and in striving to make the best of such 
 a very bad business as the universe would then be. 
 But it could not give us the sense of harmony with 
 the universe. It sounds plausible to say that the 
 
THE IMPORTANCE OF DOGMA 23 
 
 possession of virtue ought to console us for everything, 
 and that, therefore, we have only to be virtuous to 
 feel in harmony with the universe, since with the 
 virtuous man no element can disturb his happiness, 
 or, consequently, affect the harmony. But we must 
 remember that in proportion as we are devoted to 
 virtue we care for its success, and must therefore be 
 considerably affected by a world which denies success 
 to it. A man who was so extremely virtuous that the 
 defect of virtue in all its endeavours was a matter of 
 entire indifference to him, would scarcely be con- 
 sistent. And, again, not only the virtuous man him- 
 self, but other people also, are subject to misfortune. 
 A virtue which was so intense that it rendered us 
 indifferent to the sufferings of others might be held to 
 have passed into its opposite. The recognition of evil 
 is not, as we have seen, incompatible with reHgion, but 
 they can only be made compatible by convictions as to 
 the predominance of good which would involve dogma. 
 
 The difficulty cannot be avoided by saying that 
 these other people would have been happy if they had 
 been virtuous. This, if it were true, would only 
 deepen the tragedy, since it asserts that wherever we 
 find the evil of misery we find also the evil of sin. 
 Nor do we avoid it by saying that their wickedness is 
 their own fault, even if that were true. For the diffi- 
 culty is that there is evil in the universe, and that is 
 not removed by any decision as to what being in the 
 universe is responsible for the want. 
 
 Again, how are we to deal with beings whose un- 
 happiness cannot be removed by virtue because they 
 are not capable of it — at least, not yet ? Granted — and 
 it is granting a good deal — that the contemplation of 
 the moral imperative could solace a man in the spasms 
 of hydrophobia, it would be unreasonable to expect 
 
24 SOME DOGMAS OF EELIGION 
 
 such devotion to the ideal from a dog or a guinea-pig. 
 And then how am I to prevent their pain from 
 destroying the harmony? 
 
 Even if I were prepared to forgive the universe the 
 evil which I myself suffer, that would not alter the fact 
 that it was evil, and that the universe which inflicted 
 it was Hkewise so far evil. And, even if I have the 
 right to forgive the universe my own pain, it is clear 
 that I have no right to forgive it the pain of others. 
 Unless we have reason to believe that evil is out- 
 weighed by good we have no right to approve the 
 universe. And we can have no reason for believing 
 this which is not a dogma. 
 
 20. To pass from the general to the particular, it 
 has sometimes been asserted that dogma is not essen- 
 tial to the Christian religion. This behef takes two 
 forms. The first defines dogma in much the same 
 way as we have defined it, and declares that nothing 
 is essential to Christianity except its moral teaching. 
 The second form, as we shall see later, uses dogma in 
 a different sense. 
 
 Let us consider the first. The teachings of Jesus, 
 and the teachings of the Christian Church, each con- 
 tain precepts of moraHty. Why should we not assert 
 that nothing else but this moraHty is of the essence of 
 Christianity ? 
 
 In the first place, we should have to assert that the 
 vast majority of Christians — practically all of them — 
 had been entirely deceived about what was essential 
 to the religion in which they lived and died. We are 
 not here compelled to decide whether the tenets of the 
 foimder of Christianity or of the fully developed Church 
 are best entitled to be considered typical Christianity. 
 For here they are agreed. 
 
THE IMPOETANCE OP DOGMA 25 
 
 If we put no trust at all in the narrative of the 
 Gospels — a scepticism which would be highly un- 
 reasonable — we should know nothing about the 
 teaching of Jesus on any subject, morality included. 
 But if we do put any trust whatever in that narrative, 
 it is impossible to deny that Jesus held various dogmas 
 to be true and important. His disposition does not 
 seem to have been at all metaphysical, and the dogmas 
 he taught were comparatively few and simple, and 
 were advocated, as a rule, without any attempt at 
 proof. But it can scarcely be denied that he believed 
 in personal immortality. And it seems quite impossible 
 to deny that he believed in a personal God, and that 
 he believed that the relation of God to man could be 
 suitably expressed by the metaphor of fatherhood. 
 A man who held these dogmas, and who regarded 
 them as important — and if these dogmas are true they 
 must be important — was certainly not hving without 
 dogma. 
 
 It may be worth pointing out that the Sermon on 
 the Mount, which is sometimes referred to as the ideal 
 of undogmatic rehgion, in reaUty contains dogma in 
 almost every line. ' Blessed are they that mourn : for 
 they shall be comforted,' is not a moral precept at aU. 
 It commands nothing, and it forbids nothing. But it 
 does make a statement about events, and one which 
 cannot be verified by empirical experience, since many 
 people mourn for years, and die uncomforted. The 
 blessings on the merciful and the pure in heart may 
 be held to be indirectly ethical, since they certainly 
 seem to im;ply an injunction to be merciful and pure. 
 But they also assert that the merciful shall obtain 
 mercy, and that the pure in heart shall see God. These 
 are not ethical injunctions, but statements asserting 
 events. And they are not capable of empirical veri- 
 
26 SOME DOGMAS OF EELIGION 
 
 fication — indeed, in the sphere of our empirical observa- 
 tion they are not always true. 
 
 If these assertions about future events are to be 
 verified without empirical observation, they can only 
 be verified by deduction from the fundamental nature 
 of reality. (Jesus, I suppose, derived them from his 
 belief as to the nature of God.) In this case, they are 
 dogmas. 
 
 21. But, it may be answered, Jesus may have 
 believed dogmas, and yet not have believed them to be 
 essential. But essential to what ? There is certainly 
 no reason to suppose that he believed that a man 
 could not act morally, or please God, without holding 
 these dogmas. Indeed it seems tolerably clear that he 
 did not believe this. But the question is not whether 
 he beUeved that holding these dogmas was essential 
 to morality, or to God's favour, but whether he believed 
 that they Avere an essential part of his own teaching. 
 Will any one venture to suggest that Jesus would 
 have regarded his teaching as not materially altered, 
 if the belief in the existence and providence of God 
 had been withdrawn from it? 
 
 And here, at least, Christianity has followed its 
 founder. Dogma has not been, as it might be supposed 
 some people believed, the monopoly of schoolmen. 
 The simplest Christians have always believed that 
 there is a God, that good is more pleasing to him than 
 evil, that he watches and controls events. And all 
 Christians have held some dogma on the subject of 
 life after death — the great majority that such a life 
 awaits every man, a few that it is conditional only. 
 These are not all. Every Christian has always 
 believed more dogmas than these, though all of them 
 have not beheved the same. 
 
 The importance of the dogma, again, has always 
 
THE IMPORTANCE OF DOGMA 27 
 
 been at least as great as that of the moral teaching. 
 The love of God and the hope of heaven are both 
 based on dogma. And certainly moraUty has not 
 been a more potent element in the religious lives of 
 the majority of Christians than the love of God and 
 the hope of heaven. 
 
 22. It is, of course, possible to say that Jesus and 
 the Church were both mistaken as to the relative 
 importance of their own beHefs. It is, indeed, almost 
 impossible that they should have been deceived as to 
 the relative importance of those behefs for themselves. 
 At any rate they were more likely to know than any 
 one else can be. But it is possible to maintain that 
 they were wrong as to the true relative importance of 
 their beliefs, and that if they had seen more clearly 
 they would have seen that the belief in God and 
 immortality could not compare in importance with 
 the moral precepts which we owe to Christianity. 
 
 If any one held this view, he could doubtless say 
 that dogma was not essential to Christianity, provided 
 he realizes that the discovery was first made in the 
 nineteenth century, and has against it the almost 
 unanimous opinion of Christians. But in that case 
 Christianity will not be a religion. We saw reason 
 before to confine the name of religion to cases which 
 exhibited a conviction of harmony with the universe. 
 And mere moral teaching can never, as we have seen, 
 do this. It only teUs us what the good is. It does 
 not teU us how far the good is realized. And it is on 
 this that our conviction of harmony with the universe 
 depends. 
 
 It would, indeed, be absurd to adhere to a definition 
 of religion which excluded Christianity, as Christianity 
 is commonly understood. For when people speak of 
 religion they most certainly mean to include Christi-; 
 
28 SOME DOGMAS OF EELIGION 
 
 lanity. And any change in the definition which would 
 alter the application of the word so materially would 
 be intolerably inconvenient. But there is no reason 
 why we should fix on a definition to include the pro- 
 posed undogmatic Christianity. For this is something 
 quite new, quite unhke Christianity as commonly 
 understood, and quite unlike anything to which the 
 name of religion has been generally given. We sin 
 against no canon of definition, therefore, by saying 
 that, if this is the true Christianity, then the true 
 Christianity is not religion. 
 
 23. We pass to the second sense in which the 
 undogmatic character of Christianity has been main- 
 tained. In this case a religious dogma is not defined, 
 as we have defined it, as any metaphysical proposition 
 which has influence on the character of religion. 
 Dogma is sometimes defined as whatever propositions 
 of this kind the definer dislikes. Or it is defined as 
 being those propositions which are complicated, and 
 can only be understood by trained theologians. Or it 
 is defined as being those propositions on which Chris- 
 tians do not agree. And on such gi^ounds we find the 
 assertion made that Christianity does not depend 
 upon dogma — not because a belief, for example, in 
 God and in immortahty are not considered necessary 
 for Christianity, but because they are not considered 
 dogmas. 
 
 The first definition of dogma, by which it is merely 
 employed as a term of abuse for doctrines which are 
 regarded as false, may be dismissed without much 
 delay. Dogma has never meant this in the past, and 
 it does not mean this now with most people. To use 
 the word in this way creates great confusion, and has 
 sometimes the further disadvantage of creating a false 
 impression of impartiality. A man may be supposed 
 
THE IMPOETANCE OF DOGMA 29 
 
 to be asserting that certain teaching, for example, is 
 impartial, when his real meaning is only that he does 
 not consider it false. 
 
 Connected with this is the tendency to conjfine the 
 use of the word to such propositions as are asserted 
 without proof— a tendency probably due to the fact 
 that the adjective dogmatic has frequently this meaning. 
 But adjectives have often meanings which differ widely 
 from those of the corresponding substantives. And it 
 would be very inconvenient to call the same proposi- 
 tion a dogma, when no reasons are given for it, and 
 undogmatic, if an attempt is made to prove it. 
 
 24. Let us take the second suggestion, that dogma 
 should denote exclusively those propositions which are 
 comphcated and can only be understood by theologians, 
 The objections to this view are two. In the first place, 
 it has not been the most usual definition of the word 
 in the past, and would exclude much that is commonly 
 called dogma. In the second place, the propositions 
 which it excludes and the propositions which it in- 
 cludes are like one another in such important charac- 
 teristics that they ought to have some common name, 
 and no other common name seems suitable. 
 
 On this view, the existence of God would not be 
 a dogma, but the doctrine of the Trinity as expounded 
 in the Athanasian Creed would be a collection of 
 dogmas. The first of these is, at any rate, apparently 
 simple, while the second has not even the appearance 
 of simpUcity. And almost every one can attach some 
 meaning to the first, while the second would be 
 absolutely unintelligible to many people. But the 
 same differences exist between the multipHcation table 
 and an advanced treatise on mathematics, and yet both 
 are said to consist of mathematical propositions. The 
 assertion of God's existence and the Athanasian 
 
30 SOME DOGMAS OF RELIGION 
 
 doctrine of the Trinity resemble one another in being 
 assertions of a metaphysical nature, in having influence 
 on the religion of those who hold them, and in being 
 extremely controversial statements. It seems to me 
 that these similarities render it desirable that they 
 should have a common name. 
 
 But, even if we did define dogma like this, it would 
 only be with important reservations that we could say 
 that Christianity was undogmatic. In one sense it 
 would doubtless be so. Many people who would be 
 called Christians have either never heard of the com- 
 pHcated and technical dogmas, or have never attached 
 the slightest meaning to the words they heard. It is 
 clear, therefore, that these beliefs are not psychologically 
 essential to Christianity. But it would be admitted 
 on this theory that the behef in God's existence was 
 essential to Christianity. Now some people would 
 maintain, and some have attempted to prove, that the 
 existence of God would involve a contradiction, unless 
 the Athanasian theology were true. In this case, these 
 dogmas would be logically essential to Christianity. 
 In the same way, no one could maintain that Kant's 
 theory of space was psychologically essential to a behef 
 in the propositions of geometry, since it is notorious 
 that many people accept the truth of geometry who 
 have never heard of Kant's theory of space, or who 
 have deliberately rejected it. But, if Kant is right, 
 a belief in his theory of space would be logically es- 
 sential to geometry, because^ as he maintains, it is 
 impossible that geometry should be true unless his 
 theory of space were true also. 
 
 25. It remains to consider the view that dogma 
 should be defined as that on which Christians do not 
 agree. Here, once more, we must object that this is a 
 departure from usage. Dogma has not meant this, and 
 
THE IMPOETANCE OF DOGMA 31 
 
 there is no reason why it should mean it now, unless to 
 gratify the feelings of people who have a vague idea 
 that it is hberal and tolerant to use dogma as a term of 
 abuse. And then, why draw this precise line ? K the 
 existence of God is not a dogma, because all Christians 
 beheve it, why should we not say that transubstantia- 
 tion is not a dogma, because all Romanists beheve it ? 
 But if transubstantiation is a dogma, because some 
 Anglicans do not beheve it, why should we not say 
 that the existence of God is a dogma, because there 
 are Atheists who do not beheve it ? 
 
 If dogma were defined like this, would it be essential 
 to Christianity ? A dogma on which all Christians do 
 not agree is clearly not psychologically essential to 
 Christianity, since some Christians do not believe it. 
 But, as in the last case, it may be logically essential 
 to Christianity. For there may be a contradiction 
 between asserting some of the admittedly essential 
 doctrines (which on this view would not be called 
 dogmas) and rejecting this dogma. 
 
 Even on the psychological question we must make 
 a certain reservation. We often hear the argument 
 that, since two men can both be Christians although 
 they hold incompatible dogmas on a certain point, 
 therefore the point is quite indifferent to Christianity, 
 and a man may be a Christian without troubling him- 
 self about it all. The conclusion may be correct — and 
 often is. But it does not follow from the premises, 
 and may be false in cases where the premises are true. 
 Because A can be either B or (7, it does not follow that 
 it can be neither. A human being need not be male, 
 for it may be female, nor need it be female, for it may 
 be male. But it woxild be a mistake to argue from this 
 that it could be sexless. Some Christians say that 
 man can be pleasing to God only by his works, some 
 
32 SOME DOGMAS OF EELIGION 
 
 that he can be pleasing to God only by his faith, and 
 some hold other views on the matter. None of these 
 theories, then, can be psychologically essential to 
 Christianity. But it does not follow that a man could 
 be called a Christian, who had no belief as to the way 
 in which man could please God, 
 
 26. Let us now return to the more general question. 
 I have endeavoured to show that if religion and dogma 
 are defined in a manner which accords with their 
 general use, we must hold that, while dogma is by no 
 means identical with rehgion, it is absolutely necessary 
 to rehgion. This conclusion could be avoided by taking 
 different definitions of rehgion and dogma. But the fact 
 would still remain, that our beliefs on metaphysicalsub- 
 jects are of supreme importance for the determination 
 of our attitude towards reality in general,, and towards 
 our own lives in particular, and are therefore, for many 
 people, of supreme importance for their happiness. 
 
 It will depend on those beliefs, whether we shall 
 consider the universe as determined by forces com- 
 pletely out of relation with the good, or whether, on 
 the contrary, we may trust that the dearest ideals and 
 aspirations of our own nature are realized, and far 
 more than realized, in the ultimate reahty. It will 
 depend on them whether we can regard the troubles of 
 the present, and the uncertainties of the future, with 
 the feelings of a mouse towards a cat, or of a child 
 towards its father. It will depend on them whether 
 we look on our pleasures as episodes which will soon 
 pass, or on our sorrows as delusions which will soon be 
 dispelled. It will depend on them whether our lives 
 seem to us worth living only as desperate efforts to 
 make the best of an incurably bad business, or as the 
 passage to a happiness that it has not entered into our 
 
THE IMPOETANCE OF DOGMA 33 
 
 hearts to conceive. It will depend on them whether 
 we regard ourselves as temporary aggregations of 
 atoms, or as God incarnate. These questions are not 
 devoid of practical importance. 
 
 There are, indeed, people whom they do not interest, 
 but they are in a minority. Experience shows us that 
 such questions as these are of great interest to most 
 people, nor is there any reason to suppose that the 
 extent or the degree of that interest is lessening. The 
 world of our empirical knowledge is no more capable 
 of satisfying us than it used to be. It is probably less 
 capable, for though it has doubtless improved, it has 
 not kept pace with our increasing demands for im- 
 provement. And while we want more than we can 
 find around us, questions of dogma will retain their 
 influence on the happiness of mankind. 
 
 27. It is common to speak of metaphysical problems 
 as abstract and unpractical. In reality, all other 
 questions are abstract as compared with these, and 
 most, as compared with these, are unpractical. Hence, 
 indeed, arise many of the difficulties of metaphysics. 
 If it progresses more slowly than science, it is often 
 because science, by its comparative abstraction, gains 
 in ease and simplicity what it loses in absolute truth. 
 And often, again, it is because our dearest hopes hang 
 on the answers to metaphysical questions, so that we 
 are afraid to seek those answers, or to look them in 
 the face when they present themselves, or to allow 
 other people to face them. 
 
 28. The dependence of religion on dogma condemns 
 religious questions to remain controversial and doubt- 
 ful. If religion could be based on science matters 
 would be different. In science we find a consensus of 
 expert opinion, and we find progress. There is, of 
 course, a margin of uncertainty in science. There are 
 
 D 
 
34 SOME DOGMAS OF KELIGION 
 
 always questions which have been raised and not yet 
 answered, or to which we find conflicting, and so 
 uncertain, answers. But problem after problem is 
 solved, and the solution becomes part of the common 
 and undoubted property of mankind. The advance of 
 science and the certainty of its results are beyond all 
 doubt. What it all means is another matter, but we 
 cannot deny that it is there to mean something. 
 
 With morality, again, the certainty, though not so 
 striking as with science, is still very marked. It may 
 be sometimes exaggerated. But if we compare the 
 agreement of men's opinions on any metaphysical 
 question — the existence of God, of matter, of immor- 
 tality — with the agreement as to the general duty, 
 and even the details, of honesty, of truthfulness, of 
 courage, we shall find the balance overwhelmingly in 
 favour of the latter. Indeed, the possibility of civilized 
 life proves that the general agreement as to morality 
 must be considerable. 
 
 Compare all this with dogma. If we take the reli- 
 gions which claim to be revealed, we do not find 
 a single proposition laid down by any of them which 
 is not challenged by others, nor does there appear any 
 chance of discovering a common ground upon which 
 it would be possible to settle the dispute. And if 
 metaphysics is better off, it is still devoid of the agree- 
 ment without which we can have no certainty. It 
 advances, no doubt. But it only advances by changing 
 the battle-ground, not by settling any problem finally. 
 The questions evolve into different forms, but the 
 answers are still various. We may hope that the long 
 contest will eventually develop into a form where 
 opposition will cease. But such a goal must at best 
 be very distant, and many — though I cannot agree 
 with them — fail to see any hope that it can ever be 
 
THE IMPOETANCE OF DOGMA 35 
 
 realized. For centuries to come we must resign our- 
 selves to the admission that where we have dogma we 
 shall have division, 
 
 29. Dispute without any confident hope of recon- 
 cihation is always bad — worst of all when, as must so 
 often be the case in subjects bearing on rehgion, it 
 implies a certain spiritual discord. If we could put 
 dogma altogether on one side, or confine it to the 
 studies and lecture-rooms of theologians and philoso- 
 phers ; if we could say that in spite of dogmatic 
 divisions and doubts we could still have rehgious 
 unity and certainty, still lack no element for a peaceful 
 and happy life, then the world might surely count 
 itself fortunate. But to do this is impossible, for it 
 would ignore something essential to religion, and, in 
 most cases, to a peaceful and happy hfe — the conviction 
 of our own harmony with the universe. 
 
 Dogma means, now at any rate, division. But it 
 may be doubted whether we shall get any unity worth 
 preserving by the process immortalized by Mr. Saunders 
 M^Kaye, which stripped mankind of their clothes, and 
 then proclaimed them brothers *on the one broad funda- 
 mental principle o' want o' breeks'.^ It happens that 
 the things about which we most disagree are the things 
 about which we most want to know, and a unity 
 which is attained by ignoring them is valueless. 
 
 It is not, of course, certain that dogma will deliver 
 us from our troubles. We cannot set out on an 
 inquiry and settle the answer beforehand. If we 
 begin to inquire into what lies behind phenomena, the 
 answer may be even more depressing than the super- 
 ficial aspect of the phenomena themselves. But the 
 only line of escape which is even possible hes in this 
 direction. The phenomena of hfe, as we see them, 
 
 ^ Charles Kingsley, Alton Locke, chap. xxiL 
 D 2 
 
36 SOME DOGMAS OF KELIGION 
 
 can never give us the assurance of harmony that 
 we want. To get this it is necessary to go behind 
 experience, whether we take the path of revelation or 
 of metaphysics. 
 
 It is here, as it seems to me, that the strength of the 
 revealed religions of the world lies. If true, they would 
 be adequate. They have arrived at the result that the 
 nature of the universe is something with which it is 
 possible that the wise and good man may find himself 
 in harmony — usually, of course, by means of the 
 dogma that the course of the universe is controlled by 
 a benevolent God. They may be entirely unjustified 
 in their conclusions. Their conceptions of divine 
 benevolence may be of the most remarkable nature. 
 But they change the whole aspect of heaven and earth 
 for those who believe in them. The gifts they offer are 
 worth taking — or at any rate seem so to those who are 
 able to take them. And so, for all their faults, they 
 possess considerable advantages over systems which 
 have nothing to offer except rhetorical embellishments 
 of ethical maxims. 
 
 30. The result is not entirely satisfactory. The 
 only roads by which dogma has been reached in the 
 past are revelation and metaphysics, and every year 
 fewer people appear willing to accept any system of 
 asserted revelation as valid without support from 
 metaphysics. Now every one who studies meta- 
 physics does not arrive at conclusions on which a reli- 
 gion can be based. And, even if they did, the study 
 of metaphysics is only open to those who have a cer- 
 tain amount of natural and acquired fitness for it. 
 The number of people who will be left between the 
 rapidly diminishing help of revelation and the possibly 
 increasing help of metaphysics seems likely to be 
 unpleasantly large. 
 
THE IMPOKTANCE OF DOGMA 37 
 
 But it is useless not to face the facts. If the supply 
 of bread runs short, we shall gain nothing by distri- 
 buting stones. Such a course may even produce two 
 positively evil results. It may persuade the ungrateful 
 recipients, not only that there is a deficiency of food, 
 but that there is no such thing as food at all. And it 
 may prolong the scarcity, or even render it per- 
 petual, by turning men's minds to quarries rather 
 than to wheat-fields, as the source from which may 
 arise some satisfaction for their desires. 
 
 Note to p. 19. — This position is different from the position of 
 Buddhism, I conceive that Buddhism regards Nirvana, in spite 
 of its negative nature, as true good and the only true good. But 
 the view we are now discussing does not take annihilation as 
 positively good, but only as the end of the series of good and bad 
 events. It cannot compensate for the evil which has been 
 suffered, but only prevents more from coming. 
 
CHAPTER II 
 
 THE ESTABLISHMENT OF DOGMA 
 
 31. In this chapter I shall consider, in the first place, 
 that belief which claims to dispense with argument 
 altogether, and to rest itself on the immediate convic- 
 tions of the believer. Secondly, I shall inquire into 
 the validity of the argument that a dogma of religion 
 is true because it is held by all, or by most, people, or 
 because it is held by persons who can work miracles 
 or predict the future. Thirdly, I shall inquire into 
 the validity of the argument that a dogma of rehgion 
 is true because its truth is of great importance for our 
 happiness, or for the moral character of the universe. 
 And finally, I shall consider the appeal which is often 
 made to leave questions of religion to faith. I shall 
 endeavour to show that none of these are valid, and 
 that, if dogmas are to be established at all, they must 
 be estabhshed by arguments, and by arguments 
 different from those dealt with here, 
 
 32. It is not uncommon to hear the assertion that 
 certain religious dogmas — the personality of God, for 
 example, or the immortality of man, or the freedom of 
 the will — do not require proof. 'I am certain of this ', 
 some one will say, * without argument. My conviction 
 does not rest on argument, and cannot be shaken by 
 it. I decline to argue. I simply believe.' A belief 
 which does not rest upon argument, in the case of any 
 particular person, may be said to be held by such 
 a person immediately, since argument is a process of 
 
THE ESTABLISHMENT OF DOGMA 39 
 
 mediation. The assertion here is, it will be noticed, 
 not only that the belief does not rest upon arguments 
 — i.e. that it is immediate — but also that it cannot be 
 shaken by arguments. 
 
 Such a position is, no doubt, impregnable from out- 
 side. If a man's belief does not rest on reasons, and 
 cannot be shaken by them, I may believe it to be 
 mistaken, but I should be wasting time in attempting 
 to argue against it. 
 
 If the person who holds a belief in this manner 
 mentions the fact to me as a reason why I should not 
 waste his time in trying to upset it, he is acting in 
 a perfectly reasonable manner. And it is also strictly 
 relevant to mention it if he is writing an autobiography 
 — for it may be an important fact in his life. Also it 
 is relevant as a contribution to statistics. It shows 
 that one more person has this particular conviction 
 in this particular way. 
 
 33. But it is not relevant if it is put forward for 
 any other motive. Above all, it is absolutely irrele- 
 vant if it is put forward as a reason to induce other 
 people to believe the same dogma. This is sometimes 
 done. A man wiU assert his own immediate convic- 
 tion of a dogma, ^ not as a reason for checking discus- 
 sion, but as his contribution to the discussion. And 
 here it seems certain that he is wrong. 
 
 What is the good of telling B that A has an 
 immediate certainty of the truth of X? li B has 
 a similar immediate certainty he believes X already, 
 and must believe it, and, for him, ^'s certainty is quite 
 superfluous. If B has an immediate certainty of the 
 falsity of X, or of the truth of something incompatible 
 
 ^ ' I am assuming for the present that an immediate conviction 
 is also one which cannot be shaken by arguments. We shall see 
 later on that this is not always the case. 
 
40 SOME DOGMAS OF EELIGION ' 
 
 ■with X, then he cannot believe X, and, for him, A's 
 certainty is quite useless. But supposing B has no 
 immediate certainty on the matter at all, how will his 
 knowledge of A's immediate certainty help him ? It 
 cannot give him an immediate certainty, for, if he 
 beheves because A does, his belief rests on an argu- 
 ment, ' A believes this, and therefore it is true,' and 
 so is not immediate. 
 
 But can A's immediate certainty be a valid ground 
 for a reasoned certainty on JS's part ? Why, because A 
 does believe anything, ought B to believe it ? That is 
 ultimate for A, but it is not ultimate for B, Why 
 should B accept this fact of A's nature as decisive, or 
 even in the least relevant, as to a truth which does 
 not relate to ^'s nature, but, for example, to the 
 existence of a personal God? 
 
 34. An attempt has been made to show why B 
 should do this — an attempt which seems to rest 
 entirely on an analogy. The people who have not 
 this immediate certainty are compared to the blind. 
 A blind man has no means of perceiving a balloon 
 which floats above him in the air. Yet he would be 
 mistaken if he disbelieved the statement of his friends 
 that the balloon was there. Similarly, we are told, 
 if another man has an immediate conviction of a propo- 
 sition, of whose truth I am not convinced at all, I ought 
 to supply the deficiency in my nature by taking on 
 trust from him what he perceives immediately.^ 
 
 But an analogy is good enough to meet an analogy. 
 
 ^ The example I have given seems to me to do the argument 
 more justice than the one usually taken, in which the blind man 
 denies the existence of colour. For most people, whether blind 
 or not, believe that colour does not exist except in the sensations 
 of those who see. If a blind man denied the existence of colour 
 in this sense, he would not be analogous to a person who denied 
 the truth of the belief of others, but to a person who denied that 
 the others had that belief. 
 
THE ESTABLISHMENT OF DOGMA 41 
 
 A man in delirium frequently believes that he sees 
 assassins lurking in a corner, or rats leaping on his 
 bed. He is as firmly convinced that he sees them as 
 I am that I see the balloon above me. His physicians 
 do not see them. Would they do well to believe that 
 they were there, but that some limitation of their own 
 faculties prevented their seeing them ? They do not 
 believe this. They do not send for the police to arrest 
 the assassins, or for a terrier to catch the rats. And it 
 would be generally admitted that they are right. 
 
 Now which is the more correct analogy here ? 
 When some people have an immediate conviction of 
 the truth of some particular dogma, and others have 
 not, are these others in the position of the blind man 
 or of the physician ? Such cases, it seems to me, are 
 settled, outside the sphere of rehgious dogma, in one 
 of two ways. Sometimes they are settled roughly by 
 counting heads. The blind are fewer in number than 
 those who can see. But only one man can see the 
 assassins or the rats. Any one else who enters the 
 room cannot see them. 
 
 If we were to decide on this plan there would be 
 little reason for any one else to beheve a dogma because 
 of ^'s immediate certainty of it. People who have, or 
 believe themselves to have, the immediate cei^inty of 
 a rehgious dogma are always comparatively rare — much 
 rarer than the people who beheve in the dogma. Let 
 us take, as an example, the existence of God. A great 
 majority of the inhabitants of the United Kingdom 
 accept this dogma. But I should say that by far the 
 greater number of them beheve in it for some reason, 
 good or bad, — the authority of the Church, the state- 
 ment of the Bible, the argument from design, or the 
 hke. The number of those who hold it because of 
 an immediate certainty would be but smalL 
 
42 SOME DOGMAS OF KELIGION 
 
 Moreover, the people who have such immediate 
 convictions agree very httle among themselves. Some, 
 for example, have immediate convictions on the 
 subjects of God, immortality, and free will. Others 
 have them on the subject of God alone, or of free will 
 alone, or of immortality alone. Others have them in 
 different combinations, and others on different subjects 
 altogether. If the sight-perceptions of mankind varied 
 as much as this, the blind would rightly decline to 
 put much faith in them. 
 
 35. But the test of counting heads, though some- 
 times the only possible test, is crude and unsatisfactory. 
 We have generally better reasons. A blind man has 
 good reasons for believing that other people have 
 sources of knowledge which he has not. They tell 
 him, for example, that a table is six feet in front of 
 him. He cannot perceive this at the time, but by 
 walking forward he can test it by touch, and he finds 
 it correct. When this has happened several times, 
 the hypothesis that other men have a trustworthy 
 sense, which he has not, becomes far more consistent 
 with his own experience than the hypothesis that 
 they are all labouring under a delusion. When the 
 dehrious man, pn the other hand, asserts that a rat 
 has come up through a solid floor, or that he is 
 menaced by a man who has long been dead, or who 
 is breatliing flame, the hypothesis that he sees some- 
 thing, to which other people are blind, would conflict 
 with the general fabric of experience far more than 
 the hypothesis that he is mistaken. 
 
 It is in this way that we ought to test the immediate 
 convictions on religious dogma which we do not happen 
 to share, before we decide whether to accept them as 
 the basis for a reasoned, non-immediate, belief on our 
 own behalf. But to do this is to inquire whether 
 
THE ESTABLISHMENT OF DOGMA 43 
 
 the existence of the object of the immediate belief 
 harmonizes better with our experience than the non- 
 existence of it. And when we make this the test, we 
 have really given up all rehance on A's immediate 
 conviction, and are endeavouring to support our belief 
 on the direct evidence for the truth of the proposition. 
 A's conviction, at most, suggests the dogma to us, if we 
 had not heard of it before ; the ground of our belief in 
 the dogma is no longer A's belief in it. 
 
 In the case of the blind man, the matter would be 
 different. The perceptions of sight of those who sur- 
 round him are very numerous, and when he has tested 
 them a certain number of times he can believe in the 
 rest without testing them. In these subsequent cases, 
 therefore, his belief is based on the perceptions of 
 others. But no man professes to have a very large 
 number of immediate convictions on rehgious 
 dogmas. It would thus be impossible to argue here 
 that a man's immediate convictions had been proved 
 right on so many points that they might be trusted on 
 the rest. They can only be trusted by us in the cases 
 in which they can be proved. And then our belief 
 rests entirely on the proof, and his immediate conviction 
 only furnishes the suggestion for us, and not the reason 
 for our belief. 
 
 36. Thus, even granting that A has an immediate 
 conviction of the truth of some religious dogma, it is 
 quite irrelevant to me, though decisive for him. But 
 it is by no means certain that A has an immediate 
 conviction when he thinks he has. In many cases — 
 though certainly not in all — it seems very probable 
 that he has not. A man may make mistakes in judging 
 and classifying what takes place in his own mind, 
 just as much as he may make mistakes about other 
 things. 
 
44 SOME DOGMAS OF RELIGION 
 
 For example, a man may be confused as to what it 
 is of which he is immediately certain. He may think 
 that his immediate certainty is of the existence of a 
 personal God, when, in his case, it may be only that 
 the ultimate reality is spiritual. That an ultimate 
 spiritual reality must necessarily be a personal God 
 may be a proposition which he behoves on account of 
 reasons — or which, perhaps, when clearly stated, he is 
 not prepared to accept at all. 
 
 Or, again, closer analysis may convince him that 
 the proposition, of which he supposed himself to have 
 an immediate certainty, is really dependent for him 
 on other propositions. If it can be proved to him that 
 he was not justified in basing it on those other pro- 
 positions — either because they are false, or because it 
 does not properly follow from them — he will abandon 
 the result which he had previously thought beyond the 
 reach of argument. 
 
 So far we have assumed that a behef which is im- 
 mediate — that is, which does not rest on arguments — 
 cannot be shaken by them. But an immediate behef 
 may rest on prejudices or tradition. (Of course, if 
 tradition is explicitly a<5cepted as likely to lead to 
 truth, then it is a reason for behef, whether it is a 
 good reason or not, and the conclusion is not imme- 
 diate. But when a man believes a tradition merely 
 because it has never occurred to him to question it, 
 then the tradition is not a reason for belief, though it 
 is a cause.) Now a belief of this sort, although it does 
 not rest on arguments, may be shaken by them. For 
 it may be shown that it is caused by prejudice or 
 tradition, and this demonstration — though it refutes 
 no arguments for the behef, since there were none to 
 refute — may cause the behever to change his opinion. 
 
 The result at which I arrive is that the statement, 
 
THE ESTABLISHMENT OF DOGMA 45 
 
 that any man has an immediate conviction on a matter 
 of religious dogma is one which he ought not to expect 
 to have any relevance for others, and which he ought 
 only to make, even for his own guidance, after careful 
 tests have convinced him, in the first place, of what 
 his belief really is, and in the second place, that it is 
 not based on arguments. Even then, he ought not to 
 consider the matter closed, unless equally careful tests 
 have also convinced him that his immediate conviction 
 is not to be shaken by arguments. The impotency of 
 argument on matters of religious dogma is always to be 
 regretted. For it is notorious that people do differ on 
 these subjects, and, where argument is impotent, no- 
 thing can be done to promote an agreement. Where 
 nothing can be done, the evil must, of course, be 
 recognized. But we are bound in each case to make 
 ourselves quite sure that the evil cannot be removed. 
 
 37. It is certain, no doubt, that, if we are to have 
 any knowledge at all, we must have some immediate 
 convictions, and that if we are to have any true know- 
 ledge at all, some of our immediate convictions must 
 be true. For nothing can be proved unless we start 
 from something already known, and, if we could know 
 nothing unless it were proved, we could never start 
 at all. 
 
 But it would not follow from this that every man 
 must have immediate convictions on matters of religious 
 dogma, or that he must have true immediate con- 
 victions on that subject. For it is possible to have 
 knowledge without having immediate convictions as to 
 religious dogmas, although it is not possible to have 
 knowledge without immediate convictions of some 
 sort. 
 
 Again, there is nothing to be regretted in the 
 immediacy of our convictions as to the validity of a 
 
46 SOME DOGMAS OF RELIGION 
 
 syllogism in Barbara, or as to the Law of Contradic- 
 tion. We cannot doubt them, and we cannot prove 
 them. But the inability to prove them is not a disaster 
 here, because nobody denies them. Where we cannot 
 argue, we can do nothing to remove differences of 
 opinion. But there is no harm in this, where there 
 are no differences of opinion to remove. With religious 
 dogma it is different. For every religious dogma of 
 the truth of which any man has an immediate con- 
 viction, is believed by many men to be false. 
 
 38. We must now consider the second branch of our 
 subject — the arguments which urge that a religious 
 dogma is to be behoved by me, because it is held by 
 certain other people. 
 
 It has been attempted from time to time to rest the 
 truth of dogmas on the fact that every one believed 
 them. But it is clear, that, if an argument is wanted 
 at all, this argument cannot be true. K no one doubts 
 a proposition, it is superfluous to spend time in proving 
 it. If any one doubts it, then it cannot be truthfully 
 supported by the assertion that every one believes it. 
 Moreover, the assertion is notoriously false. It is 
 impossible to produce any religious dogma which no 
 one has doubted, or even any religious dogma which 
 no one has definitely disbelieved. The view that all 
 persons who profess to disbelieve in certain dogmas 
 are only lying for the sake of notoriety has fortunately 
 only historical interest, since it finds no place in the 
 controversies of the present time. 
 
 39. Again, we are invited to beUeve a dogma be- 
 cause the belief in it, though not universal, is very 
 general. We are sometimes told, indeed, that the 
 belief is universal, except in the case of those people 
 whose minds have become so deeply sophisticated, 
 
THE ESTABLISHMENT OF DOGMA 47 
 
 either by their own action or the influence of others, 
 that they are unable any longer to see the truth which 
 is normally perceptible to every one. 
 
 This argument, however, involves a vicious circle. 
 The fact that a position has been reached by a gradual 
 abandonment of primitive ideas and beliefs, effected 
 by a process of self-culture and education, does not 
 necessarily condemn it. Indeed, all advances in science 
 and moraHty are made in this way. The apparent 
 force of the argument lies in the word 'sophistication'. 
 And here the circle comes in. For how are we to 
 distinguish sophistication from healthy development, 
 except from the one leading to the false and the other 
 to the true? And, as the question to be proved is 
 just whether the conclusions to which the process leads 
 are true or false, to call the process sophistication begs 
 the question while professing to prove it. 
 
 It might perhaps be possible to avoid the circle by 
 determining a development as sophistical on account 
 of its moral characteristics, and inviting us, when this 
 has been done, to distrust the intellectual result pro- 
 duced by it. But it would not, I think, be now 
 seriously maintained that any rehgious dogma was 
 certain to all men, except to those who had disquaUfied 
 themselves by a life of exceptional wickedness. 
 
 40. Or, again, the appeal to the general opinion 
 may not criticize the intrinsic trustworthiness of the 
 opinions of the minority, but may confine itself to 
 pointing out that they are the minority, and that the 
 opinion of the majority, therefore, is to be preferred. 
 
 This ground of behef does not seem very strong. 
 For it has often been acknowledged, by general con- 
 sent, that the majority has been wrong, for long 
 periods, and on important questions, and that the 
 minority has been right. If we take the opinion 
 
48 SOME DOGMAS OF KELIGION 
 
 of the majority merely because it is their opinion, 
 we adopt a principle which would have led us 
 wrong in very many cases in the past. Why should 
 we trust it for the present or the future ? If, on the 
 other hand, we endeavour to recommend this parti- 
 cular opinion of the majority as being from its nature 
 more likely to be right than the opinion of the 
 minority, we have passed away from the argument 
 from majorities, since any intrinsic reasons for the 
 truth of the dogma would be of the same weight if 
 they had as yet only convinced a single person. 
 
 Moreover, even if the argument from the opinion of 
 the majority were legitimate, is there any case where 
 we could safely apply it? If it were to have any 
 weight at all, it would surely be only in cases where 
 there is a decisive majority. But on what religious 
 dogmas is there such a majority? If we take into 
 account the whole world (and I do not see that we 
 have any right to exclude any part of it when we are 
 counting heads) it would probably be difficult to tell 
 on which side the majority was as to such questions as 
 the existence of God or human immortaHty. And if 
 we try (a task almost, if not quite, impossible) to weigh 
 opinions as well as to count them, we find that on 
 these subjects the opinions of wise men show no 
 more agreement than those of men in general. An 
 appearance of a decisive majority for some particular 
 dogma may be gained, no doubt, by first excluding 
 certain nations on the ground of their asserted in- 
 feriority, and then insisting that the opinion of every 
 man within the circle of more favoured nations is 
 to be counted as of equal value. But such a course 
 seems to be indefensibly inconsistent. 
 
 41. There remains the argument that certain dogmas 
 should be accepted because they have been held by 
 
THE ESTABLISHMENT OF DOGMA 49 
 
 men, or beings incarnate in human bodies, who have 
 worked miracles, including the miracle of predicting 
 the future. 
 
 A miracle is an event which we camiot explain by 
 any natural law known to us, and which is therefore 
 attributed, by the behevers in its miraculous character, 
 either to a special divine interference with the course 
 of nature, or to the action of some law, differing in its 
 nature from those which explain non-miraculous 
 events. It is then argued that the occurrence of such 
 events at the wiU of, or in connexion with, a particular 
 being, is evidence, either that that being is himself 
 divine, or that he enjoys special divine favour, and, in 
 either case, that his teaching on matters of reUgious 
 dogma is trustworthy. 
 
 The evidence for the existence of miracles is an 
 inquiry beyond our purpose. But we may remark in 
 passing that, as Hume has pointed out i, if miracles 
 are to be accepted as evidence of the truth of a reli- 
 gion, then whatever evidence there is for the miracles 
 of one religion is evidence against the truth of all in- 
 compatible religions. There is perhaps no reason, if 
 there are miracles at all, why they should not occur in 
 connexion with several incompatible systems. There 
 might be reasons why a God should work miracles 
 in connexion with a false rehgion. Or again the 
 miracles of all the systems except one's own might be 
 ascribed, as they used to be ascribed, to devils. But 
 then miracles would prove nothing about the truth of 
 a religion. If, on the other hand, they can prove any- 
 thing about it, then none but a true religion can have 
 miracles connected with it. Of two religions with 
 incompatible dogmas, one, at least, must be false, and 
 
 ^ Enquiry concerning Human Understanding^ Sect. 10 (p. 99, 
 ed. Green and Grose). 
 
50 SOME DOGMAS OF RELIGION 
 
 therefore only one, at most, can have miracles con- 
 nected with it. Thus neither religion can be proved 
 true, without disproving the existence of the miracles 
 of the other religion. And in so far as these latter are 
 at all probable, they render the truth of the first 
 religion improbable. 
 
 42. Supposing that miracles were proved to exist, 
 and to exist in connexion with one religion only, 
 should we be entitled to believe that rehgion to be 
 true ? It seems to me, to begin with, that the exis- 
 tence of the miracle would not prove that it was due 
 to the action of God — meaning by God a supreme 
 being. The amount of power required for any miracle, 
 however startling, can never be proved to be more 
 than finite. And in that case it is always possible 
 that it should have been performed by some being 
 whose power, while much greater than human power, 
 might be far below the power of a supreme being. 
 
 If then a miracle were due to the action of such 
 a super-human but non-divine being, would it give any 
 reason to suppose the religion to be true ? I see no 
 reason to believe that a being who can raise the dead, 
 or prophesy the future, or assist a man to do these 
 things, would be a specially trustworthy guide on 
 matters of religious dogma. The power of influencing 
 the course of events, and the power of apprehending 
 rehgious truth, are not always closely connected. 
 Napoleon greatly excelled the average English clergy- 
 man in the first, but it would be a rash inference that 
 he excelled him in the second. 
 
 But even if it were certain that a non-divine 
 miracle-worker was our superior generally, as well as in 
 the power of working miracles, it would not follow that 
 we should go right if we trusted him on religious 
 dogmas. Men would often go wrong by following 
 
THE ESTABLISHMENT OP DOGMA 51 
 
 the teaching of their spiritual superiors. Buddha and 
 St. Paul were both the spiritual superiors of the aver- 
 age monk, Buddhist or Eomanist. But it is clear that on 
 certain points either a monk who followed Buddha or 
 a monk who followed St. Paul would have gone wrong, 
 since either Buddha or St. Paul must be wrong where 
 they differ. 
 
 Waiving this difficulty, and assuming that the 
 miracle could prove the special interference of the 
 supreme being, so that the rehgion connected with it 
 could be accepted as his revelation, should we then be 
 safe in accepting it as true ? We should not be justi- 
 fied, I submit, unless we had previously proved that 
 the supreme being was good. For we have no reason 
 to suppose that he will tell us the truth except that it 
 would not be a good act to deceive us. If he is indif- 
 ferent to the good, or if he is positively malignant, he 
 may well tell us lies, either from caprice or in order to 
 gratify his malignancy. 
 
 It is obviously impossible to trust to the revelation 
 to tell us that he is good, since we have no reason to 
 trust the revelation at all unless we know that he is 
 good. This goodness must be proved independently. 
 And thus one of the most important of dogmas cannot 
 be proved by a miracle-based revelation. 
 
 43. If, however, this dogma has been independently 
 proved, are we then entitled to accept the divine 
 revelation as true ? Even then I do not think that we 
 can do this. A God— that is, a good supreme being — 
 will doubtless regard deceit as an evil. But there is, 
 beyond doubt, much evil in the universe, and, if we 
 are satisfied that there is a God, we must regard that 
 evil as in some way compatible with his goodness. 
 And then why not that further evil of a misleading 
 divine revelation ? If, for example, we attribute the 
 
 E 2 
 
52 SOME DOGMAS OF EELIGION 
 
 existence of evil to God's limited power, and say that 
 cancer and plague exist because they are the best 
 that God can do for us under the circumstances, how 
 can we be sure that the best thing he can do for us 
 under the circumstances is not to deceive us about 
 religious dogma ? How can we be sure, for example, 
 if God tells us we are immortal, that it is not a deceit — 
 bad in itself, but good as the means of avoiding some 
 greater evil ? If it were the best thing he could do 
 for us, there would be nothing extraordinary in his 
 working miracles to do it. 
 
 44. The third branch of our subject is the attempt 
 to prove dogmas by what may be called a reductio ad 
 horrendum. Unless a certain dogma is true, it may be 
 said, the universe would be intolerably bad — either 
 intolerably miserable or intolerably wicked — and 
 therefore the doctrine must be true. 
 
 It may be remarked, to begin with, that it is not so 
 easy to estabhsh the premise of this argument as is 
 sometimes supposed. When we are told that the 
 universe would be intolerably bad unless A were 
 true, we must remember that all dogmas involve com- 
 plex ideas, and that it may be part only of A which 
 is essential to avoid the evil. If it is asserted, for 
 example, that the universe would be intolerable if 
 human actions were completely determined, analysis 
 may reveal that what is meant is only that it would 
 be intolerable if they were completely determined 
 from outside. 
 
 Again, tastes differ, and tastes change. A Viking 
 or a Maori warrior might well find that the prospect of 
 an immortality without fighting made the universe 
 intolerable. And he might well be supported by the 
 public opinion of his own society. But the hfe of the 
 
THE ESTABLISHMENT OF DOGMA 53 
 
 modern Christian — and many of these are Scandi- 
 navians, and a few are Maoris— is not rendered 
 miserable by his conviction of the peacefulness of 
 heaven. Examples like these should make us careful 
 before we assert that the universe would be proved 
 worthless if some dogma were proved false. 
 
 But there are cases, no doubt, where the premise 
 would be true. We can easily imagine dogmas — or 
 even recall them in history — which it would be 
 scarcely possible to reconcile with any judgement, 
 from any standpoint, except that the universe was very 
 bad. Of the dogmas logically contradictory to these, 
 therefore, we may assert, with scarcely a trace of doubt, 
 that unless they are true the universe is very evil. 
 
 45. But what of that ? What reason have we, at 
 this point, to assert that the universe is not very evil? 
 I say ' at this point ', because, when we have investi- 
 gated the fundamental nature of reality, we may find 
 reasons to believe that, in point of fact, the universe 
 is not very evil. But till we have investigated the 
 more important dogmas of religion we have not in- 
 vestigated the fundamental nature of reality, especially 
 as it relates to good and evil. It is only by the deter- 
 mination of such questions as the existence of God, the 
 immortahty of man, the purposiveness of things, and 
 the like, that we can determine whether the universe 
 is of a nature which, in point of fact, is not very 
 evil. To use the principle that it cannot be very 
 evil, in order to determine these very problems, would 
 be to beg the question, unless we are in a position to 
 lay down, before we know what the nature of the 
 universe is, that at any rate it cannot be very evil. 
 
 This is what the argument before us does. It takes 
 the statement that the universe cannot be very evil as 
 a truth evident in itself — evident before we know 
 
64 SOME DOGMAS OF EELIGION 
 
 what the nature of the world in other respects is — and 
 makes it the basis of further investigation. With 
 what right ? If it were very evil, the fact would be 
 very much to be regretted. If its evil were due to 
 the uncontrolled action of any conscious being, that 
 being would be very much to be blamed. But have 
 we any reason to believe that it is not very evil ? 
 Eeality and goodness are, at any rate prima facie^ 
 separable. Whenever I admit that I have faults, 
 I admit that something is real which is more or less bad. 
 What argument can be brought forward — none, I think, 
 ever has been brought forward — to show that reality 
 and extreme badness are not ultimately compatible ? 
 
 46. It may be said that, although this line of argu- 
 ment is not permissible in establishing whatever 
 dogmas are taken as fundamental, it may be used in 
 certain cases, when the more fundamental dogmas have 
 been estabhshed, to prove further dogmatic proposi- 
 tions. If, for example, it were first proved in some 
 other manner that the universe is entirely dependent 
 on the will of a benevolent God, might we not there- 
 fore conclude that the universe is not very bad, and 
 that therefore, for example, we must be immortal, 
 because, if we were not inmiortal, the universe would 
 be very bad ? 
 
 I do not, however, think that this is valid. For it 
 must be remembered that any argument which would 
 prove that there was no evil in the universe would be 
 at once condemned by the incompatibility of its 
 conclusion with notorious facts. There is evil in the 
 universe. The shghtest pang of toothache, the 
 sUghtest envious thought, proves that beyond all 
 doubt. It is, therefore, clear that the principle which 
 determines the universe does not exclude all evil from 
 it. And, since this is so, how can we venture to put 
 
THE ESTABLISHMENT OF DOGMA 55 
 
 any limits to the amount of evil which it will admit 
 into the universe ? Desirable as it would be to do 
 this, I do not see how it could possibly be done. I do 
 not mean that it is impossible to determine the nature 
 of the universe to such an extent as to be able to 
 assert that, as a matter of fact, in spite of the evil in 
 it, it is more good than evil. But I maintain that this 
 can only be done by first finding out what the nature 
 of the universe is, and by discovering in this manner 
 that it is, in fact, more good than evil. We cannot 
 argue, before we know the characteristics of the uni- 
 verse, that it cannot have a particular characteristic, 
 because it would be very evil, while the universe, or 
 its creator, is perfectly good. For if such an argu- 
 ment could prove anything, it would prove that no 
 evil exists at all. And an argument which proves 
 that is reduced to an absurdity. 
 
 The objections I have put forward above depend 
 upon the assertion that it is beyond all doubt that 
 there is real evil in the universe. They would be 
 removed, therefore, if we accepted the view, held by 
 some metaphysicians, that it is impossible that any 
 part of reality should be evil, and that there is no real 
 evil in the universe at all \ But, should this view be 
 accepted, while it would destroy my objections, it 
 would destroy no less the arguments to w^hich they 
 were objections. The considerations which induced 
 those metaphysicians to assert that there was no real 
 evil in what appears to us as toothache and envy, 
 would also — since they are always of a quite general 
 nature — induce them to assert that there was no real 
 evil in the consequences which would follow from the 
 falsity of the dogma to be proved. And thus the 
 whole argument for the truth of the dogma would 
 
 ' Cp. Section 171. 
 
56 SOME DOGMAS OF EELIGION 
 
 vanish, since it rested exclusively on the evil results 
 that would follow from the falsity of the dogma. 
 
 47. But, it may be said, why should we not argue 
 as much from our desires and aspirations as from any- 
 thing else ? Are they not as real as anj^hing else in 
 the world ? And are not those desires and aspirations, 
 which are directed to the establishment of perfection, 
 either in ourselves or in the world at large, as good as 
 anything in the world ? Are we not as much entitled 
 to argue from their existence and their goodness as 
 from any other fact? 
 
 This is a consideration which is often urged. So 
 far as the frequent exhortation to trust the heart 
 rather than the head means anything, it appears to 
 mean this. But the position rests on a misunder- 
 standing of a nature so elementary that it could 
 scarcely have escaped detection if the writers who 
 used it had not been bhnded by the supposed attrac- 
 tiveness of the conclusion. For the question is not 
 whether we can deduce anything from the existence 
 and goodness of our desires and aspirations. No 
 person in his senses ever denied that we could. The 
 question is what we can deduce. And, especially, 
 can we deduce from their existence or goodness the 
 conclusion that they will be eventually gratified ? 
 
 Our desires and aspirations are real. From this, of 
 course, we can argue that whatever is inconsistent 
 with their reality must be false. But in what way is 
 the failure of a desire to be gratified inconsistent with 
 its reality ? I can see no contradiction whatever in 
 the statement that a desire is real, but remains un- 
 gratified. The statement is often true. Many people 
 had a real desire that the Pretender should be victorious 
 in 1746, but they were disappointed. 
 
 Our desires, indeed, on such subjects as the existence 
 
THE ESTABLISHMENT OF DOGMA 57 
 
 of God, or of immortality, are of very different im- 
 portance from the desire for a political revolution. 
 But they are, as desires, no more real. The reality of 
 a desire relates to its existence as a psychical fact, 
 and from this point of view all desires are on a level. 
 If the reality of any desire is compatible with its non- 
 fulfilment, then we can never argue from the reality 
 of any desire to its fulfilment. If we wish to make 
 a distinction between the desire for heaven and the 
 desire for the restoration of the Stuarts, we can only 
 do so on account of the greater importance of the 
 object of the former. And if, on account of this greater 
 importance, we argue that the desire for heaven cannot 
 remain unfulfilled, although the other has remained 
 unfulfilled, we have abandoned the argument from the 
 reality of the desire, and gone back to the argument — 
 already considered — from the amount of evil which 
 the falsity of the dogma would imply. 
 
 It should also be noticed that, if the reality of 
 a desire ensured its eventual gratification, then, even 
 if we hmited ourselves to those desires which bear 
 on rehgious dogmas, we should be involved in contra- 
 dictions. Some men desire that an omnipotent 
 God should exist. Others feel that, in a universe 
 which contains evil, the existence of an onmipotent 
 God would be the culminating evil. Some men 
 passionately desire immortality, others the loss of 
 their personal identity. Each desire is as real as the 
 desire for the contrary result. In this, of course, 
 there is no contradiction. There is no difficulty in 
 believing that A desires that there should be a God, 
 and B that there should not be a God. But if a desire 
 involves its fulfilment there would be a hopeless con- 
 tradiction. For it is impossible that God should both 
 be and not be. 
 
58 SOME DOGMAS OF EELIGION 
 
 This last difficulty is generally concealed from view- 
 by a distortion of the evidence, which is certainly in 
 most cases unconscious, but is none the less illegiti- 
 mate. The man who argues from the reality of 
 desires to the truth of dogmas often, in the first place, 
 ignores the races to which he does not belong. In the 
 next place, he leaves out of account those people of his 
 own race whose desires are different from his own. 
 This latter elimination is the more practicable since 
 such arguments as these are generally put forward in 
 favour of the orthodox ideas of a particular time and 
 place, and therefore there is only a minority to ignore. 
 And in this way he comes to assert that ' the human 
 heart imperatively demands ' something or the other, 
 when all that he is entitled to say is that he desires it 
 himself, and that the majority of his race agree with him. 
 
 48. Similar considerations apply to the argument 
 that because our desires and aspirations are of high 
 moral value we are entitled to believe that they will 
 be fulfilled. There are two different positions which 
 it is possible to confuse here. One is that the objects 
 of the desires would, if attained, have high moral 
 value. The other is that the desires themselves have 
 high moral value. The argument based on the first 
 of these we have already considered. It is a variety 
 of the argument that certain things must exist, because 
 their non-existence would be so bad, which was dis- 
 cussed above. 
 
 But the other remains to be considered. And the 
 fact on which it is based is no doubt often true. The 
 desires and aspirations of men for the truth of various 
 dogmas may have very high moral values. And it 
 must be noticed that they may have that value in 
 cases where the objects of the desires are devoid of it. 
 One man thinks that it is right that a sinner who dies 
 
THE ESTABLISHMENT OF DOGMA 59 
 
 impenitent should be tormented unendingly. Another 
 thinks that it is not right that any one, under any 
 conditions, should be tormented unendingly. One of 
 them must be wrong as to what would be, under 
 certain conditions, a good state of things. But the 
 desire of each that what he thinks good may be 
 realized may have high moral value. 
 
 Again, if the highest good for man be an eternity 
 of conscious worship, then it must be a mistake to hold 
 that the highest good for man is to be merged into an 
 impersonal Absolute. But the desire of one man, 
 who holds the first to be best, to gain heaven, and the 
 desire of another man, who holds the second to be 
 best, to gain personal annihilation, may both have 
 high moral value, since each is a desire for what is 
 conceived to be the highest good. 
 
 But, admitting the goodness of these desires, what 
 right have we to conclude that they will be fulfilled ? 
 Is there any incompatibility between the goodness of 
 a desire and its failure ? My desire to be learned 
 may be good, though obstacles may keep me ignorant. 
 My desire to save a drowning man may be good, 
 though he may sink before I can reach him. That 
 a good desire is unfulfilled is an argument of more or 
 less weight against the goodness of the universe in 
 general, but not against the goodness of the desire. 
 Why should desires for the truth of dogmas be 
 different from the rest ? And if every good desire 
 was necessarily to be fulfilled, an absurdity would 
 arise, since desires for quite incompatible things may, 
 as we have seen, both be morally valuable. 
 
 At any rate, it may be said, any dogma which would 
 paralyse our activity cannot be true. What is meant 
 is not that the truth of the dogma would prevent any 
 one from acting, but that the belief in the truth of the 
 
60 SOME DOGMAS OF EELIGION 
 
 dogma would prevent the believer from acting. Now 
 this is no reason for doubting the truth of the belief. 
 It may be a great calamity that people should find 
 their power of action paralysed. But great calamities 
 sometimes happen. And besides, what would be evil 
 in this case would not be the truth of the dogma, but 
 the belief in its truth. So that, even if such an argu- 
 ment was legitimate at all, it could only tell us that 
 people would not beheve the dogma. It could not 
 tell us that the dogma was false. 
 
 49. And is there any dogma the belief in whose 
 truth would paralyse action? Is there, to begin with, 
 any dogma the belief in which would make action 
 absurd ? It is sometimes said that a behef in mate- 
 rialism would make action absurd. If materiahsm is 
 true, then each man ends at the death of his body, and 
 the human race will almost certainly be frozen out of 
 existence some day. No good result that we can 
 achieve will have any chance of permanence. And 
 this, it is said, will render all action — or at any rate 
 all moral action — absurd. 
 
 But why should it do so ? It would be absurd to 
 act, no doubt, if action made no difference in any 
 result which was of value. But neither materialism 
 nor any other dogma, which has ever been maintained, 
 could lead to this conclusion. We may not survive 
 the death of our bodies, and the race may be destined 
 to endure for only a few brief millions of years. But, 
 meanwhile, the race has not yet ceased to exist, and 
 here we are, particular individuals. And, while we 
 are here, whatever the future may be, it is better that 
 we should be full than hungry, better that we should 
 work than steal, better that we should read Eobert 
 Browning than that we should read Eobert Mont- 
 gomery. It is worth while to bring about these things 
 
THE ESTABLISHMENT OF DOGMA 61 
 
 for ourselves and others, and since that is so, action 
 is not absurd. Even if every man were destined to an 
 unending hell at the end of his present life, it would 
 still be better for that present Hfe to be of one sort 
 rather than another sort. 
 
 It may be said, however, that, although there are 
 no dogmas belief in which would make action logically 
 absurd, yet there might be some dogmas belief in 
 which would make action psychologically impossible. 
 Some dogmas might represent the world as inevitably 
 so bad that we should cease, in our despondency, to 
 do what was in our power to make it better, or even 
 to care for those improvements in our own condition 
 for which only selfish motives would be required. It 
 is possible that belief in some dogmas might have this 
 result, though, as I pointed out above, that could not 
 have any bearing on the question of their truth. If 
 a man did firmly believe that he himself, and every 
 one he knew, were going to hell when they died, he 
 might be paralysed with despair. Even this is not 
 certain. But what does seem quite certain is that 
 men are not paralysed in action merely because they 
 do not believe in the eventual and permanent triumph 
 of the good. If they were, it would imply that they 
 thought more about the far distant future, and less 
 about the present and near future than is reasonable. 
 For we have seen that such paralysis, even if it did 
 occur, would not be reasonable. Now it is surely 
 notorious that the average man is so far from thinking 
 more than is reasonable about the distant future that 
 he habitually thinks less than is reasonable about it. 
 All study of economic facts proclaims this. And in 
 religious matters it has been the constant theme of 
 preachers since preaching began. Many men who 
 firmly believe that death will be for them the begin- 
 
62 SOME DOGMAS OF RELIGION 
 
 ning of intense and unending happiness are profoundly 
 depressed at the thought that death is inevitable. 
 Many men have committed sins in the firm belief 
 that, should they die while committing them, they 
 would go to hell. A member of a race which discounts 
 such expectations at such a rate will scarcely be 
 deterred from action by the conviction that for him- 
 self and his friends it will be aU the same in a hundred 
 years, and for the world in general it will be all the 
 same in some thousands of centuries. 
 
 50. There is one more form which the argument 
 from consequences takes. Practice, we are told, is 
 supreme over theory. All thought is action, and all 
 action is directed towards an end. And, therefore, 
 practical considerations are the supreme test of truth. 
 And, therefore, if we want a particular conclusion to be 
 true or false, or at least if we want it very intensely, 
 we are entitled, and indeed bound, to assume it to be 
 true or false accordingly. 
 
 To discuss the truth of the view that practice is 
 supreme over theory would take us too far away from 
 our subject. I will only remark that, so far as I know, 
 all attempts to prove this have consisted of demon- 
 strations that all theory depends upon practice. 
 Admitting this to be true, from one point of view, it 
 seems to me no less true that from another point of 
 view all practice depends upon theory, and that the 
 true conclusion is that the two are reciprocally 
 necessary to each other, and that priority or supre- 
 macy can be ascribed to neither. 
 
 But it is not necessary to determine this point, for, 
 however it is determined, the argument fails. Let 
 us take a case of theory which depends on practice, 
 and let us ignore the question whether that practice 
 does not in its turn depend upon theory. If a man 
 
THE ESTABLISHMENT OF DOGMA 63 
 
 studies medicine, with a view to acting as a physician, 
 his knowledge is conditioned by practice. A know- 
 ledge of medicine is not acquired without strenuous 
 exertion. And the exertion has an end — to acquire 
 money, or to acquire fame, or to cure his patients. 
 Let us take the last. Then his possession of know- 
 ledge is dependent on his will to cure the patients who 
 consult him. But for his resolve to be able to do this 
 he would not have acquired the knowledge at all. But 
 if he were to allow himself to make practice the 
 criterion of his knowledge, and, because he desired to 
 cure every patient, to believe that he would succeed in 
 doing so, his belief would be generally admitted to be 
 indefensible. 
 
 Now I cannot see any difference between this con- 
 duct and that to which the advocates of the supremacy 
 of practice invite us. No one would consider dogma 
 unless he had some end in view. That end may be to 
 know the truth for its own sake. But it may be, and 
 no doubt very often is, the hope of discovering that 
 the universe is good. But it is one thing to investigate 
 the nature of the universe in the hope that we may 
 find that it is good, and quite another to declare that 
 we shall find this result, because we hope to find it. 
 
 51. We have now, I think, examined the principal 
 forms in which the argument from consequences pre- 
 sents itself. It only remains to apply to it the test 
 which it is so anxious should be applied to every- 
 thing. What would be the consequences of the 
 validity of this argument? 
 
 The question has sufficient practical importance to 
 justify the inquiry. The good or evil consequences 
 which would result from the truth of a proposition, 
 which so many people hold to be true, can never lack 
 interest. But it also has some bearing on the truth of 
 
64 SOME DOGMAS OF KELIGION 
 
 the argument. The argument is that whatever would 
 be very bad cannot be true. If it should turn out that 
 the argument's own truth would have destructive effects 
 on morality, then its supporters may, of course, deny 
 that it is very bad to exercise destructive effects on 
 morality. But this they are scarcely likely to do. Or 
 else they must admit that, if the argument is true, it 
 proves that it cannot be true, since it belongs itself to 
 the class of very bad things of which it denies the 
 truth. And this would be fatal to the truth of the 
 argument. 
 
 *What would be very bad cannot be true.' This 
 seems to many people — and naturally enough at first 
 sight — rather an inspiring and elevating assertion. 
 What is more flattering to morality, which deals with 
 the good, than to make goodness the test of reality ? 
 But the consequences of the proposition are far reach- 
 ing. If that which is very bad is not true, it follows — 
 it is valid even by formal logic — that what is true 
 cannot be very bad. 
 
 Extreme badness excludes truth, we are told. So 
 long as anything seems certainly bad, and is not 
 certainly true, this appears to be advantageous for 
 goodness. But supposing some day the truth should 
 be revealed without possibility of doubt — and sup- 
 posing it should turn out to be what we had previously 
 thought very bad ? Then we shall be bound to believe 
 that it is not very bad — not because our opinions on 
 the badness have changed, but because the existence of 
 the thing proves that we were mistaken. 
 
 It sounds favourable to morality to assert that Ter- 
 tuUian's hell, and the pleasure which the inhabitants 
 of heaven take in witnessing it, are impossible because 
 they are hideous. And at present, whatever the effect . 
 on morality, it is a comforting belief. But supposing 
 
THE ESTABLISHMENT OF DOGMA 65 
 
 we should wake some day to find ourselves in that hell, 
 or in a heaven where the blessed were enjoying the 
 sight of it. Then, if the argument from consequences 
 is valid, we should have to believe that neither the 
 existence of such a hell nor of such a heaven was 
 morally revolting. And such a reversal of our moral 
 judgements could scarcely, I think, be represented as 
 favourable to morality. 
 
 52. Even without any such demonstration of such 
 a truth, what would be the effect on morality? It 
 would still be true, if the argument from consequences 
 were correct, that if anything were real, it could not be 
 very bad. That is to say, our moral judgements have 
 no independence. There are things which cannot be 
 very bad, because they possess the non-moral quality 
 of being real. 
 
 Surely this destroys the very essence of morahty. 
 Morahty passes judgement on whatever is real. It 
 pronounces it to be good or bad. And it seems to me 
 that it is the very essence of morality to hold that in 
 this judgement it is independent — that it is able to say 
 of reality that it is bad because it has the quality jt, 
 and because rr, however fundamental to reality, is bad, 
 or again to say of reality that it is bad because it has 
 not the quality y^ and because y^ however ahen to 
 reality, is essential to goodness. 
 
 It is absurd to ask whether reality or goodness be 
 the more fundamental. Each is supreme in its own 
 sphere, and the spheres are so different that they can- 
 not come into conflict. What is real is real, however 
 bad it is. What is not real is not real, however good 
 it would have been. On the other hand, what is good 
 is good, however unreal it is. What is bad is bad, 
 however real it is. And so it is our duty to be 
 humble in judging of reality, and imperious in judging 
 
 F 
 
66 SOME DOGMAS OF RELIGION 
 
 of goodness. For what is real is real^ however we may 
 condemn it. But, on the other hand, what we con- 
 demn — if we condemn rightly — is bad, even if it were 
 the essence of all reality. 
 
 The moral evil of the argument from consequences 
 seems to me to be that it makes us imperious in the 
 wrong place, where our imperiousness is arrogance, 
 and, by an inevitable consequence, makes us humble 
 in the wrong place, where our humility is mean and 
 servile. When the reality of a thing is uncertain, the 
 argument encourages us to suppose that our approval 
 of a thing can determine its reality. And when this 
 unhallowed link has once been established, retribution 
 overtakes us. For when the reality is independently 
 certain, we have to admit that the reality of a thing 
 should determine our approval of that thing. I find 
 it difficult to imagine a more degraded position. 
 
 53. It remains to speak of the appeal, sometimes 
 made, to leave dogmatic questions to faith. In so far 
 as this means faith in a proposition, it presents nothing 
 new. If the faith is asserted to be based on reason, 
 then the question will arise whether the reasons are 
 valid, and it becomes a question of argument. If the 
 faith dispenses with reason, we have again the appeal 
 to immediate conviction. 
 
 We are sometimes invited to have faith on account 
 of the very limited amount of our knowledge, and the 
 possible errors in it. The argument is not, I think, 
 one which gains any importance by being advocated 
 by thinkers of repute, but it is sufficiently popular to 
 deserve some consideration. 
 
 The fact on which it is based is unquestionably true. 
 We know very little, compared to what there is to be 
 known ; and what we take for knowledge is frequently 
 
THE ESTABLISHMENT OF DOGMA 67 
 
 error. If this were assigned as a reason for not being 
 certain as to the truth of our conclusions — and especially 
 of our conclusions on obscure and disputed subjects — 
 the argument would be unquestionably legitimate. 
 And if it were given as a reason for complete scepticism, 
 the conclusion, though exaggerated, would not be very 
 surprising. But it is somewhat remarkable that our 
 want of knowledge on any subject, should be put for- 
 ward as a reason for coming to a particular conclusion 
 on that subject. 
 
 Yet this is often done. If it is suggested that there 
 is no evidence that the universe is working towards 
 a good end, the doubter is reminded of the hmitations 
 of his intellect, and on account of this is exhorted to 
 banish his doubts from his mind, and to believe firmly 
 that the universe is directed towards a good end. And 
 stronger instances can be found. An apologist may 
 admit, for example, that for our intellects the three 
 facts of the omnipotence of a personal God, the bene- 
 volence of a personal God, and the existence of 
 evil, are not to be reconciled. But we are once more 
 reminded of the feebleness of our intellects. And we 
 are invited to assert, not only that our conclusions 
 may be wrong, not only that the three elements may 
 possibly be reconciled, but that they are reconciled. 
 There is evil, and there is an omnipotent and bene- 
 volent God. 
 
 54. This hne of argument has two weaknesses. The 
 first is that it will prove everything — including mutu- 
 ally incompatible propositions — equally well. It will 
 prove as easily that the universe is tending towards 
 a bad end as that it is tending towards a good one. 
 There may be as little evidence for the pessimistic view 
 as for the optimistic. But if our intellects are so 
 feeble that the absence of sufficient evidence in our 
 
 F 2 
 
68 SOME DOGMAS OF EELIGION 
 
 minds is no objection to a conclusion in the one case, 
 then a similar absence can be no objection to a conclu- 
 sion in the other. Nor can we fall back on the asser- 
 tion that there is ^55 evidence for the pessimistic view 
 than for the optimistic, and that, therefore, we should 
 adopt the latter. For if our intellects are too feeble 
 for their conclusions to be trusted, our distrust must 
 apply equally to their conclusion on the relative weight 
 of the evidence in the two cases. 
 
 The other objection to the argument is that it implies 
 that, if we cannot trust our conclusion that A is false, 
 we have no alternative but to conclude that A is true. 
 But there is a third alternative to being confident of 
 the truth or confident of the falsity of A, It is to 
 abstain from judging about A at all. And it is this 
 which would seem to be the more reasonable alterna- 
 tive, supposing our intellects are as weak as they are 
 asserted to be. If I have only taken a hasty view by 
 twilight of my neighbour's garden, it would be rash of 
 me to place much trust in my failure to see any hlies 
 in it. But it would be even more rash if I proceeded 
 from the untrustworthiness of my negative conclu- 
 sion to a confident assertion that there were lilies in 
 it, and that there were exactly seventeen of them. 
 
 Even in this case, however, I should not have to 
 ignore an a priori conviction of the impossibility of 
 lihes. But the argument from the feebleness of our 
 intellects is often used as a reason why we should 
 believe a state of things to exist which our intellects 
 pronounce to be self-contradictory. I might as well 
 call on you and myself to believe that in some remote 
 comer of the universe the Law of Diminishing Returns 
 devours purple quadratic equations. It seems to us, 
 certainly, that a quadratic equation cannot have colour, 
 or be eaten. But then, how inadequate are our 
 
THE ESTABLISHMENT OF DOGMA 69 
 
 merely human powers to limit the resources of the 
 infinite ! 
 
 55. But there is another form of the appeal to faith 
 which requires m.ore serious consideration. We are 
 invited to have faith, not in the truth of a proposition, 
 but in the goodness of a person. A man trusts his 
 friends, or his political leader. He behoves them to 
 have good motives for actions whose motives he is 
 unable to detect. He often believes them to have 
 good motives in cases where, but for his personal faith 
 in them, the circumstances would strongly suggest 
 that they were actuated by bad motives. Subsequent 
 events often — though not always — prove that he was 
 right. Can we not, we are asked, trust God as we 
 trust our friends ? 
 
 In the first place it must be remarked that this form 
 of faith can only have a very hmited application in the 
 establishment of dogma. It assumes, to begin with, 
 that the dogma of the existence of a personal God has 
 been already established. For the appeal to trust 
 God as we trust men loses all plausibility if God is 
 not a person. If the ultimate reality of the universe 
 were an aggregate of atoms, or a chaos of sensations, or 
 a substance devoid of will, intellect, and purpose, it 
 would be futile to trust it. 
 
 Then our faith in a man does not enable us to 
 predict his actions — which, indeed, are often deter- 
 mined, as in the case of a statesman, by considerations 
 which are inaccessible to us. All that it does is to 
 render us confident that they will be wise and virtuous. 
 And a corresponding faith in God will not enable us 
 to determine whether we are immortal, whether our 
 wiUs are undetermined, or similar questions of dogma. 
 It will only give us light on one particular dogma — 
 that the world is wisely and righteously ordered. 
 
70 SOME DOGMAS OF RELIGION 
 
 And even here it will not by itself enable us to say 
 that the world is wisely and righteously ordered. For 
 my faith in a man only enables me to be confident 
 that his ends will be well chosen and pursued. They 
 may not be attained, for that may be prevented by 
 forces beyond his own control. All that such faith 
 then can teach us of God is that his end is the good 
 ordering of the universe. If we are to be certain that 
 his end will be attained, we must, in some other 
 manner, have estabhshed the dogma that God is either 
 omnipotent, or at any rate so powerful as to be free 
 from the risk of eventual defeat. 
 
 56. But, even with these limitations, should we be 
 entitled to trust God as we might trust a man whose 
 designs we could not follow ? I do not think that we 
 should. Of course, if we have convinced ourselves 
 by direct argument that God is working towards the 
 good and is strong enough to realize it, then we should 
 be sure of his intentions and his success in cases 
 where we could not divine his purpose or directly 
 perceive his success. But this would not be the faith 
 of which we speak. It would be demonstration. If 
 God's nature is proved to be such that he always 
 wills the good, then in any particular case it is clear 
 that he does will it. The faith in God which we are 
 now discussing is a substitute for demonstration, not 
 an instance of it. 
 
 We do not put faith in all men, but only in some of 
 them. Why is this ? I take it that our faith is an 
 induction from experience. If a man has always 
 acted honourably, and now acts in a way which, viewed 
 in itself, would seem to us dishonourable, we think it 
 more probable that our judgement on the character of 
 this action is mistaken, than that our mass of pre- 
 vious judgements on his former conduct should be 
 
THE ESTABLISHMENT OF DOGMA 71 
 
 mistaken, or that his character should have suffered 
 a sudden change for the worse. We say, therefore, 
 that we are confident that circumstances, at present 
 unknown to us, prevent the action from being 
 dishonourable. 
 
 Sometimes, indeed, we have faith in a man of whom 
 we know nothing. But, after all, we know that he is 
 a man. And our faith in this case is based on the 
 experience of the past which tells us that men have 
 never, or very rarely, been known to commit the 
 particular crime or to omit the particular duty in 
 respect of which we have faith in the man before us. 
 
 Faith in man, then, rests on an induction — an 
 induction from the previously observed conduct of the 
 man in question, or of men in general. Can such an 
 induction be legitimately made in the case of God, 
 supposing his existence as a person to be already 
 proved ? I do not think it can be. For consider how 
 enormous is the scope of the conclusion of the inference 
 as compared with the scope of the observations on 
 which it can be based. What are the limits of our 
 knowledge of what are, on this theory, God's acts? 
 We know a very few of those which have happened 
 on one planet for a few thousand years, together with 
 a few isolated facts about events beyond this planet. 
 On the strength of these we are invited to beheve in 
 a uniform law applying to all his actions for the whole 
 universe. We know that the universe is much larger 
 than our sphere of observation. It is perhaps even 
 infinitely larger. Surely to conclude from so little to 
 so much would be ineffably rash. Judas, according to 
 the old legend, once gave a cloak to a leper. If any 
 man who had observed this had argued that the whole 
 of the life of Judas was one continuous succession of 
 virtuous deeds, his conclusion would be admitted to 
 
72 SOME DOGMAS OF EELIGION 
 
 be unfounded. Yet that action would have borne a 
 larger proportion to the whole life of Judas, than those 
 of God's actions which we directly know can be 
 supposed to bear to the whole of his actions. 
 
 57. This objection would, I think, be fatal even if 
 every one of the divine actions which we directly 
 know was one which suggested a good purpose. But 
 it is universally admitted that this is not the case, and 
 that many of those facts of the universe which we 
 directly know do not suggest a good design as their 
 most obvious explanation. I will not inquire whether 
 the facts known to us show more good than evil. It is 
 certain that the two are so nearly balanced, that both 
 the evil and the good have been considered, by different 
 observers, to be the greater. Such a mixed experience 
 as this is possibly not incompatible with the existence 
 of a God who is working for the good and who is 
 strong enough to succeed. But it does not form an 
 adequate proof of his existence. 
 
 It is clear that we cannot put faith in God without 
 reference to our knowledge of his actions, although we 
 can, in certain cases, do this with man. We trust 
 them from our experience of other men. But if there 
 is a God, he is probably not one of a class of Gods, 
 and, if he were, we should know no more of them 
 than of him. It is true that a personal God would be 
 included with men in the class of persons, but the 
 difference between God and man would be far too 
 great to justify any inference from the conduct of the 
 one to the conduct of the other. 
 
 58. There are, of course, cases in which faith does 
 not depend on induction— or indeed on any basis of 
 reason at all. If A loves B, it is frequently the case, 
 though by no means always, that he will decline to 
 adopt any theory of B's action which involves blame 
 
THE ESTABLISHMENT OF DOGMA 73 
 
 to Bj even if the facts are such that he would have 
 adopted the theory without hesitation in the case of any 
 person whom he did not love. It has been suggested 
 that this is the right analogy for our faith in God. If 
 we love him, it is said, we shall have faith in his 
 working for good, and that faith will be justifiable. 
 
 If this is even psychologically possible, it is clear 
 that it must be possible to love God before deciding 
 whether he is good or not — ^perhaps a better expression 
 would be, to love a being otherwise answering to the 
 description of God, before deciding whether he pos- 
 sesses the remaining divine quality of goodness. For, 
 if a belief in the goodness has to precede the love, it 
 is clear that the love cannot be the ground of the 
 belief in goodness. 
 
 It is, fortunately, possible to love men regardless of 
 their wickedness. Could it be so with God ? Could 
 we love an otherwise divine being who was wicked ? 
 If he were conceived as omnipotent, or as only self- 
 limited, I should imagine it to be impossible — though 
 it is difficult to be certain. But if he were conceived 
 as a finite being of hmited power, working in a world 
 of which he was not the ultimate creator, then I can 
 see no difficulty in a man loving him, irrespective of 
 his goodness or wickedness, as a man might be loved. 
 
 This conception of God is not, of course, a very 
 common one. But for those who hold it the position 
 which we are discussing is, no doubt, psychologically 
 possible. They can love God, having satisfied them- 
 selves of his existence, with a love as ultimate and 
 super-moral as that of friend for friend. And then, 
 no doubt, they may be led on from this to assume 
 his goodness when there is no evidence, or insufficient 
 evidence, for it. 
 
 Such a position is possible, then, but is it justifi- 
 
74 SOME DOGMAS OF EELIGION 
 
 able ? The analogy, at any rate, is not encouraging. 
 For, after all, the object of any belief is to gain the 
 truth. A belief which does not do this stands 
 condemned. Now it is notorious that the faith in 
 a person which is based on love for him frequently 
 does lead to a false belief, and not to a true one. If 
 a man's mother is convinced of his innocence when 
 a jury is convinced of his guilt, experience shows that 
 the jury is much more likely to be right than the 
 mother. 
 
 Nor is faith in the goodness of the beloved essential 
 to love, as has sometimes been maintained. It is 
 fortunate that it is not. For there are cases where 
 the blindest faith in a man's goodness must give way 
 before the demonstration that he is bad. And, if 
 faith in goodness were essential to love, then love 
 would have to cease when this happened. But it 
 need not cease then, and it often does not. We are 
 better than Tennyson — or at least Guinevere inter- 
 preted by Tennyson — made us out to be. We need 
 not love the highest when we see it. And we can 
 love the lowest when we see it — when the lowest 
 happens to be the person we love. 
 
 59. The conclusion I submit is that on matters of 
 dogma we cannot dispense with proof, and we cannot 
 prove anjrfching by considering the people who beheve 
 it, or the disastrous consequences which would follow 
 from its falsity. We must prove our dogmas more 
 directly, if we are to continue to beheve them. This 
 will leave all questions of dogma more or less pro- 
 blematic, and many quite unanswered. And this 
 is doubtless unpleasant. But unpleasant things are 
 sometimes true. 
 
 Is there then no moral element involved in belief in 
 religious dogma ? I believe that to acquire true belief 
 
THE ESTABLISHMENT OF DOGMA 75 
 
 in religious dogma does require moral qualities — in 
 almost every case — in the seeker. But they are 
 required, not to show us what the truth is — for that 
 purpose they seem to me as useless to the metaphysician 
 as to the accountant — but to prevent our turning away 
 from the truth. In the first place, a man will scarcely 
 arrive at truth in these questions without courage. 
 For he must seek before he can find, and at the 
 beginning of his search he cannot tell what he will 
 find. 
 
 And he will also need — unless he is almost in- 
 credibly fortunate — a certain form of faith. He will 
 need the power to trust the conclusions which his 
 reason has deliberately adopted, even when circum- 
 stances make such a beHef especially difficult or painful. 
 There are leaden days when even the most convinced 
 idealist seems to feel that his body and his furniture 
 are as real as himself, and members of a far more 
 powerful reality. There are times when the denial 
 of immortality seems, to the firmest disbeliever in 
 immortality, a denial which he has scarcely strength 
 to make. 
 
 But, whatever is true, it is quite certain that truth 
 is not affected by incidents like these. If all reahty 
 has been proved to be spiritual, it cannot have ceased 
 to be spiritual because to-day I am ill or overworked. 
 If I had no reason to beheve in immortahty yesterday, 
 when other people's friends were dead, I have no 
 greater reason to beheve in it because to-day my 
 friend has died. 
 
 If we want to know the truth, then, we must have 
 faith in the conclusions of our reason, even when they 
 seem — as they often will seem — too good or too bad 
 to be true. Such faith has a better claim to abide 
 with hope and love than the faith which consists in 
 
76 SOME DOGMAS OF EELIGION 
 
 believing without reasons for belief. It is this faith, 
 surely, wliich is sought in the prayer, ' Suffer us not 
 for any pains of death to fall from thee/ And for 
 those who do not pray, there remains the resolve that, 
 so far as their strength may prevail, neither the pains 
 of death nor the pains of Kfe shall drive them to any 
 comfort in that which they hold to be false, or drive 
 them from any comfort in that which they hold to 
 be true. 
 
CHAPTER III 
 
 HUMAN IMMORTALITY 
 
 60. I DO not propose to offer here any arguments in 
 support of the positive assertion that men are immortal, 
 I believe that such arguments exist, and that, in spite 
 of the difficulty and obscurity of the subject, they are 
 of sufficient strength to justify a belief in our immor- 
 tality.^ But to expound these arguments would require 
 an elaborate and lengthy treatise of technical meta- 
 physics, for they could only be proved by a demon- 
 stration of some ideahst theory of the fundamental 
 nature of reality. My present design is merely to 
 consider some arguments against immortality which 
 have been based on certain facts of ordinary observa- 
 tion, and on certain results of physical science. I shall 
 endeavour to show that those arguments are invalid, 
 and that the presumption against immortality, which 
 they have produced in many people, should be dis- 
 carded. 
 
 It is better to speak of the immortality of the self, 
 or of men, than of the immortahty of the soul. The 
 latter phrase suggests untenable views. For, in speak- 
 ing of the identity of a man during different periods 
 of his bodily life, we do not usually say that he is the 
 same soul, but the same self, or the same man. And 
 to use a different word when we are discussing the 
 prolongation of that identity after death, calls up the 
 
 ^ Cp. my Studies in Hegelian Cosmology, chap. ii. 
 
78 SOME DOGMAS OF EELIGION 
 
 idea of an identity less perfect than that which lasts 
 through a bodily life. The form in which the question 
 is put thus implies that the answer is to be in some 
 degree negative — that a man is not as much himself 
 after death as he is before it, even if something escapes 
 from complete destruction. 
 
 Moreover, it is customary, unfortunately, to say that 
 a man has a soul, not that he is one. Now if our 
 question is put in the form ' Has man an immortal 
 soul ? ' an affirmative answer would be absurd. So 
 far as it would mean anything it would mean that the 
 man himself was the body, or something which died 
 with the body — at any rate was not immortal — and 
 that something, not himself, which he owned during 
 life, was set free at his death to continue existing on 
 its own account. For these reasons it seems better 
 not to speak of the soul, and to put our question in 
 the form * Are men immortal ? ' 
 
 61. What reasons are there for supposing that our 
 existence is only temporary ? I see around me bodies 
 which behave so like my own, that I conclude that 
 they are related to other conscious selves in the same 
 way that my body is related to myself. But from 
 time to time these bodies are observed to cease to 
 behave in this way, and to become motionless, unless 
 moved from outside. Shortly after this the body 
 dissolves into its constituent parts. Its form and 
 identity as a body are completely destroyed. The 
 experience of the past leads me to the conclusion that 
 the same thing will happen in the future to every 
 human body now existing, including my own. 
 
 How does this affect the question of my existence ? 
 It is clear that if I am a mere effect of my body — 
 a form of its activity — I shall cease when the body 
 ceases. And it is also clear that, if I could not exist 
 
HUMAN IMMOETALITY 79 
 
 without this particular body, then the destruction of 
 the body will be a sign that I have ceased to exist. 
 
 But, besides death, there is another characteristic 
 of nature which tends to make us doubt oiu' immor- 
 tality. Of all the things around us, from a pebble to 
 a solar system, science tells us that they are transitory. 
 Each of them arose out of something else, each of them 
 will pass away into something else. What is a man 
 that he should be exempt from this universal law ? 
 
 Thus we have three questions to consider : (1) Is 
 my self an activity of my body? (2) Is my present 
 body an essential condition of the existence of my 
 self? (3) Is there any reason to suppose that my 
 self does not share the transitory character which 
 I recognize in all the material objects around me? 
 
 62. With regard to the first of these questions, it is 
 certain, to begin with, that my body influences my 
 self much and continuously. A large part of my 
 mental life is made up of sensations. Sensations are 
 continually produced in connexion with changes in 
 the sense-organs of my body, and, so far as we know, 
 they are never produced in any other way. And the 
 course of my thoughts and emotions can be profoundly 
 affected by the state of my body. If my body gets no 
 food for twenty-four hours, they will be affected one 
 way. If I introduce whisky or opium into it, they 
 will be affected another way. If my body is very 
 fatigued, the ordinary current of my mental life will 
 be entirely suspended in profound sleep, or completely 
 broken by dreams. If any of these processes is carried 
 far enough, my body dies, and I cease to have any 
 relation to it for the future, which is certainly an 
 important event for me, whether I survive it or not. 
 
 It is equally certain that the mind acts on the body. 
 
80 SOME DOGMAS OF EELIGION 
 
 My limbs, on many occasions, move according to my 
 will. And the normal behaviour of the body can be 
 altered by the mind, as much as the normal behaviour 
 of the mind can be altered by the body. Grief, or 
 fear, or anger, can produce bodily illness, and even 
 death. 
 
 Now each of these groups of events — the effects of 
 body on mind and of mind on body — could be explained 
 on the hypothesis that the self and the body were two 
 separate realities, neither of which was the mere 
 product of the other, though each affected the other, 
 and caused changes in it. And it might be thought 
 that this would be the most natural conclusion to 
 adopt, since the action appears to be reciprocal — mind 
 acting on body as much as body acts on mind. 
 
 63. There is always, however, a very strong tendency 
 to adopt the view that the self is a mere activity of the 
 body — or at any rate to hold that the only escape from 
 this view lies in accepting some form of revealed 
 religion which denies it. The cause of this tendency 
 is, in the first place, the incomplete nature of the 
 explanation which would be furnished by the recogni- 
 tion of the self and its body as independent realities. ^ 
 All ultimate explanation endeavours to reduce the 
 universe to a unity. The self is spirit, the body is 
 matter. Spirit and matter, taken as independent 
 realities, are very heterogeneous to one another. It is 
 evident that a theory which makes either spirit or 
 matter to be the sole reality in the universe, introduces 
 a greater degree of unity than a theory which makes 
 them to be equally real. 
 
 ^ By independent I do not here mean isolated, or unconnected 
 realities, but such as stand on an equal footing, so that though 
 each is connected with the other, neither is subordinate to the 
 other. 
 
HUMAN IMMOETALITY 81 
 
 Monism then, whether it be materialism or idealism, 
 is more attractive to the majority of inquirers than 
 dualism is. We must now consider the various causes 
 which tend to make a materialistic monism more 
 plausible than an idealistic monism, and which impel 
 us to the conclusion that matter is the only reality, 
 while human spirit is nothing more than one of the 
 activities which characterize matter when it is in the 
 special form of a human body. (It is immaterial for 
 our present purpose whether the adherents of this view 
 suppose matter to exist as a substance, to which these 
 activities belong, or whether they say that the activities 
 are the matter. The difference is insignificant, although 
 the second alternative is sometimes put forward as 
 a great improvement on the first. The essential point 
 is that the spiritual is in either case reduced to a 
 temporary form of an activity whose fundamental 
 nature is non-spiritual.) 
 
 64. One of these causes is the fact, so continually 
 pressed upon the notice of every man, that the nature 
 of matter is almost entirely independent of his will. 
 I cannot create matter, and there are narrow limits to 
 the extent to which I can alter it. I cannot make 
 into bread the stone which I see and touch. How- 
 ever passionately I may desire that it should be bread, 
 however serious the consequences to myself and 
 others of its remaining a stone, a stone it remains. 
 By a transition which is natural though illegitimate 
 we tend to believe that whatever is so independent 
 of our will must be independent of us altogether. 
 
 To some extent, indeed, the will can affect matter. 
 But the amount of its effects is comparatively insigni- 
 ficant. All the exertions of human beings can only 
 affect the surface of the earth, and that very slightly. 
 On the other hand, matter seems far more powerful 
 
 a 
 
82 SOME DOGMAS OF EELIGION 
 
 in its influence on spirit. The diminution of the 
 temperature of a single planet is an absurdly trivial 
 episode in astronomy. But, if the planet were our 
 earth, it would put an end to the only conditions 
 under which, as far as our observation goes, it is 
 possible for spirit to exist. Since spirit, then, appears 
 so much weaker than matter when they are taken 
 separately, is it strange that, when an attempt is made 
 to reduce the one to the other, it is spirit that is called 
 on to give way ? 
 
 In matter, too, we can observe a unity and a per- 
 sistence which may belong to spirit but does not 
 obviously belong to it. Spirit we only know in the 
 form of separate individuals, set in the midst of 
 matter, only by means of which they are able to 
 communicate with one another. No human spirit has 
 ever, as far as we know, been open to observation for 
 much more than a hundred years, and the lower 
 animals only slightly exceed this limit. Matter forms 
 one vast system, which history informs us has existed 
 for thousands of years, while science extends the 
 period to millions. 
 
 And, again, the amount of knowledge which science 
 gives us about matter is far greater than the amount 
 which it gives us about spirit. On the one side is the 
 whole vast extent of the physical sciences. On the 
 other side we have only psychology — and not the whole 
 of psychology. For the psycho-physical side of that 
 science deals as much with matter as it does with 
 spirit. 
 
 All this increases the apparent importance of matter, 
 and seems to render it more probable that matter, rather 
 than spirit, is the sole reality. Spirit, then, would be 
 the way in which matter behaves under certain circum- 
 stances. And in support of this it may be said that 
 
HUMAN IMMOETALITY 83 
 
 the activity of matter does take different forms. The 
 same energy, science informs us, which sometimes 
 shows itself as heat, shows itself at other times as 
 motion, or, again, as electricity. And this same 
 energy, it is asserted by the materialist, is transformed 
 under other circumstances — when it is found in a 
 human body — into thought, will, and emotion. Cer- 
 tainly, he admits, thought, will, and emotion are not 
 very like heat, motion, and electricity. But then 
 heat, motion, and electricity are not very like one 
 another. And, if they can all be reduced to this 
 common unity, why should not the forms of conscious- 
 ness share the same fate ? 
 
 65. These conclusions depend, it will be seen, on 
 the proposition that matter can exist independently 
 of spirit. For if this were not so, it would obviously 
 be absurd to explain away the separate reality of spirit 
 by making it one of the temporary forms which the 
 activity of matter takes. Deeper inquiry will, I think, 
 show us that matter cannot be independent of spirit, 
 that, on the contrary, matter is only an appearance to 
 the mind which observes it, and cannot, therefore, 
 exist independently of spirit. If this is the case we 
 cannot be entitled to consider the self as the activity 
 of its body. 
 
 Of what nature is the matter supposed to be which, 
 it is asserted, can exist independently of spirit ? It is 
 not conceived as having all the qualities which, in 
 ordinary language, we ascribe to matter. We say of 
 an orange that it is soft, yeUow, sweet, and odorous. 
 But these qualities are not held to belong to the 
 orange when it is not being observed. In strictness 
 they are not held to be qualities of the orange at aU, 
 but effects excited in the observer by qualities of the 
 orange. The orange is no more yellow when no one 
 
 G 2 
 
84 SOME DOGMAS OF EELIGION 
 
 sees it than it is desired when no one know^ of its 
 existence. 
 
 But the object is conceived as having other qualities 
 which really do belong to it, and give it that nature 
 which it has independently of observation, and if no 
 one observes it. Its size, its shape, its position in 
 space, its motion, and its impenetrabihty are of this 
 nature. It is these qualities, or others of the same 
 nature, which have the power, under certain circum- 
 stances, of exciting in the observer the sensations of 
 softness, yellowness, and the hke. 
 
 The quahties which are held really to belong to the 
 matter are often called its primary qualities. The 
 others are called its secondary qualities, though, on 
 this theory, it is scarcely correct to call them qualities 
 of the object at all. 
 
 Matter, then, is held to be eitended, to have position, 
 and to be capable of motion independently of observa- 
 tion. It is also impenetrable — that is, no two pieces 
 of matter can occupy the same position in space. 
 But it has no colour, it is neither hard nor soft, it has 
 no taste, no smell, and no sound. 
 
 This is matter as it is conceived in physical science. 
 It may be said also to be the ordinary conception, for 
 although we speak of an orange as yellow, yet the idea 
 that it is not yellow in the dark is generally known 
 and generally accepted. 
 
 66. What reason can be given for a belief in the 
 existence of matter? I conceive that such a belief can 
 only be defended on the ground that it is a legitimate 
 inference from our sensations. 
 
 This view has been contested, but I believe that the 
 objection to it rests on a misunderstanding. It has 
 been said, and with perfect truth, that my belief in the 
 existence of matter does not arise as an inference from 
 
HUMAN IMMOETALITY 85 
 
 my sensations. I do not first become aware of my 
 sensations and then infer the existence of an orange. 
 On the contrary, I am aware of the existence of the 
 orange first. If I am studying psychology or am 
 doubtful of the validity of my knowledge I may then 
 consider the sensations of sight, touch, and so on, con- 
 nected with my knowledge of the object. But in most 
 cases I never do consider the sensations at all. And 
 there are young children who are quite aware of the 
 existence of a material world, but who have never 
 realized that they have sensations. 
 
 These facts are sufficient to refute the view, which 
 has sometimes been held, that our belief in a material 
 world arises as an inference from our sensations. But 
 they are quite irrelevant to the question now before 
 us — whether our belief in a material world must not 
 be justified, if it is to be justified at all, as an inference 
 from our sensations. And when such facts are used, as 
 not infrequently happens, as bearing on this question, 
 involves a very serious confusion. 
 
 The belief in a material world requires justification. 
 It is natural, in the sense that every one who has not 
 reflected on the subject holds the belief as a matter of 
 course, together with many of those who have reflected 
 on it. But it is not inevitable. It is possible to dis- 
 believe it. *Many philosophers have done so. And 
 there is, at any rate, nothing obviously self-contra- 
 dictory in its denial. Berkeley's theory on the subject 
 — to take only one out of many theories which deny 
 the existence of matter — whether true or false, is not 
 self-contradictory. 
 
 Since disbelief in the existence of matter is neither 
 impossible nor contradictory, the question becomes 
 inevitable — what is the justification of the belief? 
 And it becomes more pressing, because in many cases 
 
86 SOME DOGMAS OF KELIGION 
 
 our judgements as to the existence of matter are 
 admitted to be wrong. In the first place, the quite 
 unreflective consciousness has no more doubt that the 
 world of matter is coloured than it has that the world of 
 matter is extended. But either this or the more reflec- 
 tive judgements of science and the modern world must 
 be wrong here, since they disagree. Again, if a man, 
 who sees a cloak hanging up by moonlight, believes 
 that he sees before him the body of a dead friend, it is 
 obvious that he has completely mistaken the character 
 of the matter before him. And if our judgements as 
 to what the external object is are so often wrong, we 
 have little justification for assuming without inquiry 
 that our judgement that there is an external object is 
 ever right. 
 
 There is a stronger case than this. For in dreams 
 we do not only make wrong judgements as to the 
 nature of matter, but as to the existence of matter. 
 If a believer in the existence of matter dreams that he 
 sees a roc's egg, he no more doubts, during his dream, 
 that the roc's egg exists as independent matter, than 
 he doubts, during waking life, that his table exists as 
 independent matter. And yet, on waking, he will 
 admit that in his dream he was neither observing a 
 roc's egg nor any other really existing matter which 
 he mistook for a roc's egg. Not only was his dream- 
 belief * this is a roc's egg ' mistaken, but his dream- 
 behef ' this is independently existing matter ', was also 
 mistaken. And if this is mistaken, it is mere credulity 
 to trust his belief in the table's existence without 
 examination. For that belief is no stronger and no 
 more evident than the other had been previously. 
 
 67. On what can we base a justification of the belief 
 in the independent existence of matter ? Nothing is 
 available except the sensations. They are there, and 
 
HUMAN IMMOETALITY 87 
 
 they are certain. We never believe that we are 
 observing matter unless we experience sensations more 
 or less analogous to the qualities we believe to exist in 
 the matter. We may not be conscious of the sensa- 
 tions as such at all. Indeed, as was said above, in the 
 majority of cases we never are conscious of them. But 
 whenever we look for them, on such occasions, we find 
 them. And the sensations are certain. I may be 
 wrong in believing that matter exists independently 
 of me. But the suggestion that I am wrong in believ- 
 ing I have a sensation is absurd. The belief is not 
 sufficiently separable from the sensation for the possi- 
 bihty of error. I may, of course, be wrong in believing 
 that I had a sensation in the past, for memory may 
 deceive me. And I may be wrong in the general 
 terms which I apply to a sensation, when I attempt to 
 classify it, and to describe it to others. But my know- 
 ledge that I am having the sensation which I am 
 having is one of those ultimate certainties which it is 
 impossible either to prove or to deny. 
 
 And we find that although the sensations are gene- 
 rally ignored, as sensations, when the correctness of 
 the judgement about the matter is not doubted, yet, 
 as soon as I myself, or other people, entertain a doubt 
 of the correctness of the judgement, the situation is 
 changed. If it is suggested that what I believe to be an 
 experience of matter of a certain sort is really a dream 
 or a delusion, I fall back on the sensations which 
 I have experienced, and consider whether they can be 
 accounted for on any other hypothesis than that of the 
 existence of the matter in question. If they cannot, 
 I consider that I was right in my judgement that the 
 matter did exist. 
 
 And we must act in the same way if a doubt arises, 
 not merely of the correctness of our judgements that 
 
88 SOME DOGMAS OF EELIGION 
 
 this or that matter exists, but of the correctness of all 
 judgements that matter of any sort exists. The fact 
 which it is impossible to deny is that we have the 
 sensations. Are we entitled to conclude from this 
 that the material world really exists, and that the 
 natural judgement that it does exist — which is not, 
 however, as we have seen, an inevitable or universal 
 judgement — is correct? 
 
 68. It is evident that the sensations are not them- 
 selves the matter in question. A sensation is not 
 matter, and it cannot exist apart from the self to whom 
 it belongs. It can have no independent existence. 
 But the sensations, since they begin to exist, must 
 have causes. ^ Now it cannot be said to be obviously 
 impossible that all the causes of my sensations should 
 lie within my own nature. It is certain that they do 
 not lie within that part of my own nature of which I 
 am conscious, for I am not conscious of producing my 
 sensations. But it might be said, as Leibniz has said, 
 that all my sensations arise out of the depths of my 
 own unconscious nature, and that when a self has 
 once come into existence it is as independent of out- 
 side influences in its sensations as a clock, when once 
 wound up, is in striking. But there are difficulties in 
 the way of this view into which we have no time to 
 enter, and I do not wish to lay any weight on the 
 possibility of its truth. I am prepared to admit — 
 what seems to me by far the more probable view — 
 that all my sensations have causes which are not my- 
 self nor anything in myself. Such causes must in 
 each case be merely part-causes. I am unquestionably 
 
 ^ This step might not be accepted by any one who denied the 
 universal application of causality, A thinker who denied the 
 universal application of causality could not, as far as I can see, 
 have the least justification for a belief in the existence of matter. 
 
HUMAN IMMOETALITY 89 
 
 one of the causes of my own sensations, for, if I did 
 not exist, my sensations also would not exist. 
 
 It may thus be admitted that my sensations make 
 it, at any rate, highly probable that some reality exists, 
 which is not myself or anything within myself, but 
 exists independently of me. But we have not got to 
 matter. A reality which exists independently of me 
 need not be matter — it might, for example, be another 
 spirit. We do not call anything matter unless it 
 possesses the primary qualities of matter given above. 
 These qualities correspond to certain sensations, or 
 elements in sensations, and the presence of the sensa- 
 tion in me is held to prove the existence of the cor- 
 responding quality in the material object, 
 
 69. But is this legitimate? The independent reality 
 has been admitted to be the part-cause of the sensa- 
 tions, but that does not prove that it is like them. 
 Causes do not necessarily resemble their effects. * 
 Happiness in A does not resemble the misery which 
 it may cause to the envious B. An angry man does 
 not resemble a slammed door. A ray of sunshine 
 does not resemble a faded water-colour. 
 
 70. And, on this very theory, the external causes 
 of all mental events do not resemble those events. 
 When I see a sphere of red-hot iron I have sensations 
 of form, sensations of colour, and (if I am near enough) 
 a feehng of pain. Now the ordinary theory of matter 
 makes the matter the cause of the sensations of colour 
 and of the feehng of pain, as ranch as of the sensations of 
 form. Yet it denies that the matter is red or painful. 
 Here, therefore, is an external cause of mental events 
 which does not resemble them. It is therefore impos- 
 sible to fall back on the principle, that the external cause 
 of mental events always resembles them. And what 
 other principle have we to justify us in ascribing 
 
90 SOME DOGMAS OF EELIGION 
 
 the primary qualities to the external causes of the 
 sensations ? ^ 
 
 The distinction between the primary and secondary 
 qualities renders the theory of the existence of matter 
 less tenable than it would otherwise be. In the first 
 place, there is the inconsistency, which we have just 
 noticed, of asserting that we can argue from some of 
 our sensations to a resemblance in their causes, and 
 not from others. If our perception of the secondary 
 qualities varies from time to time, and from individual 
 to individual, so also does our perception of the primary 
 qualities. If our perception of the primary quahties 
 exhibits a certain uniformity from time to time, and 
 from individual to individual, so also does our per- 
 ception of the secondary qualities. 
 
 And, in the second place, matter, while extended 
 and impenetrable, is destitute both of colour and of 
 hardness, since these are secondary qualities. Now 
 the sensations of extension and impenetrability only 
 come to us by sight and touch. When they come 
 by sight they are invariably conjoined with sensations 
 of colour, when they come by touch they are invariably 
 conjoined with sensations of hardness. We cannot 
 even imagine to ourselves a sensation which gives 
 extension without giving either colour or hardness. 
 
 Thus the theory which makes the external causes of 
 our sensations material, reaches a climax of incon- 
 
 ^ It must be noticed that the resemblance which the theory- 
 attributes to the sensations and their external causes is very 
 limited. The causes are not sensations, nor are their qualities 
 sensations. All that can be said is that, in some way not too 
 easy to define, certain predicates of the causes resemble the 
 content of some of the sensations which are the effects of those 
 causes. But it is not necessary for my argument to follow out 
 the ambiguities and difficulties which follow from this elaborate 
 combination of similarity and difference between sensations and 
 matter. 
 
HUMAN IMMORTALITY 91 
 
 sistency. Its one defence was the principle that the 
 causes of the sensations must resemble the sensations 
 they cause. But now it turns out that that what the 
 causes are to resemble is a mere abstraction from our 
 sensations — a naked extension, which is so far from 
 being a sensation which we experience, that we cannot 
 even imagine what a sensation of it would be like. 
 
 71. Is it possible to avoid this inconsistency by 
 dropping the distinction between primary and secondary 
 qualities? Shall we say that matter has not only 
 shape, size, position, motion, and impenetrability, but 
 also colour, hardness, smell, and taste? This view 
 certainly avoids some of the objections to the more 
 ordinary theory. It does not make an arbitrary and 
 gratuitous difference in the treatment of two sets of 
 qualities. And it gives matter a nature not utterly 
 unlike our experience, and not utterly unimaginable 
 by us. 
 
 But on the other hand the theory would no longer 
 have the support of physical science. For that science 
 treats matter as devoid of the secondary qualities, and 
 it endeavours to show that the primary qualities of 
 matter, under certain circumstances, excite in us the 
 sensations of the secondary qualities. 
 
 Of course the independent existence and ultimate 
 nature of matter is a question for metaphysics and not 
 for science. And therefore a metaphysical theory that 
 matter possesses the secondary qualities as well as the 
 primary cannot be upset by the fact that science, 
 working from its own more superficial point of view, 
 finds it convenient to treat matter as possessing only 
 the primary qualities. If science keeps to its own 
 sphere, it cannot clash with any metaphysical theory. 
 If attempts are made to treat its results as if they were 
 metaphysical truths, they have no claim to validity in 
 
92 SOME DOGMAS OF KELIGION 
 
 this sphere, and a metaphysical theory is none the 
 "worse for being incompatible with these misapplica- 
 tions. 
 
 But the theory that matter exists depended very 
 largely for its plausibility on the illegitimate support 
 which it obtained by taking science as if it were 
 metaphysics ; and if it loses this support, as it must 
 in the suggested new form, it loses, indeed, no real 
 strength, but much of what caused people to believe 
 it. As has been already said, the fact that physical 
 science treats matter as independent of spirit, and that 
 physical science forms a vast system, coherent, accepted, 
 and, from its own standpoint, irrefutable, has done 
 much to strengthen the beUef that matter, at least, 
 must be real, and that, if one of the two must be 
 explained away by the other, it is spirit which must 
 go, and matter which must stay. The inference is 
 quite illegitimate, since nothing in physical science 
 touches, or can touch, the question of the independent 
 existence of matter. But it is an inference which is 
 frequently made. And when the theory of the in- 
 dependent existence of matter defines the nature of 
 that matter in a manner completely different from the 
 definitions of physical science, it will no longer be able 
 to gain apparent support in this way. 
 
 72. Nor does the amended theory, while less incon- 
 sistent than the original form, altogether avoid 
 inconsistency. The red-hot sphere of iron is now 
 admitted not only to be a sphere, independent of any 
 observer, but to be red, independent of any observer. 
 But the pain still remains. It is not asserted that the 
 iron is painful, although it causes me pain. Now the 
 pain is a result produced in the observer which is 
 quite as real as the sensations of form and colour, and 
 quite as independent of the observer's will. It is like- 
 
HUMAN IMMOETALITY 93 
 
 wise just as uniform. The iron will not give me the 
 sensations except under certain conditions. (I shall 
 not see it to be red, for example, if I am blind, or have 
 my eyes shut.) And, under certain conditions, quite 
 as definite, it will inevitably give me the feeling of 
 pain. Yet nothing resembling the mental effect is 
 attributed to the cause in this case. Why should 
 a difference be made between this case and the 
 others ? 
 
 And, even if we limit ourselves to sensations, the 
 amended theory does iiot escape inconsistency. For, 
 even if the secondary quahties are predicated of matter, 
 it remains impossible to assert that matter is Hke the 
 sensations which it causes. These sensations change for 
 me from moment to moment. If I look at a thing under 
 one set of conditions, as to light and shade, I get one 
 sensation of colour from it ; if I change the conditions 
 next minute I get quite a different sensation. And if 
 two men look at it simultaneously under the different 
 conditions of light and shade they will have, simul- 
 taneously, the two different sensations of colour which 
 I had successively. Now it is impossible to suppose 
 that the object has at once two different colours. And 
 if it has only one, then that colour must differ, at least, 
 from one of the two sensations experienced by the 
 two observers, since these sensations differ from one 
 another. 
 
 The same is the case with the other secondary 
 quahties. And it is also the case with the primary 
 quahties. Two men who look at a cube from different 
 positions simultaneously have two quite different sensa- 
 tions of its shape — not merely numerically different, but 
 sensations which do not resemble one another. Yet 
 an object cannot have two shapes at once, and each of 
 these men would, under normal circumstances, agree 
 
U SOME DOGMAS OF EELIGION 
 
 about the shape of the object, although they started 
 from non-resembhng sensations. It is clear, therefore, 
 that the shape attributed to the object cannot resemble 
 the sensations of shape which it causes, since they do 
 not resemble one anothei\ 
 
 73. Now if it is once admitted that the qualities 
 attributed to the external object do not resemble the 
 quahties of the sensations it causes, we have no reason 
 to attribute those qualities to it at all. The only 
 reason we had for supposing the causes of our sensa- 
 tions to have these qualities was the supposed resem- 
 blance of the qualities to the sensations. But now it 
 becomes clear that the qualities attributed to the 
 causes, although partially resembling the sensations, 
 do not resemble them completely. It foUows that 
 a cause of a sensation may lack some of the qualities 
 of the sensation it causes. And in that case there 
 seems no reason for den3dng the possibility of its being 
 quite different, and having none of the qualities in 
 question. 
 
 It may be replied, no doubt, that it is nevertheless 
 possible that the causes of the sensations do possess 
 qualities partially resembling the sensations. The 
 causes exist, and must have some qualities. And it 
 may be these qualities which they have, and so they 
 may be entitled to the name of matter. But such 
 a possibility would be far too vague to give any 
 support to the theory that matter exists. They may 
 possess these qualities, for there is no reason why 
 a cause should not resemble its effect in certain 
 respects. But there is no reason to beheve that they 
 do possess them, or that their possession of them is 
 in the slightest degree probable. A man who boils 
 a lobster red may have a red face — there is nothing to 
 prevent it. But his action in causing the redness of 
 
HUMAN IMMOETALITY 95 
 
 the lobster gives us no reason to suppose that his face 
 is red.^ 
 
 The result is that matter is in the same position as 
 the Gorgons or the Harpies. Its existence is a bare 
 possibility to which it would be foolish to attach the 
 least importance, since there is nothing to make it at 
 all preferable to any other hypothesis, however wild. 
 
 74. If we ask, then, of what reality the vast mass 
 of knowledge holds true which science and everyday 
 life give us about matter, we must reply that it holds 
 true of various sensations which occur to various 
 men, and of the laws according to which these sensa- 
 tions are connected, so that from the presence of 
 certain sensations in me I can infer that, under certain 
 conditions, I shall or shall not experience certain other 
 sensations, and can also infer that, under certain con- 
 ditions, other men will or will not experience certain 
 sensations. 
 
 It will be objected that this is not what common 
 experience and science profess to do. When we say 
 that this bottle contains champagne, and this vinegar, 
 we are not talking about our sensations. And physical 
 science deals with such things as planets, acids, and 
 nerves, none of which are sensations. 
 
 It is quite true that it is usual to express the conclu- 
 sions of common experience and of science in terms 
 which assume the independent existence of matter. 
 Most people in the past have beheved that matter 
 
 * The statement that the bare possibility of the external causes 
 being material still remains open must be taken as referring only 
 to the arguments in this chapter. I believe that further con- 
 sideration should convince us, for reasons closely analogous to 
 those of Hegel and Lotze, that all substance must possess certain 
 characteristics which are essential to the nature of spirit, and 
 incompatible with the nature of matter. If this view is right — 
 a question beyond the purpose of this book to investigate — the 
 existence of matter would be positively disproved. 
 
96 SOME DOGMAS OF EELIGION 
 
 does exist independently ; our language has been 
 moulded by this belief, and now it is easier and 
 shorter to express our conclusions in this way. 
 Besides this, most people at present do hold the meta- 
 physical opinion that matter exists independently, and 
 tend to express themselves accordingly. 
 
 But the conclusions remain just as true, if we take 
 the view that matter does not exist. Something has 
 been changed, no doubt, but what has been changed 
 is no part either of common experience or science, but 
 a theory of metaphysics which forms no part of either. 
 And so we sacrifice neither the experience of every- 
 day life nor the results of science by den3dng the 
 existence of matter. We only sacrifice a theory of 
 metaphysics which we have already seen cannot be 
 justified. 
 
 I say, in ordinary language, that this is champagne 
 and this is vinegar. Supposing that there is neither 
 champagne nor vinegar as matter existing indepen- 
 dently of observation, but that it remains true that 
 a certain group of sensations of sight and smell is 
 a trustworthy indication that I can secure a certain 
 taste by performing cei-tain actions, and that another 
 group of sensations of sight and smell is a trust- 
 worthy indication that I can secure a different taste 
 by performing similar actions. Does not this leave 
 a perfectly definite and coherent meaning to the 
 experience of everyday life, which fits every detail of 
 that experience as well as the more common theory 
 does, and only differs from it on a question of meta- 
 physics ? 
 
 75. It is the same with science. Every observation 
 made by science, every uniformity which is estabhshed, 
 every statement as to the past or the future which it 
 asserted, would still have its meaning. The observa- 
 
HUMAN IMMOETALITY 97 
 
 tions would inform us of what had been experienced, 
 the uniformities would tell us the connexions of 
 various experiences, the statements as to the past and 
 future would tell us what has been or will be ex- 
 perienced, or would be so if the necessary conditions 
 were present. What more does science tell us, or 
 what more could it desire to tell us ? If the language 
 in which scientific results are generally expressed does 
 seem to tell us more, and to imply the independent 
 existence of matter, that is not science, but meta- 
 physics — the unconscious and uncritical metaphysics 
 of ordinary language — and its rejection does not 
 involve the rejection or the distrust of a single result 
 of science. 
 
 Science requires, no doubt, that experience should 
 exhibit certain uniformities, so that a certain expe- 
 rience can safely be taken as an indication of what 
 other experiences will follow it under certain con- 
 ditions. But this proves nothing as to the inde- 
 pendent existence of matter. If the external causes 
 of my sensations, and I myself, have a constant nature, 
 the sensations which are their joint result will exhibit 
 uniformities. And a non-material cause can have 
 a constant nature just as easily as a material cause 
 could have. 
 
 Science also requires that experience should have 
 a community of nature between different persons, so 
 that it shall be possible for us to infer from any 
 experience what the experience of another person 
 would be under conditions more or less similar. This, 
 again, can be explained as easily without matter as 
 with it. If my nature and that of other persons were 
 not more or less the same, our experience would not 
 be similar, whatever the nature of its external cause. 
 But if our natures resemble one another, then it is 
 
 H 
 
98 SOME DOGMAS OF RELIGION 
 
 obvious that the action on us of the same external 
 cause would produce results which resembled one 
 another, 
 
 76. The denial of matter, it must also be noticed, 
 does not lead us towards solipsism — that is to say, to 
 the denial by each individual of aU reality except 
 himself. The arguments which ^rove that my sensa- 
 tions must have causes which are not myself, nor in 
 myselfj but are some other reality, lose none of their 
 force if we decide that these causes are not of 
 a material nature* And the other arguments against 
 sohpsism — the consideration of which is apart from 
 our present object — are just as strong on the hypo- 
 thesis that matter does not exist. 
 
 It might be supposed that the theory I have been 
 advocating was a form of agnosticism. Agnosticism 
 holds that we can know nothing but phenomena. 
 Beneath these phenomena lies a reality on which they 
 are based, but of this reality, agnosticism declares, we 
 can know nothing. If we only know of the external 
 causes of our sensations that they do cause the sensa- 
 tionsj have we not in effect taken up the agnostic 
 theory that the reality on which phenomena depend 
 is unknowable ? 
 
 But this is not the case. Agnosticism says that we 
 can know nothing whatever of the reality behind the 
 phenomena. And, in sajdng this, it contradicts itself. 
 For it asserts that such a reality exists, and that it 
 stands in certain relations to the phenomena. Thus 
 we do know something about it, and it is therefore 
 not the case that we can know nothing about it. 
 
 But the theory which I have put forward does not 
 say that we can know nothing about the causes of 
 sensations. It only says that we do not know that 
 they are hke the sensation they cause. Even if this 
 
HUMAN IMMOETALITY 99 
 
 should destroy all knowledge of them except of the 
 fact that they were causes, it would not be a general 
 assertion of the impossibility of any knowledge of 
 them, and so there would be no inconsistency in 
 saying that we knew they were causes. To know m 
 of anything is inconsistent with being unable to know 
 anything of it, but it is quite consistent with knowing 
 nothing about it except m. 
 
 Nor does it follow that we know nothing else about 
 the causes of our sensations if we cannot conclude that 
 they resemble the sensations. It might be possible^ 
 as various philosophers have maintained, to determine 
 the qualities which must belong to every substance in 
 virtue of its being a substance. And it might turn 
 out that this could give us a considerable knowledge 
 of the nature of these substances. We might, for 
 example, be led to the conclusion that all substance 
 Was spirit. But we cannot here do more than point 
 out the possibility of such a result. 
 
 77. And we have thus, I think, proved our original 
 contention that the self cannot be one of the activities 
 of its own body. If the self were, as such a theory 
 would require it to be, merely a way in which matter 
 behaved under certain circumstances, it would be 
 possible to explain the self satisfactorily in terms of 
 matter. And it would be possible that a state of 
 things should exist in which those circumstances, 
 which determine the activity of matter to take the 
 form of spirit, occurred nowhere in the universe, which 
 would then be a universe of matter without any con- 
 sciousness. But so far is this from being the case that 
 we can, as we now see, only explain matter in terms 
 of a conscious self, and to talk of matter existing with- 
 out consciousness is absurd. Matter is so far from 
 being the sole reality, of which the self is only an 
 
 H 2 
 
100 SOME DOGMAS OF EELIGION 
 
 activity, that, taken by itself, it is not a reality at all. 
 The only things which have, in any sense, the qualities 
 attributed to matter, are the sensations experienced by 
 selves. In place of an independent reality we find events 
 in men's minds which are real, indeed, but not an inde- 
 pendent reality. Matter is simply our illegitimate 
 inference from these events. 
 
 This may be put in another way. If my self is one 
 of the activities of my body, then, since my body is 
 only events in the life of some conscious being, my 
 self must also be events in the life of some conscious 
 being. It is clearly absurd to suppose that I am an 
 activity of my body, as my body is known to myself, 
 for then I should be events in my own life. But it is 
 equally impossible that my self should be one of the 
 activities of its own body as perceived by some other 
 self. In that case the self A would be events in the 
 life of another self B. But how about B? By the 
 same rule it also will have to be events in the life of 
 another self. If this self is A, the absurdity will recur 
 in an aggravated form. For then A would be the 
 events which happened in a self which was itself events 
 in A, But if we say that B is events in the life of 
 a third self C, the same question will arise about C, 
 and so on without end. If every self is only events in 
 the life of some other self, no self is explicable until 
 we have reached the end of an infinite series — that is, 
 no self is explicable at all. And so we are brought 
 back to the conclusion that the self cannot be an 
 activity of its body. 
 
 I may be thought to have dwelt unnecessarily on 
 this point. Surely, it may be said, it is obvious that 
 the theory that the self is an activity of the body must 
 fall with the theory of the independent existence of 
 matter. Surely no one would maintain that the body 
 
HUMAN IMMORTALITY 101 
 
 only existed for spirit, and, at the same time, that 
 spirit was an activity of body. Yet this has been 
 done. Men of ability have maintained that what I 
 call matter is nothing but my thoughts and sensations, 
 and, at the same time, that my thoughts and sensations 
 are nothing but an activity of my brain— which, being 
 matter, will itself be thoughts and sensations ! To - v^ 
 
 78. The bearing of this discussion on the question 
 of our immortality is that it disproves a hypothesis 
 which would render immortality incredible. If the 
 self was an activity of the body, it would be impossible 
 that it should continue to exist when the body had 
 ceased to exist. We might as well suppose, in that 
 case, that the digestion survived the body as that the 
 self did. But the body, as we have now seen, only 
 exists for the selves which observe it, and we cannot, 
 therefore, reduce any self to be an activity of its own 
 body. 
 
 It has been admitted, indeed, that there is reahty 
 external to myself — the reality which includes the 
 external part-causes of my sensations — although we 
 are not justified in regarding that reahty as material. 
 And nothing that we have said excludes the possi- 
 bility that my self may be a product or activity of 
 some other reality, and one which is destined to cease 
 to exist when some change takes place in its cause. 
 
 But while this view has not been refuted, there is 
 not any reason, that I can see, why it should be held 
 to be true, or even probable. There is no reason why 
 we should regard our selves as the product or activity 
 of any other reality whatever, and there is no reason 
 why, if we did regard them as such products or 
 activities; we should consider them hkely to cease. ^ 
 
 ^ It is commonly held that human selves are not products of 
 non-divine realities, but that they are all produced by God. I have 
 
102 SOME DOGMAS OF RELIGION 
 
 If the external reality had been independently exist- 
 ing matter, it would have been different. In that case 
 there would have been, as we saw above, a strong 
 tendency to regard matter as the only ultimate reality, 
 and the self as an activity of its body. The tendency 
 would not be due to a logical necessity, since the facts, 
 as we have seen, would not be inconsistent with the 
 hypothesis that spirit and matter were independent, 
 though connected, realities. But the tendency would 
 be very strong, owing to our desire to find as much 
 unity as possible in the universe. If the self is an 
 independent reality, it is a non-material reality. And, 
 granted the independent existence of matter, more 
 unity would be gained by denying the independent 
 reality of spirit. But without independently existing 
 matter the case is changed. No increased unity is 
 gained by making the self a mere activity of something 
 else, unless that something else is already known to 
 exist and to be of a non-spiritual nature. Indepen- 
 dently existing matter would, of course, be of a non- 
 spiritual nature. But, when we have rejected this, 
 I have no reason to believe that the reality outside 
 myself is non-spiritual, and so I should gain no in- 
 creased unity for the universe by denying the inde- 
 pendent reality of my self. 
 
 And, again, if the self is an activity of its body, it 
 must be a temporary activity, since the body is only 
 a temporary combination of matter. But if the self 
 were an activity of some non-material reality outside 
 itself; there would be nothing to disprove the per- 
 manence of the state of things which produces the 
 
 given in Chapter VI the reasons why this view does not seem to 
 me to be necessary. But, supposing that they are produced in 
 this way, we should have no ground for supposing that their 
 divine production involved their subsequent destruction, though 
 it is not, of course, incompatible with such destruction. 
 
HUMAN IMMOETALITY 103 
 
 self—though, of course, there would equally be nothing 
 which proves that permanence. 
 
 79. We must now pass on to our second question, 
 My self cannot be a form of the activity of my body. 
 But it is still possible that the nature of my self makes 
 the possession of my present body essential to it. 
 Granted that the body could not exist except for 
 knowledge, it may be that the knowledge of my body, 
 by myself or other selves, is a necessary condition of 
 the existence of my self. In that case it would be an 
 inevitable inference that when my body dissolves, and 
 ceases to be known as a body at all, my self must 
 have ceased also. If A, whenever it exists, is neces- 
 sarily accompanied by B, then the cessation of B is 
 a sure sign of the cessation of A, 
 
 What evidence is there in favour of such a view ? 
 In the first place, while we have plenty of experience 
 of selves who possess bodies, we have no indubitable 
 experience of selves who exist without bodies, or after 
 their bodies have ceased to exist. Besides this, the 
 existence of a self seems to involve the experience of 
 sensations. Without them, the self would have no 
 material for thought, will, or feeling, and it is only in 
 these that the self exists. Now there seems good 
 reason to suppose that sensations never occur in our 
 minds at present without some corresponding modifica- 
 tions of the body. This is certainly the case with 
 normal sensations. And, even if the evidence for 
 clairvoyance and thought-transference were beyond 
 dispute, it could never prove the possibility of sensa- 
 tion without bodily accompaninients. For it could not 
 exclude — indeed, it seems rather to suggest — the 
 existence of bodily accompaniments of an obscure and 
 unusual kind. 
 
104 SOME DOGMAS OF EELIGION 
 
 80. But, after all, these considerations would, at the 
 most, go to show that some body was necessary to my 
 self, and not that its present body was necessary. 
 Have we, after the results already reached, any reason 
 to suppose that the death of the body must indicate 
 anything more than that the self had transferred its 
 manifestations to a new body, and had, therefore, 
 passed from the knowledge of the survivors, who had 
 only known it through the old body ? The apparent 
 improbability of this lies, I think, simply in our 
 instinctive recurrence to the theory that the self is an 
 activity of the body. In that case, no doubt, it would 
 be impossible that it should be successively connected 
 with two bodies. But that theory we have seen to be 
 untenable. The most that a body can be is an essential 
 accompaniment of the self. And then the supposition 
 that the self has another body would fit the facts quite 
 as well as the supposition that the self has ceased to 
 exist. 
 
 There seems no reason why such a change should 
 not be instantaneous. But even if it were not so, no 
 additional difficulty would be created. If a body is 
 essential to the action of a self, the self would be in 
 a state of suspended animation in the interval between 
 its possession of its two bodies — a state which we 
 might almost call one of temporary non-existence. 
 But this is nothing more than what happens, as far 
 as we can observe, in every case of dreamless sleep. 
 During such a sleep the self, so far as we know, is 
 unconscious — as unconscious as it could be without 
 a body. Yet this does not prevent its being the same 
 man who went to sleep and who woke up again. 
 Why should the difficulty be greater in a change of 
 bodies ? 
 
 81. And then, have we any reason, after all, to 
 
HUMAN IMMOKTALITY 105 
 
 suppose that a body is essential to a self? It seems to 
 me that the facts only support a very different pro- 
 position—namely, that, while a self has a lody, that body 
 is essentially connected with the self's mental life. 
 
 For example, no self can be conceived as conscious 
 unless it has sufficient data for its mental activity. 
 This material is only given, as far as our observations 
 can go, in the form of sensations, and sensations again, 
 as far as our observations can go, seem invariably 
 connected with changes in a body. But it does not 
 follow, because a self which has a body cannot get its 
 data except in connexion with that body, that it would 
 be impossible for a self without a body to get data in 
 some other way. It may be just the existence of the 
 body which makes these other ways impossible at 
 present. If a man is shut up in a house, the trans- 
 parency of the windows is an essential condition of his 
 seeing the sky. But it would not be prudent to infer 
 that, if he walked out of the house, he could not see 
 the sky because there was no longer any glass through 
 which he might see it. 
 
 With regard to the connexion of the brain with 
 thought, the chief evidence for it appears to be that 
 diseases or mutilations of the brain affect the course of 
 thought. But this does not prove that, even while 
 a man has a brain, his thoughts are directly connected 
 with it. Many things are capable of disturbing thought, 
 which are not essential to its existence. For example, 
 a sufficiently severe attack of toothache may render 
 all consecutive abstract thought impossible. But if 
 the tooth was extracted, I should still be able to think. 
 And, in the same way, the fact that an abnormal state 
 of the brain may affect our thoughts does not prove 
 that the normal states of the brain are necessary for 
 thought. 
 
106 SOME DOGMAS OF EELIGION 
 
 Even if the brain is essential to thought while we 
 have bodies, it would not follow that when we ceased 
 to have brains we could not think without them. The 
 same argument applies here as with the organs of 
 sense. It might be that the present inability of the 
 self to think except in connexion with the body was 
 a limitation which was imposed by the presence of the 
 body, and which vanished with it. 
 
 82. We have now considered the two arguments 
 against the immortality of the self which spring from 
 the death of the body. But we have said nothing as 
 to the bearing on this question of stories as to the 
 ghosts of the dead. Such stories, however numerous 
 and well authenticated, could never give us any positive 
 evidence that the self was undying. At the most 
 they could prove that it survived its body for a few 
 centuries. But indirectly the evidence could be of 
 considerable importance. For it might possibly prove 
 that the self survived the death of its body. Now 
 the death of its body is by far the strongest reason 
 that we have for doubting the self s immortality. And 
 if the appearances of ghosts could prove that this 
 reason had no weight, they would have removed the 
 greatest difficulty in the way of the belief- 
 Much of the evidence offered on this subject is 
 doubtless utterly untrustworthy. But there is a good 
 deal which investigation has failed to break down. 
 And there is much to be said in support of the view 
 that, after all deductions have been made for fraud, 
 error, and coincidence, there is still a sufficient resi- 
 duum to justify the belief that such apparitions are in 
 some cases due to the action of the dead man whose 
 body they represent. 
 
 But the mere proof that there was this causal con- 
 nexion between the dead man and the apparition 
 
HUMAN IMMOKTALITY 107 
 
 would not suffice to prove that the dead man had 
 survived his death. A chain of eflPects may exist long 
 after its original cause is destroyed. Chatham may 
 be one of the chief causes of the pride which England 
 excites in an Englishman to-day, but this proves 
 nothing as to Chatham's present existence. And, as 
 far as I know, all stories of apparitions would be 
 equally well explained by the theory that a man 
 might, before his death, initiate a chain of circum- 
 stances which would cause his apparition to appear, 
 after his death, under certain conditions, to men still 
 alive. In this case, nothing would be proved about 
 his existence after death. 
 
 This may appear improbable. But, on the other 
 hand, any attempt to prove empirically that man could 
 survive death would have to struggle with such an 
 enormous mass of negative evidence that its ante- 
 cedent improbability would also not be small. In- 
 vestigation may give us more evidence, and evidence 
 incompatible with any theory except that of suivival. 
 But at present it seems to me that we have much 
 more chance of proving our immortahty by meta- 
 physics than by psychical research. 
 
 83. We now come to the third question. Is there 
 any reason to suppose that my self does not share 
 the transitory character which I recognize in all the 
 material objects around me ? 
 
 What exactly is this transitory character? When 
 science says that a material object — a planet, or a 
 human body — ceases to exist, what does it mean ? It 
 does not mean that anjiihing is annihilated. It means 
 that units, which were combined in a certain way, are 
 now combined otherwise. The form has changed. 
 But everything which was there before is there now. 
 
108 SOME DOGMAS OF EELIGION 
 
 "We need not inquire whether this distinction be- 
 tween an unchanging matter and a changing form can 
 have more than a rough approximate correctness. It 
 is suflficient to note that the analogy of science — what- 
 ever weight may be attached to it — does not give us 
 reason to suppose anj^hing to be transitory except 
 combinations.^ 
 
 84. Is the self a combination? It certainly re- 
 sembles a combination in one respect, for it is dif- 
 ferentiated and contains a plurality. We can have 
 different sensations at the same moment, and sensations, 
 thoughts, and desires can exist simultaneously. But it 
 does not follow from this that a self is a combination. 
 For if a whole is a combination it is built up of parts 
 which could exist without being combined in that 
 way, while the combination could not exist without 
 them. If the bricks of a wall, for instance, were 
 destroyed, the wall would be destroyed too. But the 
 wall might be destroyed by being taken to pieces, 
 and the bricks would remain unchanged. 
 
 Do the parts of the self stand in this relation to it ? 
 Could my thoughts, my volitions, my emotions, exist 
 isolated, or in new combinations, when my self had 
 ceased to exist? It seems clear to me — the point is 
 too ultimate for discussion — that they cannot. It is 
 inconceivable that a thought, a sensation, a volition, 
 or an emotion should exist outside of a self. And it 
 is inconceivable that the same thought, sensation, 
 volition, or emotion which was once part of my mind 
 could ever be part of somebody else's. The self, we 
 
 ^ I do not mean to imply that science necessarily accepts any 
 units as indivisible and imperishable. My point is that it tells us 
 that whatever does perish does so only by the separation of the 
 parts of which it is composed. Those parts may themselves be 
 combinations. Thus it is possible that they may perish, and so 
 on ad infinitum* But nothing perishes but combinations. 
 
HUMAN IMMORTALITY 109 
 
 must say, is complex, but not a compound. It has 
 parts, but it is not built up out of them. For, while 
 it depends on them, they depend just as much on it. 
 
 The self, therefore, cannot cease by the separation 
 of its parts. For its parts only exist as united in it, 
 and therefore could not separate from it. If it did 
 cease to exist, it could only be by annihilation. It is 
 not only that the form would have changed, but that 
 the form and content alike would have perished. 
 
 Now there is no analogy in science to suggest the 
 probabihty of this. For science treats nothing as 
 perishable except combinations. This, indeed, does 
 not give us any safe analogy for the persistence of the 
 self. In the first place, there is reason to doubt the 
 absolute validity of the distinction between content 
 and form, which science finds it convenient to make. 
 And, in the second place, the difference between a self 
 and matter is too great for an analogy from one to the 
 other to be very conclusive. But at any rate science 
 gives no analogy against us. 
 
 85. All this still leaves us very far from a positive 
 assertion of immortality. Even though the death of 
 the body is no argument for the destruction of the 
 self, and the self cannot be decomposed into its parts, 
 it is still possible that the self should not be immortal. 
 And this view has been held in many systems of 
 idealism. It may be maintained, for example, that 
 finite individuals only exist to carry out some divine 
 purpose, and that it is possible that an individual may 
 cease to be necessary for such a purpose, and so cease 
 to exist. This was Lotze's view. Or again, it may 
 be maintained that there is something contradictory 
 in the idea of a self, which prevents us from regarding 
 it as an adequate expression of reality, and that there- 
 fore there is no reason to suppose that any particular 
 
110 SOME DOGMAS OF EELIGION 
 
 self shares the eternity which is characteristic of true 
 reality. 
 
 To meet such doubts as these it would be necessary 
 to construct a complete metaphysical system. We 
 should have to determine what was the general nature 
 of all reahty, and whether that nature involved the 
 existence of finite selves. And if in this way we 
 reached the conclusion that the existence of finite 
 selves was eternally necessary, the question would 
 arise whether each self was eternal, or whether, on 
 the other hand, there was an unending succession of 
 transitory selves. And, if the former alternative were 
 accepted, we should have to consider the relation 
 between eternity and immortality. All that I have 
 endeavoured to do here has been to show that the 
 more obvious arguments against immortality — those 
 which have most weight with most people — have no 
 validity* 
 
 In spite of all arguments, however, the idea that the 
 self cannot be immortal continually returns to us. 
 Reflection may drive it away, but in unreflective 
 moments it besets us again. We seem so small, and 
 the transitory seems so great. It is always hard — 
 there are times when it seems impossible — to be- 
 lieve that each of us can be a permanent element 
 in a universe in which nations and planets are but 
 momentary shapes. 
 
 And the belief in immortality seems all the more 
 incredible when we consider many of the believers. 
 Many people believe in it because they wish it to be 
 true, their desires blinding their judgements. Many 
 believe in it on the authority of some religion claiming 
 to be revealed — most of which must, on any hypothesis, 
 be untrustworthy. It is illogical to conclude that a 
 behef cannot be true because it has generally been 
 
HUMAN IMMORTALITY 111 
 
 believed for mistaken reasons, but it is difficult, in 
 practice, to keep our distrust from spreading from the 
 reasons to the belief. Yet I think that reasons for 
 the belief in immortality may be found of such strength 
 that they should prevail over all difficulties. 
 
CHAPTER IV 
 
 HUMAN PEE-EXISTENCE 
 
 86. In this chapter I wish to point out some reasons 
 for thinking that, if men are immortal, it is more prob- 
 able that the beginning of the present life, in which each 
 of us finds himself now, was not the beginning of his 
 whole existence, but that he lived before it, as he will 
 live after it. I wish, then, to consider the explanation 
 which this theory, if true, would afford of some of the 
 facts of our experience, and to consider what would 
 be the practical value of such immortality as it can 
 offer us. 
 
 The present attitude of most western thinkers to 
 the doctrine of pre-existence is curious. Of the many 
 who regard our hfe after the death of our bodies as 
 certain or probable, scarcely one regards our life before 
 the birth of those bodies as a possibility which deserves 
 discussion. 1 And yet it was taught by Buddha and 
 by Plato, and it is usually associated with the belief 
 in immortahty in the far east. Why should men who 
 are so anxious to-day to prove that we shall live after 
 this life is ended regard the hypothesis that we have 
 already survived the end of a life as one which is 
 beneath consideration ? 
 
 The explanation of this, I suppose, is that in modern 
 western thought the great support of the belief in 
 immortality has been the Christian religion. Under 
 
 ^ Lotze, for example, treats it as a serious objection to a 
 particular argument for immortality, that it would lead to the 
 ' strange and improbable ' conclusion of pre-existence. Meta^hysiCj 
 Section 245. 
 
HUMAN PKE-EXISTENCE 113 
 
 these circumstances a form of the belief which was 
 never supported by that religion was not hkely to be 
 considered of any importance. And, for some reason, 
 Christians have almost unanimously rejected those 
 theories which placed pre-existence by the side of 
 immortality, although there seems nothing in pre- 
 existence incompatible with any of the dogmas 
 which are generally accepted as fundamental to 
 Christianity. 
 
 87. The most effective way of proving that the 
 doctrine of pre-existence is bound up with the doctrine 
 of immortality would be to prove directly that the 
 nature of man was such that it involved a life both 
 before and after the present life. But, as I said at the 
 beginning of the last chapter, such a demonstration, if 
 it is possible at all, as I believe it to be, would be far 
 beyond the scope of this book, since it would involve 
 a determination of some of the most fundamental 
 characteristics of reahty. 
 
 I must content myself with stating in a more general 
 manner my grounds for believing that any evidence 
 which will prove immortahty will also prove pre- 
 existence. There are two ways in which a proof of 
 immortahty may be attempted. The first is the 
 directly metaphysical way. We may attempt to show 
 that the nature of man is such that he cannot cease to 
 exist while the universe continues to exist ; or that his 
 nature is eternal, and that an eternal nature cannot 
 have an end in time ; or pursue some similar Une of 
 argument. 
 
 In this case it seems to me that, if we succeed in 
 proving immortality, it will be by means of considera- 
 tions which would also prove pre-existence. I do not 
 see how existence in future time could be shown to 
 
114 SOME DOGMAS OF KELIGION 
 
 be necessary in the case of any being whose existence 
 in past time is admitted not to be necessary. If the 
 universe got on without me a hundred years ago, what 
 reason could be given for denying that it might get on 
 without me a hundred years hence ? Or if it is con- 
 sistent with my eternal nature that its temporal mani- 
 festation should begin at some point in time, coidd we 
 find any reason for supposing that the cessation of that 
 manifestation at some point in time would be incon- 
 sistent with that nature ? I do not see of what kind 
 such a reason could be, nor do I know of any attempt 
 that has been made to establish one. 
 
 88. There is another way in which attempts have 
 been made to prove immortality. This consists in 
 demonstrating that the universe is the work of a 
 benevolent creator, or has a purpose harmonious with 
 our ideals of morality, and then arguing that the 
 absence of immortality would be inconsistent with the 
 benevolence of such a creator, or with such a moral 
 purpose. Arguments of this type would prove immor- 
 tahty more readily than they could prove pre-existence. 
 No wrong can be done to the non-existent, and it could 
 hardly be made a reproach to the goodness of the uni- 
 verse that it had waited a long time before it produced 
 a particular person. But, once produced, any person 
 has certain moral claims, and if it could be shown that 
 his annihilation was inconsistent with those claims, 
 we could argue from the goodness of the universe to 
 the impossibility of his annihilation. 
 
 Can we, however, validly conclude from the good- 
 ness of the universe to the impossibihty of a particular 
 evil? It cannot be denied that some evil does exist. 
 The ultimate nature of reality, then, is not incompatible 
 with the existence of some evil. And when this is 
 once admitted, can we hope for an a priori proof that 
 
HUMAN PEE-EXISTENCE 115 
 
 any particular evil is too bad to be consistent with the 
 nature of the universe? It seems to me that we 
 cannot, and that we must therefore reject all argu- 
 ments which attempt to prove that a thing is unreal 
 because it would be evil. 
 
 We may call arguments of this sort ethical, since 
 they involve the conception of the good. Modern 
 demonstrations of immortality have almost always 
 been of this character, and are not purely metaphysical, 
 and this explains why it has often been held in modern 
 times that immortahty was proved, although pre- 
 existence has almost always been disbeheved. Even 
 the arguments of the eighteenth century, which were 
 attacked by Kant, had an ethical element in them. 
 Their supporters endeavoured, indeed, to prove by 
 purely metaphysical considerations that the nature of 
 man's spirit was such that it could not be destroyed 
 in the ordinary course of nature. But they held that 
 each man had been created by an act of the divine 
 will, and they admitted that a similar act could destroy 
 him. In order to show that God never would will 
 to destroy a man whom he had once created, they 
 either fell back on the asserted evidence of revelation, 
 or contended that such destruction would be incon- 
 sistent with what we knew of God's moral character, 
 in which case their argument had passed over into the 
 ethical class. 
 
 If, as I have maintained, ethical arguments of this 
 sort are invalid, we are forced back on the purely 
 metaphysical arguments, and here we seem unable to 
 treat the past and the future differently. My conclu- 
 sion is, then, that any demonstration of immoi-tality 
 is hkely to show that each of us exists tlirough all 
 time — past and future — whether time is held to be 
 finite or infinite, 
 
 I 2 
 
116 SOME DOGMAS OF EELIGION 
 
 89. We must now inquire what consequences would 
 follow from the truth of pre-existence and immortality. 
 Each man would have at least three lives, his present 
 life, one before it, and one after it. It seems more 
 probable, however, that this would not be all, and that 
 his existence before and after his present life would in 
 each case be divided into many lives, each bounded by 
 birth and death. This doctrine of a plurality of future 
 lives and of past lives may be conveniently referred to 
 as the doctrine of plurality of Hves.^ 
 
 90. There is much to be said for the view that a 
 plurality of lives would be the most probable alterna- 
 tive, even on a theory of immortahty which did not 
 include pre-existence. We do not know what is the 
 cause which produces the limitation of our present 
 lives by birth and death, but some cause there must 
 be, and a cause which produces so important an effect 
 is one which plays a great part in our existence, as 
 long as it continues to act. 
 
 If we accept immortality and reject a plurality of 
 lives — and this is the most common opinion, though 
 plurahty of lives is accepted more frequently than pre- 
 existence — we must hold that the causes, whatever 
 they are, which operate on each of us so as to cause 
 his death once, will never operate again on any of us 
 through all future time. This is, of course, not 
 impossible. The true nature of death may be such 
 that there is no need, and no possibility, of its repeti- 
 tion. But I do not see that we have any reason to 
 believe this to be even probable. 
 
 ^ In one sense, of course^ a belief in pre-existence and immor- 
 tality is itself a belief in a plurality of lives, since it is a belief in 
 three at least. But it will, I think, be more convenient to reserve 
 the name for the belief mentioned above — that for each of us 
 existence on either side of the present life would be divided into 
 more lives than one. 
 
HUMAN PEE-EXISTENCE 117 
 
 It is quite clear that a life wHcli stretched on 
 unendingly without death would in many respects be 
 enormously different from our present lives. An 
 attempt to imagine how our present lives would be 
 transformed if neither we ourselves, nor our fellow 
 men, had in future any chance of death, will make 
 this evident. A believer in immortaHty who denies, 
 or regards as improbable, the doctrine of the plurality 
 of lives, must assert, or regard as probable, that the 
 death which ends his present hfe for each of us will 
 change profoundly and permanently the conditions 
 of all future life. And for this there seems no 
 justification. 
 
 If we are immortal, the value of our existence 
 either remains permanently at about its present level, 
 or rises or falls after death. In the first case, we 
 should have no reason to suppose that it was so 
 changed that death would not recur. As I have said, 
 it is not impossible that it should be so. But when 
 anything has a particular characteristic, the presump- 
 tion is that, if that thing continues to exist, its 
 characteristic will not suddenly vary. The presump- 
 tion is certainly not strong, and it can give us no firm 
 behef. But it is, I think, suflScient to render it rather 
 more probable that the characteristic of periodic 
 mortality wiU not be left behind at the end of our 
 present lives. 
 
 91. I do not think that this would be very generally 
 denied. The denial of the pluraUty of lives is generally 
 based on the behef that oux lives do not remain at the 
 same level after death. It is not because men have 
 died once that it is held that they cannot die again. 
 It is because it is beheved that after death they are in 
 heaven or hell, the one much above the level of 
 earthly life, the other much below it. It is contended 
 
118 SOME DOGMAS OF RELIGION 
 
 that the change effected in this manner renders further 
 deaths improbable. This is especially maintained with 
 regard to heaven. 
 
 It might be admitted that a state of absolute perfec- 
 tion would render further death improbable. But 
 even the best men are not, when they die, in such 
 a state of intellectual and moral perfection as would 
 fit them to enter heaven immediately, if heaven is 
 taken as a state of absolute perfection which renders 
 all further improvement unnecessary and impossible. 
 This is generally recognized, and one of two alter- 
 natives is commonly adopted to meet it. The first is 
 that some tremendous improvement — an improvement 
 out of all proportion to any which can ever be observed 
 in life — takes place at the moment of death, at any 
 rate in the case of those who die under certain condi- 
 tions. For this, so far as I know, there are no argu- 
 ments. The other and more probable alternative is 
 that the process of gradual improvement can go on 
 in each of us after the death of our present bodies. 
 
 But if our existence immediately after our present 
 life is imperfect, and a state of improvement and 
 advance, it has not yet reached that absolute perfection 
 which might make future deaths improbable. And it 
 seems to me that the natural inference from this view 
 — though it is not drawn by the majority of those who 
 hold it — is that this life will be followed by others 
 like it, each separated from its predecessor and 
 successor by death and re-birth. For otherwise we 
 should be limited to the hypothesis that a process of 
 development, begun in a single life bounded by death, 
 should be continued as an indefinitely long life not 
 divided by birth and death at all. And to suppose, 
 without any reason, such a change from the order of 
 our present experience seems unjustifiable. 
 
HUMAN PEE-EXISTENCE 119 
 
 Should any persons be destined to attain a state of 
 great and permanent degradation, there would be still 
 less reason for supposing that this would exclude all 
 death from their future existence. Death may possibly 
 be incompatible with absolute perfection, but it has no 
 characteristic which can be suggested as incompatible 
 with the extreme of human degradation. In addition 
 to this we may urge, as in the case of heaven, that it 
 is unreasonable to suppose an extreme change at the 
 moment of death, and that, even if the completed 
 degradation was likely to exclude death, there could 
 be no reason for supposing that the process towards 
 it would do so from the first. 
 
 92. Again, processes begun in this life are some- 
 times finished in it, and sometimes left incomplete. 
 We continually find that death leaves a fault without 
 a retribution, a retribution without a repentance, a pre- 
 paration without an achievement, while in other cases, 
 where the life has lasted longer, a similar process is 
 complete between birth and death. If men survive 
 death, we must expect that these processes, when 
 not worked out before death, will be worked out in 
 a future hfe. And if the content of our existence after 
 death has so much similarity, in essential features, 
 with the content of our present lives, the presumption 
 is increased that they have not changed so far as to 
 have shaken off the necessity of periodical death.^ 
 
 93. There seems, therefore, good reason for regard- 
 ing plurality of lives as the least improbable alter- 
 native, even if we accept immortality without accepting 
 pre-existence. But if pre-existence is also accepted, 
 the case for a plurahty of lives becomes stronger. For 
 then the death which alters my present life is no longer 
 an unique event in my existence. One hfe, if no more, 
 
 ^ On this subject we may refer to Browning's Evelyn Hope, 
 
120 SOME DOGMAS OF KELIGION 
 
 came to an end for me before my present life could 
 begin. Thus any theory would be false which should 
 try to reject the plurality of hves on the ground that 
 it was probable that death could only occur once in 
 a man's existence. And the pluraUty of lives could 
 only be regarded as improbable, if there was reason 
 to suppose that an event, which happened twice in 
 a man's existence, would never happen a third 
 time. Now while it might be contended — though, as 
 I have said, I do not think it could be rightly con- 
 tended — that there were features about death which 
 made it probable it would only occur once in a man's 
 existence, it is difficult to see the slightest ground for 
 the suggestion that there is anything about death 
 which should make it improbable that it should occur 
 three times, although it was known that it occurred 
 twice. We can only accept immortality and pre- 
 existence, wliile rejecting the plurahty of lives, if we 
 hold that the causes which break off a life by a death, 
 after remaining dormant from the beginning of our 
 existence, act twice within an interval of from five 
 minutes to about a hundred years, and then never act 
 again through all future time. 
 
 The result seems to be that, even granting that pre- 
 existence is certain, there can be no absolute demon- 
 stration of pluraUty of lives, but that the plurality of 
 lives is the more probable supposition in any case, 
 and is still more probable on the hypothesis of pre- 
 existence. 
 
 94. There are various features of our present life 
 which can be explained more satisfactorily on the 
 theory of pre-existence than on any other. I do not, 
 however, wish to suggest that the ease of explaining 
 them on this hypothesis, or the difficulty of explaining 
 them without it, is so great as to form any proof of 
 
HUMAN PEE-EXISTENCE 121 
 
 the doctrine of pre-existence. That doctrine, I beheve, 
 can only be proved by metaphysical arguments of the 
 type mentioned at the beginning of this chapter. 
 
 The most important of these features is to be found 
 in personal relations. Two people who have seen but 
 little of each other are often drawn together by a force 
 equal to that which is generated in other cases by 
 years of mutual trust and mutual assistance. 
 
 The significance of this fact has been, I think, very 
 much underrated.^ It is generally explained, when 
 any attempt at explanation is made, by the capricious- 
 ness of sexual desire. This explanation is inadequate, 
 because the fact to be explained is found with as great 
 proportional frequency in friendships which have no 
 connexion with sexual desire. 
 
 On the theory of pre-existence such relations would 
 naturally be explained by the friendships of past lives. 
 The love wliich comes at first sight, and the love which 
 grows up through many years in this life, would be 
 referred to similar causes, whose similarity would 
 account for the similarity of the effects. Each would 
 have arisen through long intimacy, and the only dif- 
 ference between them would be that in one case the 
 intimacy had been suspended by death and re-birth. 
 
 95. Again, as a man grows up certain tendencies 
 and qualities make themselves manifest in him. They 
 cannot be entirely due to his environment, for they 
 
 ^ The same may be said of all facts connected with the love 
 of one particular human being for another. Philosophy and 
 theology, when they profess to take men's love seriously, generally 
 confine it either to a love for God, or to a passion for mankind as 
 a whole. It is rarely that the writings of a philosopher or a 
 theologian find anything in a young man's love for his sweetheart 
 except a mixture of sexual desire and folly, or anything in a young 
 man's love for his comrade except folly pure and simple. Hegel 
 is, I think, to be regarded as an exception. Possibly, also, the 
 writer of the first epistle of St. John. 
 
122 SOME DOGMAS OF EELIGION 
 
 are often very different in people whose environment 
 has been very similar. We call these the man's 
 natural character, and assume that he came into life 
 with it. Such tendencies and qualities, since they are 
 not due to anything which happens after birth may be 
 called innate, as far as the present life is concerned, 
 
 Now when we look at the natural characters of 
 men, we find that in many cases they possess qualities 
 strongly resembling those which, as we learn by direct 
 experience, can be produced in the course of a single 
 life. One man seems to start with an impotence to 
 resist some particular temptation which exactly re- 
 sembles the impotence which has been produced in 
 another man by continual yielding to the same 
 temptation. One man, again, has through life a calm 
 and serene virtue which another gains only by years 
 of strenuous effort. Others again have innate powers 
 of judging character, or of acting with decision in 
 emergencies, which give them, while yet inex- 
 perienced, advantages to which less fortunate men 
 attain, if they attain to them at all, only by the ex- 
 perience of years. Here then we have characteristics 
 which are born with us, and which closely resemble 
 characteristics which, in other cases, we know to be 
 due to the condensed results of experience. If we 
 hold the doctrine of pre-existence, we shall naturally 
 explain these also as being the condensed results of ex- 
 perience — in this case, of experience in an earlier life. 
 
 96. But, it may be said, can we not explain these 
 features of our life quite as well by means of the 
 theory of heredity, without accepting pre-existence? 
 In the case of personal relations, I do not see that 
 heredity would help us at all. I have admitted that 
 it is not impossible to explain the facts otherwise than 
 by pre-existence. The attraction may be simply due 
 
HUMAN PEE-EXISTENCE 123 
 
 to something in the character of each of the two 
 persons, though in many cases we cannot see what 
 that something could be. And then it is possible 
 that the element in question has been produced by 
 heredity. But there is nothing in heredity which 
 should make it likely that it should produce this result 
 rather than another, and so the abstract possibility that 
 the attraction is due to some undetected element in the 
 two characters is not increased by the suggestion that 
 the characters were produced by heredity. On the 
 theory of pre-existence, however, we can regard the 
 effects as produced by a course which would be likely 
 to produce this result rather than another — that is, by 
 relations formed in an earlier life. 
 
 Heredity, however, can produce a more satisfactory 
 explanation of innate aptitudes. My ancestors can- 
 not — if pre-existence is false — have loved my friend, 
 and therefore there is nothing in the fact that I inherit 
 from them that explains my loving him at first sight. 
 But my ancestors may have yielded to certain sins, 
 or resisted them, or practised certain activities, and 
 then, supposing that I can inherit the results which 
 they have acquired,^ there would be a reason why 
 I should have an innate strength or weakness in 
 certain directions, which closely resembled similar 
 characteristics which other men have acquired by 
 their own action in the course of their present lives. 
 
 We must, however, remember that such innate 
 dispositions often occur in cases where nothing of the 
 sort can be traced among the ancestors — even if, as 
 sometimes happens, the ancestors themselves can be 
 traced for many generations back. It is possible, no 
 doubt, that the acquirement of some more remote 
 
 ^ The possibility of this, however, is, to say the least, highly 
 uncertain. 
 
124 SOME DOGMAS OF EELIGION 
 
 ancestor may have remained dormant through the 
 intervening generations and have now re-appeared, 
 but the explanation is naturally much less probable 
 than it is in the cases where the ancestral acquirement 
 is known to have existed. 
 
 But, as I have said, while I regard the explanation 
 of these facts by pre-existence as better than any 
 which can be offered without it, I do not regard the 
 superiority of the explanation as sufficient to give by 
 itself any appreciable probability to pre-existence, 
 which, if established at all, must be established by 
 more directly metaphysical arguments. 
 
 97. Whether acquired qualities can be inherited or 
 not, there is no doubt that there is a certain tendency 
 for men — not merely their bodies, but themselves — 
 to resemble their ancestors. And it may be thought 
 that this would be an objection to our theory of pre- 
 existence. If a man's character is determined by his 
 previous lives, how can it be also determined by the 
 character of the ancestors by whose bodies his body 
 was generated? 
 
 There is, however, no real difficulty here. We may 
 have reason to believe that a man's character resembles 
 to some extent that of his ancestors, but it would be 
 impossible to demonstrate, and there is no reason to 
 beheve, that there are no elements in it which could 
 not be derived from that source. On the other hand, 
 the doctrine of pre-existence does not compel us to 
 deny all influence on a man's character of the characters 
 of his ancestors. The character which a man has at 
 any time is modified by any circumstances which 
 happen to him at that time, and may well be modified 
 by the fact that his re-birth is in a body descended 
 from ancestors of a particular character. 
 
 98. Thus the two ways in which the character in 
 
HUMAN PEE-EXISTENCE 125 
 
 this life is said to be determined need not be incon- 
 sistent, since they can both co-operate in the deter- 
 mination, the tendencies inherited with the body 
 modifying the character as it was left at the end of 
 the previous life. But there is no impossibility in 
 supposing that the characteristics in which we resemble 
 the ancestors of our bodies, may be to some degree 
 characteristics due to our previous lives. In walking 
 through the streets of London, it is extremely rare to 
 meet a man whose hat shows no sort of adaptation to 
 his head. Hats in general fit their wearers with far 
 greater accuracy than they would if each man's hat 
 were assigned to him by lot. And yet there is very 
 seldom any causal connexion between the shape of the 
 head and the shape of the hat. A man's head is never 
 made to fit his hat, and, in the great majority of cases, 
 his hat is not made to fit his head. The adaptation 
 comes about by each man selecting, from hats made 
 without any special reference to his particular head, 
 the hat which will suit his particular head best. 
 
 This may help us to see that it would be possible to 
 hold that a man whose nature had certain character- 
 istics when he was about to be re-born, would be re- 
 born in a body descended from ancestors of a similar 
 character. His character when re-bom would, in this 
 case, be decided, as far as the points in question went, 
 by his character in his previous life, and not by the 
 character of the ancestors of his new body. But it 
 would be the character of the ancestors of the new 
 body, and its similarity to his character, which deter- 
 mined the fact that he was re-born in that body rather 
 than another. The shape of the head— to go back to 
 our analogy— does not determine the shape of the hat, 
 but it does determine the selection of this particular 
 hat for this particular head. 
 
126 SOME DOGMAS OF EELIGION 
 
 But how, it may be asked, would each person, in 
 this case, be brought into connexion with the new 
 body that is most appropriate to him ? I do not see 
 any difficulty here. We know that various substances 
 which have chemical affinities for one another will meet 
 and combine, separating themselves, to do so, from 
 other substances with which they have been previously 
 in connexion. And we do not find anything so st]"ange 
 or paradoxical in this result as to make us unwilling 
 to recognize its truth. There seems to me to be 
 nothing more strange or paradoxical in the suggestion 
 that each person enters into connexion with the body 
 which is most fitted to be connected with him. 
 
 99. And, if there were any difficulty in this suppo- 
 sition, it is a difficulty which would be just as serious 
 for the theory adopted by most behevers in immortality 
 who reject pre-existence. If no man existed before 
 the formation of his present body, the question still 
 arises — how did he become connected with a body such 
 that his character resembles the characters of the 
 ancestors of that body ? The question would not arise 
 if we supposed that the whole character of the self 
 was simply produced by the body. But this is not 
 the ordinary view. Indeed, it would be difficult to 
 hold this without also holding that the self, as well as 
 its character, was produced by the body. And such 
 a view as this would be all but incompatible — if not 
 quite incompatible — with the belief in immortality. 
 
 Again, the question of how the connexion is deter- 
 mined, might be considered to have been answered if 
 it were held that the parents created the new person 
 at the time that they generated the body. I wiU not 
 discuss the difficulties which, as it seems to me, are 
 involved in this view, since I am dealing with the 
 consequences of pre-existence, and not with the 
 
HUMAN PEE-EXISTENCE 127 
 
 theories which deny it. It is sufficient to remark here 
 that this is not the view most generally adopted. 
 The common belief is that the person is created, not 
 by the parents of his body, but directly by the supreme 
 power of the universe. 
 
 And then the question still remains — how does this 
 person get into connexion with the appropriate body, 
 since they come into existence independently? It 
 seems that there are only two alternatives. It may 
 be said that the connexion is due to a special act of 
 divine providence in each case. But, if it is legitimate 
 to invoke such a special act at all, it is surely just as 
 legitimate to invoke it to make a connexion for a pre- 
 viously existing person as for a newly created person. 
 Or else it may be said that the appropriate connexion 
 is brought about by some general law. And there can 
 be no greater difficulty in supposing such a law to act 
 on persons who had previously existed than in sup- 
 posing it to act on persons newly created. The diffi- 
 culty, therefore, if there is one, is no greater for those 
 who accept pre-existence than for those who deny it, 
 
 100. Is the truth of pre-existence desirable? How 
 much would an immortality be worth to us which 
 was coupled with pre-existence? The most serious 
 objection relates to memory. We do not now re- 
 member anything of any previous life. If, neverthe- 
 less, we have lived previously, and have forgotten it, 
 there seems no reason to expect that we shaU remember 
 our present life during subsequent lives. Now an 
 existence that is cut up into separate hves, in none of 
 which memory extends to a previous life, may be 
 thought to have no practical value. We might as 
 weU be mortal, it has been said, as be immortal with- 
 out a memory beyond the present life. The question 
 
128 SOME DOGMAS OF RELIGION 
 
 becomes more serious if not only pre-existence, but 
 also the plurality of lives is true. For then it might 
 reasonably be feared that we might lose memory, not 
 only twice in our existence, but an indefinitely large 
 number of times. 
 
 Sometimes, indeed, it has been asserted that such 
 a state would not be immortality at aU. Without 
 memory of my present life, it is said, my future life 
 would not be mine. If memory ceases at the death of 
 my body, I cease with it, and I am not immortal. 
 
 If each life had no continuity with its successors, 
 and no effect on them, then indeed there would be 
 httle meaning, if any, in calling them lives of the same 
 person. But we cannot suppose that this could be the 
 case. If the same self passes through hves, any change 
 which happens to it at any time must affect its state in 
 the time immediately subsequent, and, through this, in 
 all future time. Death and re-birth, no doubt, are of 
 sufficient importance to modify a character consider- 
 ably, but they could only act on what was already 
 present, and the nature with which each individual 
 starts in any life would be moulded by his experiences 
 and actions in the past. And this is sufficient to make 
 the identity between the different lives real. 
 
 101. It has also been objected that the re-birth of 
 a person without a memory of his previous life, would 
 be exactly equivalent to the annihilation of that person 
 and the creation of a new person of exactly similar 
 character. (By this it is not meant that the new 
 person would be exactly similar to the old one at the 
 moment of the latter's annihilation, but that he would 
 be exactly similar to what the old person would have 
 been if he had undergone the process of re-birth.) 
 Now, it is argued, I should not regard myself as 
 immortal if I knew that I was to be annihilated at 
 
HUMAN PEE-EXISTENCE 129 
 
 death, even if I knew that an exactly similar person 
 would then be created. And therefore, it is concluded, 
 re-birth without memory cannot be considered as real 
 immortality of the self. 
 
 But the objection supposes an impossibility. There 
 could not be another self which would have a character 
 exactly similar to what mine would have been under 
 exactly similar circumstances. The self is not a ' thing 
 in itself', whose nature is independent of its qualities. 
 The self is a substance with attributes, and it has no 
 nature except to express itself in its attributes. If the 
 character of the new self, under certain circumstances, 
 were exactly similar to my character under the same 
 circumstances, its attributes would be exactly similar 
 to my attributes. Then the substance also would be 
 the same, and I should not be annihilated at all.^ 
 But if there were a new self, then the difference 
 between the selves must be expressed by some differ- 
 ence in the attributes. Then the new self would not 
 be of exactly similar character to what I should have 
 been under the same circumstances, and therefore the 
 creation of a new self would not be exactly equivalent 
 to my rebirth. Thus exact similarity of attributes is 
 always sufficient to prove personal identity, not because 
 it would be sufficient if the substance were different, 
 but because it proves that the substance is the same. 
 
 ^ It will be seen that I am assuming here that there cannot be 
 two different substances with exactly similar attributes. It does 
 not lie within the scope of this book to discuss this principle, 
 commonly known as the Identity of Indiscernibles. It is 
 sufBcient for our present purpose to remark that the principle 
 is accepted by most philosophers of the present day. And those 
 who deny it, and assert that things which are exactly similar 
 may yet be numerically different, would not hold that the 
 annihilation of one thing and the creation of another could be 
 exactly equivalent to the continuance of the first, even though 
 the second was exactly similar to the first. For, although exactly 
 similar, they would be numerically different. 
 
130 SOME DOGMAS OF EELIGION 
 
 102. We may say then that, in spite of the loss of 
 memory, it is the same person who Kves in the succes- 
 sive lives. But has such immortality as this any value 
 for the person who is immortal ? 
 
 I do not propose to discuss whether any immortality 
 has any value. Some people maintain that all human 
 existence is evil, however favourable the conditions. 
 Others regard existence as of such value that they 
 would be prepared to choose hell rather than annihila- 
 tion. Among those who differ less violently, some 
 regard the life of the average man on earth at present 
 as of positive value, while others will only regard it as 
 valuable if it is the necessary preparation for a better 
 life which is to follow. Such differences as to the value 
 of life must obviously produce great differences as to 
 the value of its unending prolongation. All that 
 I shall maintain here is that the loss of memory need 
 not render immortality valueless if it would not have 
 been valueless without the loss of memory. 
 
 If existence beyond the present life is not expected 
 to improve, and yet immortality is regarded as valu- 
 able, it must be because a hfe no better than this is 
 looked on as possessing value. Now it is certain that 
 in this life we remember no previous lives, whether 
 it be because we have forgotten them, or because there 
 have been none to remember. And if this hfe has 
 value without any memory beyond itself, why should 
 not future lives have value without memory beyond 
 themselves ? In that case a man wiU be better off for 
 his immortahty, since it will give him an unlimited 
 amount of valuable existence, instead of a hmited 
 amount. And a man who believed that he had this 
 immortality would have a more desirable expectation 
 of the future than if he did not believe it. If, indeed, 
 a man should say that he takes no more interest in his 
 
HUMAN PEE-EXISTENCE 131 
 
 own fate, after memory of his present life had gone, 
 than he would take in the fate of some unknown per- 
 son, I do not see how he could be shown to be in the 
 wrong. But I do not beheve that most men would 
 agree with him, and to most men, therefore, the pros- 
 pect of a continuance of valuable existence, even with 
 the periodical loss of memory, would still seem to be 
 desirable. 
 
 103. But desire for immortality is not only, or 
 chiefly, because it will give us more life like our 
 present life. Its attraction is chiefly for those people 
 who beheve that the future life will be, at any rate for 
 many of us, a great improvement on the present. 
 Heaven is longed for, not merely because it will be 
 unending, but because it will be heaven. 
 
 Now it might be said that our chief ground for 
 hoping for a progressive improvement after death 
 would be destroyed if memory periodically ceased. 
 Death, it might be argued, would not only remove us 
 from the field of our activity, but would deprive us of 
 all memory of what we had done, and therefore what- 
 ever was gained in one life would be lost at death. 
 We could no more hope for a permanent improvement 
 than a man on the treadmill can hope to end higher 
 than he started. 
 
 We are not discussing the chance of future progress, 
 but only the relative chance of such progress if memory 
 ceases at death. We must ask, therefore, what 
 elements of value are carried on by memory from the 
 present to the future. And then we must consider 
 whether they can be carried on without memory. 
 
 I think I shall be in agreement with most people 
 when I say that memory is chiefly of value in our lives 
 in three ways. In the first place, it may make us 
 wiser. The events which we have seen, and the 
 
 K 2 
 
132 SOME DOGMAS OF EELIGION 
 
 conclusions at which we have arrived, may be pre- 
 served in memory, and so add to our present know- 
 ledge. In the second place, it may make us more 
 virtuous. The memory of a temptation, whether it 
 has been resisted or successful, may under various 
 circumstances help us in resisting present temptation. 
 In the third place, it may tell us that people with 
 whom we are now related are the people whom we 
 have loved in the past, and this may enter as an 
 element into our present love of them. 
 
 The value of memory, then, is that by its means the 
 past may serve the wisdom, the virtue, and the love of 
 the present. If the past could help the present in 
 a Uke manner without the aid of memory, the absence 
 of memory need not destroy the chance of an improve- 
 ment spreading over many lives, 
 
 104. Let us consider wisdom first. Can we be 
 wiser by reason of something which we have for- 
 gotten ? Unquestionably we can. Wisdom is not 
 merely, or chiefly, amassed facts, or even recorded 
 judgements. It depends primarily on a mind qualified 
 to deal with facts, and to form judgements. Now the 
 acquisition of knowledge and experience, if wisely 
 conducted, may strengthen the mind. Of that we 
 have sufficient evidence in this life. And so a man 
 who dies after acquiring knowledge — and aU men 
 acquire some — might enter his new life, deprived 
 indeed of his knowledge, but not deprived of the 
 increased strength and delicacy of mind which he had 
 gained in acquiring the knowledge. And, if so, he 
 will be wiser in the second life because of what has 
 happened in the first. 
 
 Of course he loses something in losing the actual 
 knowledge. But it is sufficient if he does not lose all. 
 Most progress is like the advance of a tide, whose 
 
HUMAN PRE-EXISTENCE 133 
 
 waves advance and retreat, but do not retreat as far as 
 they advanced. And is not even this loss reaUy a gain ? 
 For the mere accumulation of knowledge, if memory 
 never ceased, would soon become overwhelming, and 
 worse than useless. What better fate could we 
 wish for than to leave such accumulations behind 
 us, preserving their greatest value in the mental 
 faculties which have been strengthened by their 
 acquisition. 
 
 105. With virtue the point is perhaps clearer. For 
 the memory of moral experiences is of no value to 
 virtue except in so far as it helps to form the moral 
 character, and, if this is done, the loss of the memory 
 would be no loss to virtue. Now we cannot doubt 
 that a character may remain determined by an event 
 which has been forgotten. I have forgotten the 
 greater mmiber of the good and evil acts which I have 
 done in my present hfe. And yet each must have 
 left a trace on my character. And so a man may 
 carry over into his next life the dispositions and 
 tendencies which he has gained by the moral contests 
 of this life, and the value of those experiences will not 
 have been destroyed by the death which has destroyed 
 the memory of them, 
 
 106. There remains love. The problem here is 
 more important, if , as I beheve, it is in love, and in 
 nothing else, that we find not only the supreme value 
 of life, but also the supreme reality of hfe, and, indeed, 
 of the universe. The gain which the memory of the 
 past gives us here is that the memory of past love for 
 any person can strengthen our present love of him. 
 And this is what must be preserved, if the value of 
 past love is not to be lost. The knowledge we acquire, 
 and the efforts which we make, are directed to ends 
 not themselves. But love has no end but itself. If it 
 
134 SOME DOGMAS OF RELIGION 
 
 has gone, it helps us little that we keep anything 
 it has brought us. 
 
 But past hours of love are past, whether we re- 
 member them or not. Yet we do not count their value 
 to be lost, since their remembrance makes the love 
 of the present stronger and deeper. Now we know 
 that present love can be stronger and deeper because 
 of past love which we have forgotten. Much has been 
 forgotten in any friendship which has lasted for 
 several years within the limits of a single Ufe — many 
 confidences, many services, many hours of happiness 
 and sorrow. But they have not passed away without 
 leaving their mark on the present. They contribute, 
 though they are forgotten, to the present love which 
 is not forgotten. In the same way, if the whole 
 memory of the love of a Ufe is swept away at death, 
 its value is not lost if the same love is stronger in 
 a new hfe because of what passed before. 
 
 Thus what is won in one life may be preserved in 
 another, if the people who love in the first life love the 
 same people in the second, and if their love is greater 
 in the second because it was there in the first. Have 
 we any ground to hope that these two conditions will 
 befulfiUed? 
 
 107, Let us take the first. We shall, if my theory 
 is right, have many lives — perhaps many milhons of 
 lives, and perhaps an infinite number. Now if the 
 fact that I loved a person in this hfe gave me no 
 reason to suppose that I should love him in any 
 other, then the whole value of love would be as much 
 confined to a single life as if there were no immortality. 
 And in that case it might perhaps be said that the 
 value of life was equally confined, and that immor- 
 tality, though real, was worthless. 
 
 The chance of a love recurring in any future life, 
 
HUMAN PRE-EXISTENCE 135 
 
 must depend primarily on the conditions which deter- 
 mine where and how the lovers are born in the future 
 life. For if memory does not survive death, it will be 
 impossible for love to occur in any life in which 
 people do not meet. If the conditions which deter- 
 mine the circumstances of our birth, and through them 
 our juxtapositions throughout life, were themselves 
 determined by chance, or by some merely mechanical 
 external necessity, the probabihty of meeting our friends 
 in another life would be too small to be regarded. 
 
 108. This is a consideration of great importance, 
 but it does not affect the question of the comparative 
 value of immortality with or without loss of memory. 
 Let us take the more ordinary view according to which 
 our existence after this life will be one and unbroken, 
 with a possibility of remembering in it, not only the 
 events which occur in it, but also the events of this 
 life. If the course of that future life is determined by 
 chance, or by mechanical necessity, there will be no 
 reason for hoping that we shall meet beyond death the 
 people whom we have loved in this hfe. Nor would 
 there be any reason for hoping that the love thus 
 denied fruition would be able to remain unextinguished 
 through unlimited ages of separation and new activities. 
 Once admit events to be determined in this way, and 
 there is no comfort to be gathered from immortality, 
 whether with or without memory, either for love or 
 for any of our other interests. 
 
 If immortality is to give us an assurance or a hope 
 of progressive improvement, it can only be if we have 
 reason to believe that the interests of spirit are so 
 predominant a force in the universe that they will 
 find, in the long run, satisfaction in the universe. 
 And, in this case, the constitution of the universe 
 would be such that, whether with or without memory. 
 
136 SOME DOGMAS OF RELIGION 
 
 love would have its way. I will not here inquire 
 whether the ultimate significance of spirit is anything 
 except love. But it will scarcely be denied — least of 
 all by those who feel the difficulties which I am now 
 considering — that the significance of love for spirit is 
 very great. And, if this is so, then the emotional 
 relations which exist between people must be highly 
 significant of their real positions towards one another 
 in the scheme of the universe. 
 
 In other words, people who are joined by love 
 cannot be dependent for their proximity to each other 
 — and consequently for the possibility of their love — 
 on some chance or mechanical arrangement whose 
 recurrence we could have no reason to expect. Their 
 love is not the effect of proximity, but its cause. For 
 their love is the expression of the ultimate fact that 
 each of them is more closely connected with the other 
 than he is with people in general. And proximity in 
 a particular hfe, like everj^thing else, is the effect — or, 
 rather, the manifestation under particular circumstances 
 — of those relations which make up the eternal nature 
 of the universe. 
 
 If, therefore, love has joined two people in this life, 
 we have, on the assumption we have been discussing, 
 good reason for believing that their existences are 
 bound up with one another, not for one hfe only, but 
 for ever. This would not involve their meeting in 
 every life, any more than it would involve their 
 meeting every day of each life. Love can survive 
 occasional absences, and is often even stronger for 
 them. And the universe is on a large scale, which 
 might require long absences. What we are entitled 
 to believe is that, while time remains, their eternal 
 nearness must continually find its expression in 
 proximity in temporal life. 
 
HUMAN PEE-EXISTENCE 137 
 
 109. As for the second condition ; if friends are not 
 to be separated, then certainly the value of love in one 
 life need not perish because there is no memory of it 
 in the next. If by means of it we make our relations 
 stronger and finer, then they will be stronger and finer 
 at the next meeting. What more do we want ? The 
 past is not preserved separately in memory, but it 
 exists, concentrated and united, in the present. Death 
 is thus the most perfect example of the ' collapse into 
 immediacy ' — that mysterious phrase of Hegel's — where 
 all that was before a mass of hard-earned acquisitions 
 has been merged in the unity of a developed character. 
 If we still think that the past is lost, let us ask our- 
 selves, as I suggested before, whether we regard as 
 lost all those incidents in a friendship which, even 
 before death, are forgotten. 
 
 110. I do not deny that in each particxilar life the 
 prospect of the loss of memory at the end of it will 
 appear to some extent a loss and a breach of con- 
 tinuity. In losing memory we lose that in which we 
 have found great value. Arguments may convince 
 us — as I have said, I think that they ought to con- 
 vince us — that we do not lose all the value, or any of 
 the highest value, but only the comparatively worthless 
 form, a form which the lapse of years would change 
 to a positive evil. But no doubt we shall always have 
 a tendency to shrink from the loss of memory. Yet 
 I believe that, as we come to understand life better, 
 we shall shrink from such a loss less and less. 
 
 We may, I think, fairly conclude that the value of 
 immortaUty would not be lessened much, if at aU, 
 by pre-existence. For the loss of memory which pre- 
 existence renders probable, seems to me the only 
 ground on which it has been held to diminish the 
 value of inunortality. 
 
138 SOME DOGMAS OF RELIGION 
 
 111. Pre-existence, indeed, as we have seen, renders 
 more probable a plurality of future lives. And the 
 prospect of a great number of lives — perhaps an in- 
 finite number, though this is not a necessary part of 
 the theory^ — gives us the prospect of many dangers, 
 many conflicts, many griefs, in an indefinitely long 
 future. Death is not a haven of rest. It is a starting- 
 point for fresh labours. But if the trials are great, so 
 is the recompense. We miss much here by our own 
 folly, much by unfavourable circumstances. Above all 
 we miss much, because so many good things are in- 
 compatible. We cannot spend our youth both in the 
 study and in the saddle. We cannot gain the benefit 
 both of unbroken health and of bodily weakness, both 
 of riches and of poverty, both of comradeship and of 
 isolation, both of defiance and of obedience. We cannot 
 learn the lessons ahke of Galahad and of Tristram and of 
 Caradoc. And yet they are all so good to learn. Would 
 it not be worth much to be able to hope that what we 
 missed in one life might come to us in another ? And 
 would it not be worth much to be able to hope that 
 we might have a chance to succeed hereafter in the 
 tasks which we failed in here ? 
 
 It may be that the change, the struggle, and the 
 recurrence of death, are endless, or, again, it may be 
 that the process will eventually destroy itself,^ and 
 merge in a perfection which transcends all time and 
 change. Such an end may come, perhaps, but at any 
 rate it cannot be near. 
 
 But though the way is long, and perhaps endless, 
 it can be no more wearisome than a single life. For 
 with death we leave behind us memory, and old age, 
 and fatigue. And surely death acquires a new and 
 
 ^ *As a Grod self-slain on his own strange altar, 
 Death lies dead.' — Swinburne, A Forsaken Garden* 
 
HUMAN PKE-EXISTENCE 139 
 
 deeper significance when we regard it no longer as 
 a single and unexplained break in an unending life, 
 but as part of the continually recurring rhythm of 
 progress — as inevitable, as natural, and as benevolent 
 as sleep. We have only left youth behind us, as at 
 noon we have left the sunrise. They will both come 
 back, and they do not grow old. 
 
CHAPTER V 
 
 FREE WILL 
 
 112. The question whether a man is free in his 
 actions is ambiguous. We may distinguish four senses 
 in which it may be taken. In the first place, we may 
 say that a man is free to do anything which nothing 
 but his own nature prevents him from doing. In this 
 sense I am not free to draw a triangle with two right 
 angles, since this would be impossible whatever my 
 nature might be. Nor am I, in this sense, free to 
 save a man's life if I am tied at a distance from him 
 by a chain which I cannot break. For I should be 
 equally unable to do it, whatever my nature might 
 be. It is true that I could do it, if my body was 
 sufficiently strong to break the chain. But a man's 
 body is not generally held to be part of himself in 
 such a sense as to make its characteristics part of his 
 nature. Any one who made no distinction between 
 a man and his body would, I suppose, hold that I was 
 free, in this sense of freedom, to break the chain. 
 
 But if I do not save a man's life because I mistake 
 the nature of his illness, or because it is too much 
 trouble, then I am acting freely. For if my nature 
 were such as to give me greater discernment, or such 
 as to make me think more of human hfe and less of 
 my own trouble, I should have saved him. And I 
 should still, in this sense of freedom, be held to be 
 acting freely, however certain it was that my action 
 was absolutely determined by the character with which 
 
FEEE WILL 141 
 
 I was born, combined with the circumstances of my 
 hfe. I was not determined from without to be unable 
 to save the life, and internal determination does not, 
 in this sense of freedom, prevent me from acting 
 freely. 
 
 113. Freedom, when the word is used in this sense, 
 may be called freedom of self-determination. The use 
 has some importance from a metaphysical point of 
 view, but in ordinary life it is not often employed. 
 Much more common is the second use of the word, 
 according to which a man is free to do anything which 
 nothing but his own will prevents him from doing. 
 So far as a man is free in this sense he can direct his 
 life as he chooses. We may call this, then, freedom 
 of self-direction. 
 
 When the word is used in this sense, I should not 
 be called free to save a man when I fail to do so 
 because I mistake the character of his illness. This 
 mistake comes from my nature, but not from my will. 
 I may make it in spite of wilUng not to make it. But 
 if I do not save his life because it would be too much 
 trouble, I am acting freely in not saving it. I abstain 
 from saving it because I choose to avoid trouble rather 
 than save a life. If my choice — that is to say, my 
 will — had been different, I should have saved him. 
 
 Here, as with the freedom of self-determination, 
 I should not be held less free because my choice was 
 absolutely determined by my character and my past 
 history. I am free because I act as I choose, whether 
 the choice is completely determined or not. 
 
 114. In the third place a man is said to act freely 
 when he acts according to the ultimate ideal of his 
 nature. The implication here seems to be that free- 
 dom is essentially the absence of such limitation as is 
 felt to thwart and constrain the being who is limited, 
 
M2 SOME DOGMAS OF RELIGION 
 
 and that no person can be completely free from such 
 constraint except by attaining the ultimate ideal of 
 his nature. Thus it is often said that only the wise 
 or the good are free, although it would be admitted 
 that foUy or wickedness might issue from the character 
 of some men by the same inevitable determination 
 which produced in others wisdom and virtue. 
 
 This use is extended by analogy to beings which 
 are not believed to feel constraint. Thus we say of 
 a tree that it grows freely when it grows to its normal 
 size in the ground, and that it does not grow freely 
 when it is stunted by being planted in a pot And 
 yet it is equally the nature of the tree to be stunted 
 in the one case as not to be stunted in the other. 
 The unstunted growth is called free because it is 
 looked on as the ideal. 
 
 Virtue is certainly part of the ideal of man. And 
 therefore virtuous action will be called free, in this 
 sense, when wicked action is not. So, also, will a 
 reasonable action as against a foohsh action. For this 
 there is also another ground. We often regard reason 
 and conscience as more truly parts of our nature than 
 desires or passions, which we speak of as though they 
 were forces acting on us from outside. When I act 
 according to a passion it is said to master me. When 
 reason or conscience prevents me from doing so, I 
 am said to master my passion. Thus virtuous and 
 reasonable action will appear, by the suggestion of 
 such phrases, as more truly self-determined than other 
 action. 
 
 Freedom in this sense may be called freedom of 
 self-reahzation. It has considerable importance in 
 philosophy, but we shall not have much concern with 
 it in the present chapter. 
 
 115. Freedom, finally, is used in a fourth sense, 
 
FEEE WILL 143 
 
 which is the one which mainly concerns us here. In 
 this sense a man is free in any action, if his choice 
 of that action is not completely determined. The 
 supporters of this view do not, I conceive, maintain 
 that a man can ever act without a motive, nor do 
 they consider that the existence of a motive is in- 
 compatible with freedom. But if the motive com- 
 pletely determined the act — either because there was 
 no other motive, or because it was determined to be 
 more effective than any other — then the act would not 
 be freely done. It is essential for freedom that there 
 should be motives prompting to different courses, 
 between which the agent chooses. And it is essential 
 that this choice should not be determined. We may 
 call this freedom of indetermination. My object in 
 this chapter is to consider whether it exists. 
 
 Freedom of indetermination is commonly spoken of 
 as Free Will. This seems to be justified. If freedom 
 were defined otherwise, the proper question might be 
 *Am I free to act?' not 'Am I free to wiU?' But, 
 if freedom is to imply the absence of complete de- 
 termination, it can only be the will that is free. The 
 voluntary act is completely determined — in so far as 
 it is not determined by outside circumstances it is 
 determined by the volition on which it follows. No 
 indeterminist would deny this. It is only the volition 
 which is undetermined, and only the volition which 
 is free. When an indeterminist says, for example, 
 that a man has freely committed a murder, he means 
 that he was free in willing to do it. Indeed, in so 
 far as the wiU did not completely determine the act, 
 as when a bullet meant for a tiger kills a man, the 
 indeterminist would deny that the shooter had acted 
 freely in killing the man. 
 
 116. The law of Causality asserts that every event 
 
144 SOME DOGMAS OF RELIGION 
 
 is determined bj previous events in such a way that, 
 if the previous events are as they are, it is impossible 
 that the subsequent event should not be as it is. 
 If this was the only general principle valid as to 
 causality, however, we should not be able to accept 
 as valid any of the laws of science which deal with 
 causation. For these are all general laws, which 
 assert that whatever has a particular quality produces 
 an effect with a particular quahty. Thus it is said 
 that all alcohol, taken in large quantities, produces 
 intoxication. But claret differs from whisky, and the 
 effects produced by drinking them are not completely 
 alike, nor are the effects exactly the same with all 
 men. In order to have any warrant for such generali- 
 zations we need the additional principle that for any 
 quahty, B, in an effect, there is always a quahty A 
 in the cause, of such a nature that every other cause 
 which has the quahty A produces an effect having 
 the quality B. In other words, the knowledge that 
 exactly similar causes will produce exactly similar 
 effects has no practical utihty, since we could never 
 know two causes to be exactly similar, even if it were 
 possible that they should be so. What is required 
 is the knowledge that partially similar causes will 
 produce partially similar effects. 
 
 I do not propose to consider whether Causahty and 
 the Uniformity of Nature are valid of events other 
 than volitions. To deny that they had any validity 
 at all would involve almost complete scepticism, since 
 no expectation of any future event would have the 
 least justification, and all arguments for the existence 
 of anything not perceived at the moment would be 
 absolutely baseless. The indeterminist does not, as 
 a rule, deny that all events except volitions must be 
 completely determined by previous events. Indeed, 
 
FEEE WILL 145 
 
 all his arguments as to the goodness or badness of 
 particular volitions imply that such volitions will 
 result in consequences which will inevitably follow 
 from them, unless interfered with by fresh volitions. 
 He only maintains that vohtions are not subject to 
 the law of Causality in so far as to be themselves 
 completely determined. It is this view which I 
 propose to discuss in this chapter. 
 
 Some persons, no doubt, are indeterminists as to the 
 will, because they reject the law of Causality altogether, 
 and are indeterminists as to everything. But this 
 position has not sufficient influence on rehgious thought 
 in general to be of importance for our present purpose. 
 
 Determinists, on the other hand, maintain that our 
 volitions are as completely determined as all other 
 events. From this it is generally, and I think correctly, 
 held to follow that it would be ideally possible to de- 
 duce the whole of the future course of events from the 
 present state of reality — though, of course, a mind 
 enormously more powerful than ours would be required 
 to do it. 
 
 117. We have now to consider the arguments 
 advanced by indeterminists in favour of their con- 
 tention that volitions are not completely determined. 
 These arguments may be brought under five heads. 
 Firstly, it may be asserted that I have an immediate 
 certainty of the proposition that my will is free. 
 Secondly, that each volition is accompanied by a feel- 
 ing of freedom in volition, and that this cannot be 
 accounted for except on the supposition that the will 
 is free. Thirdly, that with those volitions which are 
 recognized by us as morally right or wrong is con- 
 nected a judgement of moral obligation, which cannot 
 be accounted for except on the supposition that the will 
 is free. Fourthly, that the freedom of the wiU must be 
 
146 SOME DOGMAS OF RELIGION 
 
 true, because its falsity would make all choice absurd. 
 Fifthly, that it must be true, because of the disastrous 
 consequences which would follow if it were not true. 
 
 With regard, first, to the persons who assert that 
 they have an immediate conviction that the will is 
 free, and that that conviction cannot be shaken by any 
 arguments, I have given reasons in the second 
 chapter for thinking that ^uch a certainty, while of 
 course decisive for the people who have it, has no 
 force for those who do not have it. And it is beyond 
 doubt that many people — including many beUevers in 
 free will — have no certainty of this sort. It seems, 
 indeed, not improbable that some, at least, of those 
 who assert that they have such an immediate certainty, 
 have made some mistake as to what it is of which they 
 are immediately certain. They may have an imme- 
 diate certainty that they can do (in certain cases) what 
 they will to do — which is by no means the same as an 
 immediate certainty that their will is not completely 
 determiiaed. Or they may have an immediate certainty 
 that moral laws are binding on them, without an 
 immediate certainty that this requires an undetermined 
 will for its explanation. But of course if a man has, 
 as some men may have, a certainty of freedom of 
 indetermination, which is neither based on arguments 
 nor capable of being shaken by them, all discussion is 
 at an end as far as he is concerned* No one has any 
 means to convince him, and he has no means to 
 convince anybody. 
 
 118. We come next to the argument from that feeling 
 of freedom which we experience in action. To prove 
 that we have this sense of freedom an appeal must be 
 made to introspection, and thus this argument is apt 
 to be confounded with the assertion — just now con- 
 
FEEE WILL 147 
 
 sidered— of an immediate certainty of the proposition 
 that the will is not completely determined. They are, 
 however, quite distinct. The position which we are 
 now discussing does not say that we have an imme- 
 diate belief, but that we have a feeling. It proceeds 
 to argue that the existence of this feeling cannot be 
 explained, except on the hypothesis that freedom of 
 indetermination really exists. And so it affords some 
 possibility of discussion between opponents, since 
 arguments can be met by arguments. 
 
 The fact to which it appeals will scarcely be denied. 
 When I make a false accusation or refrain from making 
 it, when I draw my chair nearer to the fire or leave it 
 where it was, I have a feeling of freedom in my deci- 
 sion which I do not have when I am thrown from 
 a horse, or when I abstain from leaping over St. Paul's. 
 Now, it is said, this feeUng of freedom is incompatible 
 with complete determination, and the only legitimate 
 inference is that the decision which is accompanied by 
 the feehng of freedom is one in which I am not 
 completely determined. 
 
 119, I believe that we have this feehng of freedom be- 
 cause we are, in fact, free in these cases. But 1 believe 
 that the freedom of self-direction is quite sufiicient to 
 explain the feeling, and that there is no necessity to 
 accept freedom in the sense of indetermination. 
 
 On any theory, determinist or indeterminist, it is 
 clear that I shall not make the accusation or draw up 
 my chair unless I choose to do so. I may, in one sense 
 of the word, do such an act unwillingly, because all 
 the possible alternatives may be for some reason or 
 the other repugnant to me. But whatever I do in 
 these cases will be what 1 will to do in preference to 
 the other alternatives which I recognize as being 
 open to me. And my sense of freedom is just in 
 
 L 2 
 
148 SOME DOGMAS OF RELIGION 
 
 proportion to the extent to which my action does 
 depend on my will. If I draw a figure on paper I feel 
 freedom in drawing a triangle and not a cncle, but 
 I do not feel free in abstaining from giving my triangle 
 two right angles. Now I could draw either a triangle 
 or a circle if I willed to do so, but no volition would 
 enable me to draw a triangle with two right angles. 
 
 My sense of freedom, then, is proportionate to the 
 extent to which my action is determined by my will. 
 And I maintain that it is quite accounted for by 
 the fact that the action is determined by the will, and 
 that there is no need to hold that the determining 
 volition is itself undetermined. The feeling of free- 
 dom which we experience is a feeling that constraint 
 is absent. And constraint is absent in all cases where 
 a man only acts because he wills to do so. How can 
 constraint enter except in the form of a discord be- 
 tween the will and the facts? And when the facts 
 depend upon the will there can be no such discord. 
 If I were thrown from my horse because I had willed 
 to be thrown from it, where would the discord be ? 
 
 120. But, it is sometimes said, if we examine our voli- 
 tions we perceive that they are not completely deter- 
 mined. If this is asserted to be an immediate con- 
 viction, we have gone back to a contention which has 
 been already considered. If it is asserted on the basis 
 of an examination of our voKtions, I do not see how 
 such a conclusion could ever be justified. We find, 
 indeed, that all our most careful examination fails to 
 show us the whole sum of conditions which are neces- 
 sary in order that a particular volition should be com- 
 pletely determined. But this by no means proves that 
 the whole sum of conditions is not there. For there 
 is another alternative — that we have not powers of 
 observation sufficient to discover them. And, except 
 
FREE WILL 149 
 
 in the case of the human will, this alternative is always 
 adopted. It is perfectly impossible for any one to ex- 
 plain why a particular drop of rain falls where it does 
 rather than half an inch away. Yet no one supposes 
 that this event is not completely determined. We all 
 conclude that the series of events is so compHcated, 
 so numerous, and so unfavourably placed for observa- 
 tion, that our intellects are not able to follow it. It is 
 clear, therefore, that we cannot logically come to a 
 different conclusion with regard to our volitions unless 
 we have already reason to think that the case for their 
 complete determination is less strong than the case for 
 the complete determination of the falhng raindrop. 
 And therefore such a conclusion must rest on a proof 
 of free will, and cannot itself be used as such a proof. 
 
 121. It is sometimes urged, as an argument against 
 the complete determination of vohtions that a man 
 can act as he believes to be right, although he would 
 get, and knows that he would get, far more happiness 
 by acting otherwise. This, I think, cannot be denied. 
 And this is sufficient to disprove what is called 
 Psychological Hedonism — that is, the doctrine that 
 a man must will to do that which he believes will 
 produce the greatest pleasure for himself. But this is 
 quite distinct from determinism. We may maintain 
 that a man can act from other motives than his own 
 happiness, and yet quite logically assert that it is 
 completely determined both what motives will be 
 present to him, and which he will eventually choose. 
 
 122. We now pass to the third head. With those 
 vohtions which are recognized by us as being morally 
 good or bad there is connected a judgement of moral 
 obligation. I pronounce that one thing ought to be 
 willed, and that another thing ought not to be willed. 
 
150 SOME DOGMAS OF EELIGION 
 
 And I approve of whomever wills the first or does not 
 will the second, while I condemn whomever wills the 
 second or does not will the first. ^ With this judge- 
 ment of moral obhgation the absence of complete 
 determination in the will is asserted to be connected. 
 
 123. But what is the precise connexion ? Is it the 
 existence or the validity of these judgements on which 
 the argument is based ? And with what is it said to 
 be incompatible? Is it the truth of complete determina- 
 tion or the belief in complete determination? It is 
 always, I believe, the validity of judgements of obliga- 
 tion which is asserted to be incompatible with the 
 truth of complete determination. 
 
 With regard to the other three possible combina- 
 tions, it would, in the first place, be inconsistent with 
 notorious facts to say that the belief in complete 
 determination was incompatible with the existence of 
 judgements of obligation. It is well known that many 
 people are, and have been, determinists, and that the 
 great majority of them, if not all, have made judge- 
 ments of obligation respecting their own conduct, and 
 that of others. This is decisive. For if it is replied 
 that they ought not, from their own standpoint, to have 
 made such judgements, the contention is changed, 
 since it is no longer the existence of the judgements 
 which is denied, but their consistency. 
 
 In the second place, it cannot be seriously main- 
 tained that the belief in complete determination is 
 incompatible with the validity of judgements of obliga- 
 
 ^ The judgement of moral obligation is often spoken of as if it 
 referred directly to actions. But it seems to me to refer only to 
 volitions. For I do not blame myself for failing to carry out the 
 most imperative duty^ if my will to do so has been thwarted 
 by external forces. And if external forces have hindered the 
 realization of my will to do wrong I do not regard myself as 
 innocent. 
 
FREE WILL 151 
 
 tlon. In that case, if A, being a determinist, says 
 that Nero ought not to have burnt Eome, the judge- 
 ment is false, while if B, who is an indeterminist, 
 makes the same statement at the same minute, it is 
 true. This is clearly absurd. Or if I say this niorning, 
 being a determinist, that I ought not to have lost n^y 
 temper last Saturday, it is false. But if I am conr- 
 verted to indeterminism this afternoon, and then repeat 
 the same statement about the past, it will have changed 
 from false to true. 
 
 In the third place, the statement that the existence 
 of judgements of obligation is incompatible with the 
 truth of complete determination is never, so far as 
 I know, maintained, except on the ground that the 
 truth of complete determination would make them 
 invalid, and that, somehow or other, their existence 
 proves that they are not invahd. Nor do I see how it 
 could be maintained on any other ground. But if it 
 is maintained on this ground it rests, of course, on 
 the proposition that the validity of judgements of 
 obligation is incompatible with the truth of complete 
 determination. 
 
 And it is in this form — that the validity of the 
 judgements is incompatible with the truth of the deter- 
 mination — that the argument, as I said above, is 
 generally put forward, or intended to be put forward. 
 It is, however, often stated inaccurately, so as to pro- 
 duce confusion between the validity and the existence 
 of the judgements, and between the truth of the 
 determination and the belief in it. 
 
 The argument from the judgement of obligation is 
 perhaps the most usual argument for free will. It 
 seems to me that it is also the strongest, though 
 I cannot regard it as satisfactory. 'If the will is 
 completely determined, judgements of obhgation 
 
152 SOME DOGMAS OF EELIGION 
 
 cannot be valid.' The first question which this 
 suggests is the question whether judgements of 
 obhgation are vahd. 
 
 I do not think that we need trouble to inquire how, 
 if at all, the vaUdity of judgements of obligation could 
 be proved to any one who denied it — should such 
 a person be found. Determinism has always, as a 
 matter of fact, been defended, and, as I believe, can 
 be successfully defended, on the basis that judgements 
 of obligation are valid. 
 
 Let us, then, admit their validity. This does not, 
 of course, mean that no such judgement is ever mis- 
 taken, but that there are possible judgements of 
 obligation which would be valid when made, even 
 if we have not as yet succeeded in our efforts to find 
 them. Is this compatible with the truth of complete 
 determination ? 
 
 124. If the truth of complete determination were 
 incompatible with the truth of any proposition logically 
 presupposed in judgements of obligation, it is clear 
 that it would be incompatible with the validity of 
 those judgements, since they cannot be valid unless 
 their presuppositions are true. Every judgement of 
 obligation seems to me to have two such presupposi- 
 tions, (a) Something is such that its existence would 
 be good or bad.^ (&) The person as to whom the 
 judgement of obligation is passed can exercise, by his 
 will, some effect in determining the existence or non- 
 existence of that thing. No one would say that a man 
 ought to will the existence of anything unless the 
 thing willed was judged to be such that its existence 
 
 ^ When I use good and bad without any qualification, I do 
 not mean only moral good and bad, but good and bad in the 
 widest sense— that in which it may be said that happiness and 
 beauty are good. I use virtuous and wicked as synonyms of 
 moral good and bad, when the objects spoken of are volitions. 
 
FEEE WILL 153 
 
 would be good. And, again, no man can will anything 
 (though he may desire it) if he knows that he cannot 
 possibly have any influence on the matter. I should 
 not will that an eruption of Vesuvius should cease, 
 though under certain circumstances I might desire it 
 most passionately. And no sane man would say that 
 I ought to will it, or blame me for not doing so. But 
 if I were a magician, with powers so great that they 
 might possibly stop an eruption, I might then will to 
 stop it, and might possibly be morally bound to will it. 
 
 125. Would either of these presuppositions be 
 necessarily false if complete determination were true ? 
 I cannot see that either of them would. Would the 
 existence of anything cease to be good or bad because 
 it was completely determined whether I should will 
 its existence or not ? Would my own possession of 
 knowledge, or the satisfaction of my own hunger, or 
 the reUef of the distress of others, cease to be good 
 because it was absolutely certain that I should will to 
 bring them about, or because it was absolutely certain 
 that I should not will to bring them about ? Surely 
 this cannot be maintained. 
 
 As to the second presupposition, it is clear that 
 the complete determination of my will can make no 
 difference to the question of the effect of my will on 
 the result contemplated. Whether my will is com- 
 pletely determined or not, it is clear that I shall not 
 learn classical Greek or satisfy my hunger unless I 
 will to do so, while it is not improbable that I shaU do 
 both if I will to do them. Again, if I will to relieve 
 the distress of others, it is at least possible that some 
 distress will be relieved which would not have been 
 relieved otherwise, and this is not in the least affected 
 by the question whether my will is inevitably deter- 
 mined to take the course which it does take. 
 
164 SOME DOGMAS OF RELIGION 
 
 126. We pass from the presuppositions of the 
 judgement to the judgement itself. Is a volition to 
 produce a good result, or the man who makes it, to 
 be less approved, is the volition to produce a bad 
 result, or the man who makes it, to be less con- 
 demned, because the volition is completely deter- 
 mined? It does not seem to me that this should 
 make any difference to the approval or condemnation. 
 We approve or condemn whatever tends to produce 
 good or evil results, without further consideration. 
 
 If I do not save a man's life because he died before 
 I was born, I do not condemn myself for not saving 
 him, since it is not my individual nature but the 
 general nature of reality which prevents me from 
 altering the past. If I do not save him, because I 
 am tied with a chain I cannot break, do I condemn 
 myself? If I had a stronger body, I could break the 
 chain, and this would be a good thing to do. I shall 
 therefore condemn the nature of my body for being 
 unable to do it. If I regard the body as a part of 
 myself, I shall condemn myself for this bodily im- 
 perfection. If I only regard it as an external reality 
 with which I am in close connexion, I shall condemn 
 the body and not myself. ^ 
 
 If I fail to save a man's life because I mistake the 
 nature of his iUness, and so treat him in the wrong 
 manner, this is a purely intellectual mistake on my 
 part, unless my ignorance is due to past or present 
 misconduct. Postponing this latter possibility, I shall 
 
 ' The intensity of the condemnation will, of course, vary 
 according to the standard attained by similar bodies. We should 
 think very badly of a man's body which was so weak that it 
 could not break sewing-cotton. But if the chain that defied 
 my efforts were the cable of a battle-ship, my condemnation would 
 be no more than the recognition that I should have approved of 
 a strength which is absent from my body, and from the bodies of 
 all other men. 
 
FEEE WILL 155 
 
 certainly condemn myself for the failure. For the 
 failure is due to my want of knowledge, or of acute- 
 ness. These are qualities in my mind which tend to 
 produce evil, and as my mind is certainly myself, 
 whatever my body is, I shall condemn myself. I 
 shall pronounce myself a worse person than I might 
 have been. But I shall not condemn myself morally. 
 
 But now suppose that I do not will to save him, 
 because I should be enriched by his death. Or suppose 
 that I do will to save him, but that my efforts are 
 frustrated by ignorance due to past indolence, or to 
 confusion caused by intoxication. In these cases I 
 shall again condemn myself. And in these cases the 
 condemnation will be — as it was not before — moral 
 condemnation. For the result does in these cases — 
 as it did not before — depend upon my will. ^ If I had 
 willed differently in the past, I should not now be 
 ignorant or a drunkard. If I had willed differently 
 at the moment, I should not have preferred my own 
 wealth to the life of another. 
 
 I do not think that it would be denied by in- 
 determinists that, even if the will were completely 
 determined, moral condemnation of this sort would 
 be possible. The will would still be, as the body and 
 intellect had been in the previous cases, something 
 which tended to produce a bad result. And there 
 could be no reason why it should not be condemned, 
 as they, although regarded as completely determined, 
 were condemned in the previous cases. 
 
 127. The indeterminist, I conceive, would say that 
 while the complete determination of the will would 
 not destroy the validity of all approval or condemna- 
 tion of volitions, yet it would destroy the validity of 
 
 ^ I do not say that all moral qualities are qualities of volition, 
 but that all volitions have moral qualities. 
 
156 SOME DOGMAS OP RELIGION 
 
 the particular variety of approval or condemnation 
 which is found in judgements of obhgation. 
 
 Judgements of obligation are, of course, different 
 from other judgements of approval or condemnation, 
 or else they could not be distinguished as a class. 
 But they are distinguished as a class by the fact 
 that they are judgements which approve or condemn 
 volitions. And we have seen that the complete de- 
 termination of vohtions could not destroy the validity 
 of all approval or condemnation of them. It must be 
 some other characteristic of judgements of obligation, 
 and not their reference to volitions, which is incom- 
 patible with the complete determination of the volitions 
 to which they refer. 
 
 Two such characteristics have been suggested. The 
 first is the supreme value of right volition, which, as 
 it is said, is affirmed by our judgements of obligation. 
 The second is the sense of responsibility which follows 
 on those judgements. 
 
 128. It is said that we approve right volition more 
 than any other excellence, and that we condemn wrong 
 voUtion more than any other defect. Some reason, 
 it is said, is required for this fact, and the reason is 
 found in the incomplete determination of the will. 
 There was nothing which made it certain beforehand 
 that we should will rightly or wrongly. And this 
 makes right volition more precious and wrong volition 
 more detestable. 
 
 I can see no reason whatever why the moral quality 
 of an act should be regarded as intensified because it 
 happened without complete determination. It seems 
 to me that the moral quality of the act would be just 
 the same, while that of the agent would vanish. The 
 latter point will be dealt with later. ^ But however 
 ^ Cp. Sections 146-8. 
 
FREE WILL 157 
 
 this may be, it would be impossible to pi'ove Free 
 Will in this way, for two reasons. Firstly, if the 
 alleged fact were true, it would admit of another 
 explanation. Secondly, there is reason to believe that 
 the alleged fact is not true. 
 
 Our judgement of the value of excellences which 
 are not excellences of volition is different in different 
 cases. We regard the intellectual excellence of 
 Shakespeare with more approval than the excellence 
 shown by the most brilliant punster. Each of them 
 has excelled all other men in a particular direction, 
 but we admire Shakespeare most, because we regard 
 excellence in his direction as more important, in the 
 general scale of values, than excellence in punning. 
 Yet it would be universally admitted that Shake- 
 speare's genius, on the one hand, and the absence of 
 equal genius in myself, on the other hand, were facts 
 completely determined. They do not depend on voh- 
 tion, and it is only in volition that the indeterminist 
 denies complete determination. 
 
 If excellences which are admitted to be completely 
 determined can be judged to have different values, so 
 that one is placed above another, then the fact that one 
 excellence is placed above all others is quite compatible 
 with its determination. It would be quite adequately 
 explained by our judgement that the presence of this 
 excellence, however certainly determined, was better, 
 and its absence, however certainly determined, was 
 worse, than the presence or absence of any other ex- 
 cellence. 
 
 129. And, again, is it the fact that right volition is 
 always placed above all other excellences ? I think 
 that few people would be prepared to assert this. A 
 man who gives water to a thirsty dog has willed 
 rightly. If that man were Shakespeare, or Newton, 
 
158 SOME DOGMAS OF RELIGION 
 
 or Kant, should we be prepared to say that that 
 volition had more value than anything in his nature 
 except some other volition ? Surely most people would 
 regard the intellect which was capable of producing 
 Hamlet, or the Principia, or the three Critiques, as of 
 greater value. 
 
 Can we even say that the most important right 
 vohtions are approved more than any other excellence, 
 and the most important wrong volitions are condemned 
 more than any other defect ? I doubt if we can say 
 even this. It might be possible to maintain another 
 proposition which is sometimes confused with the 
 former — namely that the greatest moral excellence is 
 approved more than any other excellence, and that the 
 greatest moral defects are condemned more than any 
 other defects. But the two propositions are very 
 diffeJrent. 
 
 The argument, it will be remembered, rests on the 
 assertion that we can place nothing higher than right 
 volitions, from which it is argued that they must be 
 undetermined. Now, if we must place the gresLtest 
 moral excellence highest, it is clear that the argument 
 breaks down unless the greatest moral excellence 
 consists in right voHtion. 
 
 The simplest way of proving this would be to show 
 that all moral excellence was right volition. But 
 what, in this case, are we to say of a loving disposition, 
 a fervent patriotism, or a passion for humanity? They 
 are not volitions, or tendencies to vohtions, or habits of 
 volition. Nor can they be obtained by willing. (They 
 must, of course, be distinguished from resolutions to act 
 in particular ways. A man's will can cause him to act 
 as he would act if he loved his wife, or his country, 
 or mankind. But it cannot make him love them.) 
 
 Love and patriotism, then, are qualities which, by 
 
FEEE WILL 159 
 
 the indeterminists' own position, are as completely- 
 determined as artistic or literary excellence. Will 
 indeterminists be prepared to say that, while justice 
 and beneficence are moral excellences, love and 
 patriotism are mere gifts of fortune, and have no 
 moral import at all ? I think that few of them would 
 do so, in spite of the inconsistency in which their 
 refusal plunges them. Kant, indeed, accepted the 
 paradox rather than the inconsistency, but he had few 
 precursors, and he has few successors. His attempt to 
 prove that the teaching of Jesus is on his side can 
 only be described as astounding. ^ 
 
 The indeterminist might save his position if he were 
 prepared to maintain that, although certain completely 
 determined qualities are to be called moral excellences, 
 yet they are not to be ranked as the highest moral 
 excellences, a position which is to be reserved exclu- 
 sively for excellences of the will. But if he maintained 
 this he would have against him the authority of most 
 of the churches — certainly of the Christian church — 
 and of most of the philosophers. 
 
 130. We have now to consider the second ground 
 on which judgements of obligation are considered 
 incompatible with the complete determination of voli- 
 tion, A man who is condemned by a judgement of 
 obhgation — who is condemned, that is, as having done 
 wrong in what he has willed or omitted to will — is 
 held responsible for his conduct. And people are not 
 held responsible for anything else except an error of 
 voKtion, In ordinary language we say that a man may 
 be responsible for his ignorance, unskilfulness, or 
 some other defect which is not a defect of volition. 
 But what we really hold him responsible for is his vnll 
 not to remove the defect, or his abstention from willing 
 ^ Critique of Practical Measofij Part I, Book I, chap. iii. 
 
160 SOME DOGMAS OP EELIGION 
 
 to remove it. If his circumstances were such, for 
 example, that he could not have ceased to be ignorant, 
 however much he had tried, we should not call him 
 responsible for his ignorance. Even if the cause is not 
 external, but internal, we do not call him responsible 
 unless it is a volition or abstention from one. If 
 I write a play, I am responsible for writing it, for I 
 should not have done so if I willed not to do it. But, 
 if I write it, I am not responsible for its inferiority to 
 Hamlet, for the cause of that inferiority, though it is 
 to be found in my nature, is not in anyway dependent 
 on my will. 
 
 It is asserted that this responsibility would be in- 
 compatible with complete determination, and that 
 any one who is not prepared to reject responsibility 
 must be prepared to deny complete determination of 
 volitions. 
 
 131. Three sorts of responsibility have been asserted 
 — to our fellow men, to God, and to self. With respect 
 to the first I suppose that eveiy determinist who need be 
 reckoned with would admit that we are responsible to 
 our fellow creatures for defects of will, and that this 
 responsibility is only for defects of will and for their 
 results. 
 
 But is this inconsistent with determinism ? I can- 
 not see that it is. For although the determinist does 
 not hold that volitions are distinguished from all other 
 events by not being completely determined, he admits, 
 like every one else, that they are marked off from other 
 events by being determinable by expectation of plea- 
 sure and pain. Expectations of pleasure and pain are 
 not the only motives to will, but every one knows 
 that they are motives. On the other hand, nothing 
 but volitions can be directly determined by those 
 expectations, though other things may be determined 
 
FEEE WILL 161 
 
 by them through volitions. The fear of pain may 
 make a boy will to learn his lesson, or it may make 
 him will it more earnestly. And if the only obstacle 
 in the way was the absence or weakness of will, he 
 will now learn it. But if it is entirely beyond his 
 powers, the fear of pain may make him unhappy, but 
 will not make him successful. 
 
 Society, therefore, is quite justified in giving rewards 
 for right vohtions and in inflicting punishments for 
 wrong volitions, whether those rewards and punish- 
 ments are dehberately bestowed by the state, or 
 whether they are the less deliberate, but scarcely less 
 powerful rewards and punishments of social praise and 
 blame. For the expectation of such rewards and 
 punishments may encourage right volitions and dis- 
 courage wrong volitions. But to carry out a system of 
 this sort with regard to good and evil qualities not 
 dependent on vohtions would be foolish, and (in so far 
 as it was a system of punishments) brutal, since in 
 these cases the expectation would produce no effect on 
 the results. 
 
 132. Now I submit that my responsibility to my 
 fellow men for my volitions consists in the fact that it 
 is reasonable for them to reward and punish me for 
 my volitions, and in that fact only. And, in support 
 of this, we may notice that it is universally agreed that 
 a man is not responsible' in cases where his action 
 cannot be affected by considerations of pleasure and 
 pain. A lunatic who suffers from acute homicidal 
 mania is not hanged for murder, because the expecta- 
 tion of such a punishment would not deter a man in 
 such a condition. (It cannot, I imagine, be said that 
 he is not punished because he has not willed the 
 action. He has willed it as much as a sane murderer 
 has.) But when the same murderer is in the asylum 
 
 M 
 
162 SOME DOGMAS OP EELIGION 
 
 it is not thought wrong that he should be punished 
 for infraction of rules by exclusion from an entertain- 
 ment, because experience shows that the expectation 
 of this may affect his conduct for the future. The 
 homicidal maniac, then, is not held responsible for 
 murder, but is held responsible for untidiness, because 
 punishment will not prevent him from murdering, but 
 may keep him tidy. On the other hand, the cases of 
 certain of the more tyrannical despots, such as Nero, 
 support the contention from the other side. Psycho- 
 logically their states may have been quite as abnormal 
 as those of the ordinary homicidal maniac. Prom 
 a medical point of view they might perhaps be called 
 mad. But no court would hold them to have been 
 legally mad. It would have declared them responsible 
 for their actions. And this, I think, would have 
 been right. It is in the highest degree improbable 
 that Nero would have committed any of his crimes 
 if he had known that he would certainly have been 
 executed for them within a month or two. And 
 since the volitions of such men can be affected by the 
 expectation of punishment, it is right to punish them — 
 unless, of course, the punishment required to affect 
 them is so severe as to be a greater evil than the 
 crime. 
 
 133. It is clear from all this that the determinist 
 is not in the least inconsistent in advocating that 
 crimes should be punished. A preventive punishment 
 is obviously defensible in exactly the same way for 
 determinists and for indeterminists. Whether the 
 will is free or not, it is clear that while a man is in 
 prison he cannot be robbing on the highway or 
 breaking into houses. Deterrent punishment is justi- 
 fied for the determinist by the fact that experience 
 shows that the expectation of punishment will deter 
 
FKEE WILL 163 
 
 men from committing crimes which they would other- 
 wise have committed. And other sorts of pimishment 
 are justified for him by the fact that experience also 
 shows that a man's moral nature may in some cases 
 be improved by influences brought to bear on him 
 during the period of his punishment, or perhaps even 
 by the punishment itself.^ 
 
 There remains vindictive punishment. With regard 
 to this it need only be said here that the justifica- 
 tion of it is at least as easy for the determinist as for 
 the indeterminist — or, rather, not more impossible. 
 So far as punishment is vindictive, it makes a wicked 
 man miserable, without making him less wicked, and 
 without making any one else either less wicked or less 
 miserable. It can only be justified on one of two 
 grounds. Either something else can be ultimately 
 good, besides the condition of conscious beings, or the 
 condition of a person who is wicked and miserable is 
 better, intrinsically and without regard to the chance 
 of future amendment, than the condition of a person 
 who is wicked without being miserable. If either of 
 these statements is true — to me they both seem 
 patently false — then vindictive punishment may be 
 justifiable both for determinists and indeterminists. 
 If neither of them is true, it is no more justifiable for 
 indeterminists than it is for determinists. 
 
 As to vindictive punishment then, the two schools 
 are, so far as we have yet seen, on a level. ^ But they 
 are not on a level as to other sorts of punishment. 
 The indeterminist may, indeed, believe in preventive 
 punishment, since he attributes no indeterminate 
 volition to bolts and locks. But, as we shall see later 
 
 ^ Cp. my Studies in Hegelian Cosmology, chap. v. 
 2 Cp. Section 151. 
 
 M 2 
 
164: SOME DOGMAS OF RELIGION 
 
 on, he is quite inconsistent in supporting either 
 deterrent or reformatory punishment, 
 
 134. Eesponsibility towards man, then, is not 
 affected by determinism. But how about responsibiHty 
 towards God? God's judgement — on the hypothesis 
 that there is a God to judge — about the moral state of 
 any man could not be affected by determinism. If the 
 man is bad, he is bad, even if he is so necessarily, and 
 an omniscient being would recognize this badness. 
 But responsibility, as we have seen, involves more 
 than this. A man is not called responsible to his 
 fellow men because they do right to judge him evil, 
 but because they do right to punish him. Now, it is 
 argued by the indeterminists, it could not be right for 
 God to punish men if their actions were inevitably 
 determined by the natures which he had given them, 
 and the circumstances in which he had placed them. 
 
 It seems to me that the answer is this. If there is 
 an omnipotent God, we are not responsible to him for 
 our sins either on the determinist view or the indeter- 
 minist. If there is a God who is not omnipotent, then 
 we can as well be responsible to him for our sins on 
 the determinist view, as we can on the indeterminist 
 — or, indeed, better, as we shall see later on. 
 
 Punishment is painful, and pain is evil. Punish- 
 ment, again, does not abolish the sin for which it is 
 inflicted — that is in the past, and irrevocable. And 
 sin is evil. Consequently no person can be justified 
 in inflicting punishment if he might have avoided the 
 necessity by preventing the offence, unless the flnal 
 result of the sin and the punishment should be some- 
 thing better than would have happened without either 
 of them.^ And in this case he will not be justified 
 
 ^ Cp. my Studies in Hegelian Gosmologyj chap. vi. 
 
FEEE WILL 165 
 
 unless the good result which arises from the sin and 
 the punishment could not be attained without them. 
 For both the sin and the punishment ai'e intrinsically 
 evil. 
 
 Now on the determinist hypothesis an omnipotent 
 God could have prevented all sin by creating us with 
 better natures and in more favourable surroundings. 
 And any good result which might follow from the sin 
 and the punishment could be obtained by such a God, 
 in virtue of his omnipotence, without the sin or the 
 punishment. Thus God would not be justified in 
 punishing sin, though man would be, because God 
 could attain the desired results without the punish- 
 ment, while man could not. Hence we should not be 
 responsible for our sins to God. 
 
 135. But neither should we be responsible to an 
 omnipotent God on the indeterminist theory. For 
 such a God could have created us without free will, or 
 without any temptation to misuse it, and then there 
 would have been no sin. The common answer to this 
 is that a universe in which we inevitably did good 
 would be lower than one in which our action, whether 
 for good or for evil, was not completely determined. 
 Thus God is said to be justified in giving us free will, 
 and in punishing us when we misuse it. 
 
 I cannot see what extraordinary value lies in the 
 incompleteness of the determination of the will, which 
 should counterbalance all the sin, and the consequent 
 unhappiness, caused by the misuse of that will. If 
 God had to choose between making our wills imdeter- 
 mined and making them good, I should have thought 
 he would have done well to make them good. But we 
 need not decide this point. For the defence is one 
 which is obviously inconsistent with the idea of an 
 omnipotent God — if the word omnipotent is taken 
 
166 SOME DOGMAS OF EELIGION 
 
 seriously. The defence says that God could not secure 
 the benefits — whatever they are — of undetermined 
 vohtion without also permitting the evil of sin. But 
 there is nothing that an omnipotent being cannot do. 
 Even if the two were logically contradictory, a really 
 omnipotent being cannot be bound by the law of con- 
 tradiction. If it seems to us absurd to suggest that 
 the law of contradiction is dependent on the will of 
 any person, we must be prepared to say that no person 
 is really omnipotent. 
 
 Thus, even on the indeterminist hypothesis, we are 
 not responsible for our sins to an omnipotent God. 
 For he could have prevented the sins without intro- 
 ducing any counterbalancing evil into the universe. 
 And, consequently, he would not be justified in check- 
 ing sin by pain, since pain is intrinsically evil. If God 
 is omnipotent, then, responsibility is impossible on 
 either theory of freedom. 
 
 136. But if there is a God who is not omnipotent, 
 it would be quite possible for the determinist to hold 
 that we are responsible to him for our sins. Such 
 a God might be unable to create a universe without 
 sin, or at any rate unable to do so without producing 
 some greater evil. And he might find it possible, as 
 men do, to check that sin by means of a system of 
 punishments. In that case he would be justified in 
 doing so — provided, of course, that the necessary 
 pimishments were not so severe as to be a greater evil 
 than the sin. 
 
 And here the determinist is in a better position 
 than the indeterminist. For the indeterminist, as we 
 shall see later on,^ has no right to assert that there is 
 even a probability that the expectation of punishment 
 
 ^ Section 149. 
 
FEEE WILL 167 
 
 will alter our volitions. And, without such a proba- 
 bility, no punishment can be justified except vindictive 
 punishment. 
 
 It has sometimes been held that the freedom of the 
 human will was the only way in which the goodness 
 of God could be made compatible with the evil in the 
 universe. The evil we perceive in the universe con- 
 sists — at any rate chiefly, perhaps entirely— of sin and 
 misery. If sin was due to man's free will, and not to 
 God's decree, God, it was said, could not be condemned 
 on account of the existence of sin. And the misery 
 could be explained as the justifiable punishment of 
 sin. On the other hand, it was said, if there was no free 
 will, not only would it be impossible to justify the 
 existence of misery, but the sin also must be referred 
 to God as its ultimate cause. And it would become 
 impossible to regard a being as good, to whose nature 
 we must attribute the existence of all the evil of the 
 world. 
 
 But, as was shown above, if God is omnipotent, it is 
 impossible to account for the evil of the universe in 
 this way. Indeed, if God is omnipotent, it is impos- 
 sible that he can be good at all.^ This would not be 
 affected by the freedom of the human will, since the 
 gratuitous permission of evil would be as fatal to the 
 divine goodness as the gratuitous creation of evil. 
 On the other hand, if God is not omnipotent, his good- 
 ness would not be impossible upon either theory as to 
 the human wiU. For a being of limited power but 
 perfect goodness might well create evil, and not merely 
 allow it, supposing that the creation of the evil was 
 the only way of avoiding a greater evil or attaining 
 a greater good. 
 
 ^ Cp. chap. vi. 
 
168 SOME DOGMAS OF EELIGION 
 
 137. There remains the question of responsibility to 
 self. I think that this must be admitted to exist — that 
 a man does feel a responsibility to himself for defects 
 of volition, or for defects caused by defects of volition, 
 which he does not feel in cases of a defect with which 
 volition has nothing to do. The analogy between this 
 and responsibility to others seems to be that in the 
 latter I recognize that the others do well to punish me, 
 and in the former I recognize that I do well to feel 
 shame and remorse. Now why should I recognize 
 that it is well to feel shame and remorse for defects 
 which are — directly or indirectly — defects of voHtion, 
 and not for other defects ? The indeterminist would 
 suggest that it is because the defects of volition are not 
 completely determined. But why should I judge it 
 less good to feel shame or remorse for a defect because 
 it is an essential part of my character ? Surely, the 
 more closely a defect is bound up with me, the more it 
 is essential to my nature, the more reason I have to 
 feel ashamed of it. 
 
 It seems to me that the real reason why it is good 
 that I should feel shame and remorse in one case and 
 not in the other is the same as the reason why it is 
 good that other people should punish me in one case 
 and not in the other — namely, that in the one case it 
 may improve matters, and in the other case it cannot. 
 The only part of our nature which is influenced by the 
 expectation of pleasure and pain is the will. If, there- 
 fore, a defect is not a defect of volition, or dependent 
 on one, it will not be in any way affected by the fact 
 that I am miserable about it. And the misery, being 
 useless and painful, will be evil, and it wiU be well to 
 avoid it. 
 
 It is well that I should recognize my defects of all 
 sorts, since ignorance of one's limitations often pro- 
 
FREE WILL 169 
 
 duces evil. But when I have once recognized that 
 I cannot write a play as good as Hamlet, it is profit- 
 less self-torture if I am miserable about it, since 
 my misery will certainly not remove this particular 
 limitation. 
 
 With defects of vohtion, the matter is different. 
 My will can be affected by expectations of pleasure 
 and pain, and so, if the contemplation of a defect of 
 volition, or its consequence, gives me pain, I may be 
 led to cure the defect to escape from the pain. Or if 
 the defect is in the past, and irrevocable, the dread of 
 experiencing similar pain may keep me from similar 
 faults. In such cases shame and remorse may bring 
 advantages outweighing the evil of their painfulness, 
 and, since they will be profitable, it will be good to feel 
 them. 
 
 138. We now pass from the considerations drawn 
 from the validity of judgements of obligation to the 
 fourth division of the indeterminist arguments. This 
 was that the will must be free because, if it were not 
 free, aU choice would be absurd. It may be conceded 
 that scarcely any determinist would admit that all 
 choice was absurd, and therefore, if it can be shown 
 that this would be a result of determinism, they are 
 logically bound to give up determinism. 
 
 It is said that it is inconsistent for a determinist to 
 take duty as a motive for action. For he believes, it 
 is argued, that it is already completely determined 
 whether he will act according to his conception of 
 duty, or whether he will act otherwise. And this, it 
 is said, will render it unreasonable to choose to do his 
 duty. 
 
 The absurdity of this particular choice is the one 
 which is most often emphasized. But similar con- 
 
170 SOME DOGMAS OF EELIGION 
 
 siderations would prove that any other choice is as 
 absurd as the choice to do one's duty. For in each 
 case the determinist would believe his action to be 
 already completely determined, and if this made choice 
 absurd in one case it would do so in all others. 
 
 139. I cannot, however, see the least ground for the 
 conclusion that the belief in determinism makes choice 
 unreasonable. Of course, if the belief of the deter- 
 minist was that the end at which he was aiming was 
 completely determined to occur or not to occur, irre- 
 spective of what he chose, then choice would be 
 unreasonable.^ I should be very unreasonable to 
 choose that the sun should rise to-morrow, or that it 
 should not rise. But the ordinary determinist, like 
 everybody else, believes when he chooses any course 
 that his choice may have some effect on the event. 
 And he is quite consistent in this belief. He is a 
 determinist because he believes that, while the event 
 may well be determined by his choice, his choice is in 
 its turn completely determined. 
 
 Why should the belief that, if I choose to shut the 
 door, my choice to shut it was completely determined 
 beforehand, make it unreasonable of me to choose to 
 shut it ? (This is all the information my determinism 
 
 ^ Such a belief has been held, Napoleon, for example, seems 
 to have believed that the time of each man's death was fixed, 
 independently of all other events. If I go here, I may be 
 drowned, if I go there, I may be shot, but, wherever I go, I shall 
 die somehow at that hour. Such a belief does render absurd all 
 choice directed to the preservation of my life. Why should 
 I protect my life ? I shall either lose it to-day, even if I protect 
 it, or keep it till another time, even if I do not protect it. 
 
 But this is not identical with determinism. It does not even 
 involve it, for a man could hold this who held that my choice to 
 protect life or not to protect it was not completely determined. 
 It is compatible, no doubt, with determinism, but in this case 
 the absurdity of choice would not be due to the determinism, but 
 to this quite separate belief. 
 
FEEE WILL 171 
 
 can give me on the question. For, so long as I am not 
 omniscient, I can never be absolutely certain before- 
 hand what I shall choose. My certainty may be very 
 great but it can never be quite complete.) I cannot 
 see that it should have any such paralysing effect. 
 The contention that it ought to do so is, I think, due 
 to a confusion of this belief with the other behef, men- 
 tioned above, which treats the choice as impotent to 
 affect the result, and which asserts that the result is 
 determined irrespective of it. 
 
 But suppose that I was omniscient, so that I could 
 not choose anything without knowing beforehand that it 
 was certain that I should choose it, would that render 
 choice unreasonable? I find it rather difficult to con- 
 ceive what would happen in circumstances so unlike 
 those of which I have any experience. But I can see 
 no absurdity in a choice which is preceded by a perfect 
 knowledge that it would be made. It is to be observed 
 that most theists would hold that God could predict 
 with absolute certainty how he would will in any 
 circumstances, and that they would not hold that this 
 made it absurd in him to will. 
 
 140. There remains the defence of free will which 
 is based on the disastrous consequences which, as it is 
 asserted, would follow from the falsity of that doctrine. 
 I have endeavoured to show, in Chapter 11, that such 
 arguments are never valid. But the practical conse- 
 quences of the truth of any doctrine are always inter- 
 esting, even if they tell us nothing as to its truth or 
 falsity. Let us consider if any disastrous results would 
 follow from the truth of determinism. 
 
 It would doubtless be a great disaster if it were 
 absurd to recognize virtue as binding on us, or to adopt 
 a course of action because we believed it to be our duty 
 to do so. But we have seen that there is nothing in 
 
172 SOME DOGMAS OF RELIGION 
 
 determinism which makes judgements of obUgation 
 absurd, or which renders it absurd to be moved by 
 a regard for duty. 
 
 It would also be a great disaster if people did in fact 
 — even without any logical excuse — no longer recognize 
 virtue as binding on them, or no longer adopted a course 
 of action because they believed it to be their duty. It 
 would be also a great disaster if they did these things 
 less frequently or less intensely. But there is no 
 reason, that I know of, to suppose that a belief in 
 determiaism would tend to bring about these calami- 
 ties. Many men, and many communities, have, at 
 different times and for long periods together, accepted 
 determinism, and I do not think that their moraUty 
 has been observed to be inferior to that of communities 
 whose general circumstances were similar, but who 
 rejected determinism. Indeed, the usual course of 
 indeterminist polemics has admitted that determinists 
 are no less virtuous than other persons, and has de- 
 voted itself to attack what it asserts (for the reasons 
 which we have already considered) to be their incon- 
 sistency in being virtuous. 
 
 141, It is true that a belief in determinism will 
 tend to modify the emotion with which we regard 
 those whom we hold to be acting wrongly. There is 
 nothing inconsistent in an indeterminist pitying a 
 sinner for his sinfulness. But, ceteris paribus, a deter- 
 minist is more likely to do so. For he does not, like 
 the indeterminist, hold that among the differences which 
 separate sin from other defects is to be found the fact 
 that sin is not completely determined, and that this 
 difference is of fundamental moral importance. It is 
 more natural, therefore, for him, than it is for the 
 indeterminist, to look at sin, like other defects, as 
 a calamity to the man who suffers from it. And the 
 
FREE WILL 173 
 
 greater his hatred of sin, the greater will he consider 
 the calamity, and the greater will be his pity. 
 
 The increase of pity for sinners, however, is not 
 a calamity. It might be so if it were incompatible 
 with hatred of the sin, or with the resolve to extirpate 
 it by any remedy which would not be worse than the 
 disease. But it is clear that there is no logical incom- 
 patibility between pity for sinners and the resolve to 
 extirpate sin. And they are not psychologically in- 
 compatible, for they are often found together in great 
 intensity. The httle, for example, that we know about 
 the life of Jesus suggests that he combined an invari- 
 able intolerance of sin with an almost invariable 
 compassion for the sinner. 
 
 142, It is said that it would be intolerable for a man 
 to believe that perhaps he was inevitably determined 
 to be wicked in the future. But has the man in 
 question any desire to be virtuous, or has he not ? If 
 he has no desire to be virtuous, he is not likely to find 
 his life made intolerable by the possibility (it can 
 never be a certainty for a being whose knowledge is 
 finite) that he is inevitably determined not to be 
 virtuous in the future. 
 
 If on the other hand he has a desire for virtue, he 
 knows by experience that such a desire is an important 
 factor in determining a man to be virtuous and not to 
 be wicked, so that the chance that he is determined 
 to be wicked is diminished by the existence of the 
 desire. And the stronger such a desire is, the more 
 hkely is it to prevent his future existence from being 
 wicked. Thus the more intolerable the prospect of 
 future wickedness is to any man, the less fear is there 
 that such a fate should befall him. Such a state of 
 things is no great calamity. 
 
 Of course, a man's present desire to be virtuous is 
 
174 SOME DOGMAS OF RELIGION 
 
 far from being an absolute guarantee that he will not 
 in the future be wicked. Many men have had such 
 a desire who have afterwards become wicked. My 
 present desire is one factor in determining my future 
 moral state, and a very important factor. But it is 
 not the only factor, and the tendency may be over- 
 come by others. 
 
 But here the determinist is no worse off than the 
 indeterminist. For the indeterminist, who desires 
 to-day to be virtuous, cannot deny that other men 
 who have desired to be virtuous have subsequently 
 become wicked. The same fate may be his. He is in 
 the same position of uncertainty as the determinist, 
 except that the determinist may logically speak of 
 a probability one way or the other, while the indeter- 
 minist has no right to do so. The indeterminist, 
 indeed, can assure himself that, if he does fall into 
 wickedness, no one will be able to give a complete 
 explanation of why he has fallen. I must confess 
 that I fail to see what comfort can reasonably be 
 derived from that assurance. 
 
 If a man desires food, or love, or anything else, the 
 important question for him is whether he will get it 
 or fail to get it, — not whether his success or failure 
 can be explained as the inevitable result of what has 
 preceded. The question as to the manner in which 
 the result has been determined may have a theoretical 
 interest, but has no practical importance. Surely, in 
 the same way, all that interests the man who desires 
 to be virtuous, in so far as he desires it, is to know 
 whether he will be virtuous or not. Whether virtue 
 or wickedness comes to him from his own uncaused 
 caprice in the future, or from the eternal nature of the 
 universe, it will stiU be virtue or wickedness. 
 
 It is clear that neither determinism nor indeter- 
 
FEEE WILL 175 
 
 minism, taken by itself, can settle the question whether 
 virtue or wickedness will finally prevail, either with a 
 particular person, or with the universe. But it is also 
 clear that, if the question is to be settled at all by 
 metaphysics, it can be settled only on the basis of 
 determinism. For virtue and wickedness are depen- 
 dent on the will, and if indeterminism is true it is 
 impossible to predict the future state of the will of 
 any being, or of all beings. 
 
 143. It has been said that, if we do not possess the 
 freedom of indetermination, we shall be no better than 
 machines, Kant speaks thus of the freedom of the 
 Monads, of which Leibniz had asserted that they were 
 free because they were determined internally by their 
 own spiritual nature. *It would', he says, 'be really 
 nothing better than the freedom of a roasting-jack, 
 which also, when it is once wound up, performs its 
 motions of itseK.' ^ 
 
 It would, doubtless, be a calamity if men stood no 
 higher in the scale of values than roasting-jacks. But 
 Kant seems to me to be wrong here. If a roasting- 
 jack goes right, a joint is well cooked, if it goes wrong, 
 it is badly cooked. If a will goes right, the man is 
 virtuous ; if it goes wrong, he is wicked. If the 
 difference between virtue and wickedness is no more 
 important than that between a well-cooked leg of 
 mutton and one which is badly cooked, then, certainly, 
 internal determination (which I have called above the 
 freedom of self-determination) is of no more worth in 
 a man than in a roasting-jack. But if, as Kant would 
 certainly have admitted, the first difference is very 
 much more important than the second, then freedom 
 of self-determination in a man, which determines 
 
 ^ Critique of Practical Beason, Part I, Book I (p. 101, Harten- 
 stein's ed.). 
 
176 SOME DOGMAS OF EELIGION 
 
 between virtue and wickedness, is much more impor- 
 tant than the freedom of self-determination in a 
 roasting-jack, which only determines whether the 
 joint shall be well or badly cooked. 
 
 It must also be remembered that a man may be 
 held to be free because he possesses, in many cases, 
 what I have called freedom of self-direction. A roast- 
 ing-jack cannot have this, since it has no will. 
 
 144. It would be difficult, indeed, to imagine any- 
 thing more horrible than some of the determinist 
 views of the universe which have been held in the 
 Christian Church. But what made them horrible was 
 not the belief in determinism, but the belief in hell. 
 The belief that we are predestined to salvation is, as 
 the 17th Article of the Anglican Church remarks, a 
 belief which produces happiness, and not misery. 
 
 But is not a behef in hell even more horrible when 
 combined with determinism than it would be without 
 determinism? This has been asserted, but I think it 
 rests on a misconception, or rather on two. The first 
 is the belief that hell would be less unjust in a universe 
 where our wills were not completely determined than 
 in one where our wills are completely determined. 
 But we saw above that our responsibility to God was 
 not affected by the truth or falsity of determinism. 
 This view, therefore, must be rejected. 
 
 The second misconception ascribes to determinism 
 a horror which is really due to another belief — the 
 belief that salvation cannot be obtained by any action 
 dependent on the will. So far as people believe that 
 salvation can be obtained by such actions, their fears — 
 about themselves at any rate — would have some 
 alleviation even if they were determinists. For in 
 proportion as they were really anxious to obtain 
 salvation they would be likely to obtain it, since 
 
FEEE WILL 177 
 
 a will to act in a particular way makes it likely that 
 we shall act accordingly. On the other hand, if some- 
 thing is required for salvation which is in no way 
 dependent on the will to obtain it, the tortures of 
 apprehension might rise to any intensity for indeter- 
 minists, as well as for determinists. For then the 
 desire for salvation, while it made the fear of losing it 
 more intense, would not in any way make the chance 
 of losing it less. 
 
 145. I have now considered the principal arguments 
 brought forward by indeterminists in favour of their 
 position. It remains to consider some arguments 
 which may be brought forward against that position. 
 The main argument against it is that which proceeds 
 by establishing the universal validity of the law of 
 Causality, and so showing that volitions, like all other 
 events, must be completely determined. This argument, 
 as I said at the beginning of the chapter, I do not 
 propose to consider here. I shall only poiDit out two 
 inconsistencies in the position of the average indeter- 
 minist. The average indeterminist, like other men, 
 admits the vaUdity of morality ; and he attaches some 
 value to expectations that men will, under certain 
 circumstances, act in certain ways. I maintain that in- 
 determinism is inconsistent with both these positions. 
 
 146. In the first place indeterminism is inconsistent 
 with the vahdity of morality. Determinism, as we 
 have seen, was reproached with its inconsistency with 
 the vahdity of judgements of obhgation, and reproached 
 wrongly. The accusation can be retorted with greater 
 truth. 1 Judgements of obligation are judgements 
 
 ^ It is interesting to note in passing that this view is main- 
 tained by thinkers so different from one another as Hume and 
 Green. Compare Treatise of Human Nature, Book II, Part III, 
 Section 2; and Frolegomena to Ethics j Section 110. 
 
 N 
 
178 SOME DOGMAS OF EELIGION 
 
 which approve or condemn the person who wills 
 a certain thing. I say that I myself, or some one else, 
 is better or worse on account of a particular volition 
 than would have been the case if the volition had not 
 occurred. The approval or condemnation of the agent 
 is essential to morahty. If we are not entitled to say 
 that a man is virtuous or wicked, what is left ? 
 
 But how are we justified in passing from the volition 
 to the person who makes it? There is, perhaps, no 
 difficulty on any theory in saying that a man is good 
 or bad at the moment when he is willing well or 
 badly. But this is not all that we do say. Half an 
 hour after ordering a murder, Nero may be eating 
 his dinner, and thinking about nothing else. In the 
 intervals of his labours, St. Francis, too, must eat, and 
 may be too fatigued even to plan fresh labours. Yet 
 we should call the one wicked, on account of his past 
 crimes, and the other good, on account of his past 
 services. The whole fabric of morality would be 
 upset, if our approval or condemnation of a man for 
 his voHtion had no right to last longer than the volition 
 itself. Nor would any indeterminist, I imagine, be 
 prepared to deny its right to last longer. 
 
 147. The determinist can explain this consistently 
 with his position. According to him the volitions 
 of each man spring from his character, and are the 
 inevitable result of that character when it finds itself 
 in a certain situation. The approval or condemnation 
 of the agent is based on the belief that the character, 
 indicated by the past acts, survives in the present, 
 and is ready, on appropriate occasions, to manifest itself 
 in similar acts. Nero is condemned in the present, 
 because he still has the character which will probably 
 cause him, when he is tired of eating, to amuse him- 
 self with another murder. St. Francis is approved in 
 
FEEE WILL 179 
 
 the present, because he still has the character which 
 will probably cause him, when he has satisfied his 
 hunger, to perform fresh works of benevolence. 
 
 But how can the indeterminist defend his judge- 
 ment ? According to him the vohtion in each case is 
 a perfectly undetermined choice between two motives. 
 When the volition is over, it has ceased to exist, and 
 it has not, on the indeterminist theory, left a per- 
 manent cause behind it. For, according to that 
 theory, it has no permanent cause at all. Directly 
 Nero has ceased to think of a murder, nothing at all 
 connected with it remains in his moral nature, except 
 the mere abstract power of undetermined choice, which 
 is just as likely to be exercised on the next occasion 
 in an utterly different way. How then can the 
 indeterminist venture to call Nero a wicked man 
 between his crimes? And yet he certainly would 
 call him so. 
 
 Are we to say that it is, after all, the same person 
 who committed the murder and who is now being 
 condemned, and that this forms a sufficient justification 
 of the condemnation ? I cannot see that this should 
 justify it. For the judgement passed on Nero at 
 dinner is not only that he was wicked when willing 
 the murders before dinner, but that he is wicked now. 
 But what is wicked in him now? Not his volition, 
 for he is now only willing to gratify his palate, which 
 is not wicked. Not his character, for his previous 
 volition consisted in an undetermined choice of the 
 wrong alternative, and this has no root in his character 
 — or anywhere else. 
 
 148. Moreover, if the indeterminist adopts this de- 
 fence, he involves himself in fresh inconsistencies. If 
 a man is to be approved or condemned now, simply 
 because he is the same man who willed well or badly 
 
 N 2 
 
180 SOME DOGMAS OF EELIGION 
 
 in the past, then all past volitions are equally grounds 
 for such an approval or condemnation. But there are 
 cases in which every one, including indeterminists, 
 would admit that past volitions are not equally grounds 
 of approval or condemnation in the present. 
 
 We often will in our dreams to commit evil actions. 
 And men often will, in a state of intoxication, to 
 commit evil actions which they would not will to 
 commit when sober. Now we never consider a man to 
 be wicked, after he has awaked, on account of his evil 
 volitions in dreams. And we do not consider him 
 wicked, when sober, on account of the wickedness 
 which he willed when drunk. In ordinary cases, 
 indeed, we should blame him for getting drunk. But 
 when this is not his fault —as in the case of a savage 
 who tastes alcohol for the first time — then no blame 
 whatever is attributed to him after he has got sober 
 for the volitions he formed when drunk. 
 
 Such judgements are quite inconsistent with the 
 theory in which the indeterminist has now taken 
 refuge — the theory that a man may be approved or 
 condemned in the present for any volition which 
 he has made in the past. For the man who dreamed 
 or was drunk is certainly the same man who is now 
 awake and sober. 
 
 But the determinist avoids all inconsistency. Ex- 
 perience shows us that the conditions of dreaming or 
 of intoxication so affect the moral character that no 
 inference can be drawn from volitions made in one of 
 these states as to the probable volitions of the same 
 man when he is not in that state. I have no reason 
 to think that the brutal murder, which I planned in 
 a dream on Monday night, gives even the least in- 
 dication that I should be likely to yield, when awake 
 on Tuesday morning, to any temptation to commit 
 
FEEE WILL 181 
 
 murder. And therefore, when awake on Tuesday 
 morning, I do not condemn myself for it. 
 
 Again, in so far as we believe a man to have really 
 altered his character, we no longer blame him for 
 what happened before it was altered. If, since he 
 willed a certain crime, his conduct has shown that he 
 can resist temptations similar to those to which he 
 3delded before, we no longer condemn him. Even 
 without this experience, if we have reason to believe 
 that he has repented his crime so sincerely and effec- 
 tually that he would not in future yield to a similar 
 temptation, we condemn him no longer. It may be 
 necessary to punish such men, either as an example to 
 others, or because the law cannot safely take account 
 of such delicate and doubtful matters. But from the 
 point of view of morality, he is not condemned. 
 
 For determinists this is completely consistent. They 
 pronounce a man to be wicked in the present, who has 
 committed a crime in the past, because they regard it 
 as evidence of a still-existing character of the kind which 
 tends to produce crimes. But if they have reason to 
 beHeve — such as amendment or repentance can give 
 them — that his character has changed since the crime, 
 and is no longer such as tends to produce similar 
 crimes, they have no longer any reason to condemn 
 him. 
 
 The indeterminist recognizes that amendment and 
 repentance may remove wickedness as much as the 
 determinist recognizes it. To deny this would be to 
 break with every religion in the world's history, and 
 with the moral judgement of all mankind. But his 
 recognition of this is quite inconsistent with his in- 
 determinism. We have seen that he can only condemn 
 a man for a past crime at all, on the basis that it is 
 sufficient groimd for condemnation that he is the same 
 
182 SOME DOGMAS OF RELIGION 
 
 man who committed the crime. And if this is a 
 sufficient ground, then it is clearly unjustifiable to 
 condemn him the less on account of his amendment 
 or repentance. 
 
 149. We pass to the second inconsistency involved in 
 the position of the indeterminist. The indeterminist, 
 Hke every one else, assumes that it is possible to pre- 
 dict, with some probability, though not with absolute 
 certainty, how men will act under particular circum-' 
 stances. To reject this would render impossible all 
 trade, all government, and all intercourse with our 
 fellow men. 
 
 We continually act on the faith of such predictions. 
 We assume that a postmaster will sell a penny stamp 
 for a penny, that he will not sell two for a penny, 
 that a policeman will not try to kill us for walking 
 along the Strand, that a soldier in battle wiU try to 
 kill the enemy, and the like. There is, of course, 
 no certainty. The postmaster may be drunk, the 
 policeman a homicidal maniac, the soldier a disciple 
 of Tolstoi. But we are confident of the probabihties. 
 I am more likely to get one penny stamp for a 
 penny than to get two. My life is more hkely to 
 be attempted in a fight at close quarters than in 
 a walk through London. 
 
 The indeterminist admits that on his theory there 
 can be no certainty of prediction. For all practical 
 purposes the determinist must admit the same, since 
 only an omniscient person could be quite certain what 
 causes were at work, and with what strength. The 
 indeterminist, however, thinks that his theory admits 
 of statements of probability as to volitions. 
 
 If, however, indeterminism is true, there is no 
 justification whatever for making any statement as to 
 the probability of future volitions. The indeterminist 
 
FKEE WILL 183 
 
 theory assumes that in every case the choice between 
 motives is undetermined. There cannot then be the 
 slightest probability that this choice will be of one 
 motive rather than another. Our only ground for sup- 
 posing that a particular man will choose in a particular 
 way, under particular circumstances, is that experience 
 has shown us that he has previously acted in a similar 
 way under similar circumstances, or else that most 
 men, or most men who resemble him in certain ways, 
 have previously acted in a similar way under similar 
 circumstances. Now why should we suppose that 
 similar circumstances will be followed by similar 
 results? There is no reason to do so unless the cir- 
 cumstances determine the results, or the circumstances 
 and the results are both determined by the same 
 cause. Otherwise the expectation that the similar 
 results would follow would be as foohsh as the ex- 
 pectation that I should win at cards on one Lord 
 Mayor's day because I had won at cards on the 
 previous one. 
 
 According to the indeterminist theory our choice 
 between motives is not determined by anything at all. 
 And thus it follows that all groimd for predicting the 
 action of any man, so far as it depends on his volition, 
 vanishes altogether. 
 
 One result of this is that the indeterminist is quite 
 inconsistent in expecting one line of conduct from one 
 man and another from another. It is just as probable 
 that an English general to-day should eat his prisoners, 
 as it was that a Maori chief should do so a hundred 
 years ago. It is just as probable that the drunken man 
 in the street should be Johnson as that it should be 
 Boswell. 
 
 But this is a trifle. If the indeterminist is right we 
 have no reason to expect any line of conduct from 
 
184 SOME DOGMAS OF EELIGION 
 
 any one, rather than any other line of conduct which 
 is physically possible. It is just as likely that the 
 majority of Londoners will burn themselves aUve 
 to-morrow, as it is that they will partake of food to- 
 morrow. I am just as likely to be hanged for brushing 
 my hair as for committing a murder. "When men 
 commit suicide, or eat, or hang other men, their action 
 depends on their volition, and their volition cannot be 
 anticipated. 
 
 150. In the general chaos to which this would 
 reduce our life, it is scarcely worth mentioning the 
 result as to punishment. Preventive punishments 
 might still be justified, because they trust to keeping 
 the prisoner's body in a place where he will be 
 harmless, and because their effect does not depend on 
 his volition. Yet even here the punishment could not 
 be carried out if the gaolers decided one morning to 
 release all the prisoners. And on the indeterminist 
 theory they are just as likely to will this as they are 
 to will anything else. 
 
 All other non-vindictive punishments would be 
 completely absurd. For they all depend on judge- 
 ments as to the probability of future vohtions. It 
 is believed that a man is less likely to will to commit 
 a crime which he desires to commit, if he knows that 
 the commission of the crime will produce consequences 
 which will be painful to him. And it is believed that 
 a man who has been subjected to certain influences is 
 less likely to desire to commit crime. If indeterminism 
 is true, both these beliefs are absolutely baseless, 
 
 151. There remains vindictive punishment. It has 
 often been believed that indeterminism would at 
 least justify this — a belief which has sometimes been 
 taken as the chief support of indeterminism. We 
 admitted provisionally, earlier in this chapter, that, 
 
PEEE WILL 185 
 
 if vindictive punishment could be justified at all, it 
 might be as well justified for the indeterminist as for 
 the determinist. But there is a difficulty. Vindictive 
 punishment is inflicted on a man because of his 
 wickedness. It is seldom, however, that it can be 
 inflicted at the very moment when he is sinning. 
 And we have seen that to regard a man as wicked 
 on account of a past volition is entirely inconsistent 
 with indeterminism. Yet vindictive punishment does 
 stand in a specially close relation to indeterminism. 
 For at least we can say that, in those cases where the 
 punishment begins before the sin ceases, vindictive 
 punishment is not more absurd for the Ofitgnnizust 
 than it is for other people. iMDeTEeM \Mi ST 
 
 S e^ PC J. 
 
CHAPTER VI 
 
 OOD AS OMNIPOTENT 
 
 152. By God I mean a being who is personal, 
 supreme, and good. In calling him personal, I mean 
 to assert that he is self-conscious, that he has that 
 awareness of his own existence which I have of my 
 own existence. In calling him supreme, I do not 
 mean to assert that he is omnipotent, but that he is, 
 at the least, much more powerful than any other being, 
 and so powerful that his volition can profoundly affect 
 the whole sum of existence. In calling him good, I do 
 not mean to assert that he is perfect, but that he is, at 
 the least, of such a nature that he would be rightly 
 judged to be more good than evil. 
 
 All these three characteristics are, I think, implied 
 in the word God, as used in Western theology. A 
 reality which was not a person would not here be 
 given the name of God. Cases can be quoted, no 
 doubt, where the name has been used of an impersonal 
 reality. But I think that such statements do not mean 
 more than that the reality in question is a worthy 
 substitute for a God, or that the behef in it is a 
 worthy substitute for a belief in a God. They do not 
 mean that the name is appropriate in a strict sense 
 to an impersonal reality. When the schoolboy was 
 told that the school was his father, his mother, and 
 the rest of his relations, the meaning was that it ought 
 to take their place with him. It is in this way, I be- 
 lieve, that we must interpret the statement that Law, 
 or Nature, or Goodness, or Duty, is God. 
 
 It is sufficiently clear that a person is not called 
 
GOD AS OMNIPOTENT 187 
 
 God, unless he is supposed to be the supreme being in 
 the universe, or, at least, one of a limited number of 
 such beings. And it cannot, I think, be doubted that 
 a supreme being, who was not believed to be good, 
 would not be called God. It is true that some 
 thinkers have admitted that the action of their God is 
 not what would be called good according to human 
 standards. But then they pronounce those standards 
 to be unfit for application to God. In spite of those 
 standards, they tell us, we must call him good, we 
 must praise and worship him, and we have no need to 
 despise ourselves for doing so. If a man believed in 
 a supreme person whom he behoved to be bad, he 
 would not think him worthy of the name of God — 
 a name which, when used by a behever in God, always 
 implies reverence. 
 
 153. The usage in philosophy, however, is sometimes 
 different from the usage in theology. In philosophy 
 we have high authority — including Spinoza and Hegel 
 — for a different practice. God is frequently defined 
 by philosophers as the true reality, of whatever nature 
 that reality may be, provided only that it possesses 
 some sort of unity, and is not a mere chaos. If the 
 word is used in this way, every person, except absolute 
 sceptics or the most extreme plurahsts, must be said 
 to beheve that a God exists. The question of the 
 existence of God, on this definition, becomes very 
 trivial. The important question is not whether there 
 is a God, but what sort of nature he, or it, possesses. 
 
 If the usage of theology and philosophy differ, which 
 ought to give way ? It seems to me that it should be 
 philosophy. It is possible to change the meaning of 
 terms which are used by a comparatively small number 
 of students, but impossible to change the meaning of 
 so common a word as God in popular usage. Now 
 
188 SOME DOGMAS OF EELIGION 
 
 popular usage is distinctly in favour of the narrower, 
 or theological, use of the word, and philosophy ought 
 to accommodate itself to this, to avoid a dangerous 
 ambiguity. 
 
 Again, while the conception of the true reality is of 
 fundamental importance for philosophy, the conception 
 of a supreme and good person is also of great philo- 
 sophical significance. It is desirable that each of them 
 should have a separate name in philosophical termino- 
 logy. For the second no name but God has ever 
 been proposed, while the first is often called the Abso- 
 lute. If God is used as a synonym for the Absolute, 
 we should have two philosophical names for one 
 important idea, and no name for the other. 
 
 Finally, philosophical usage is by no means uniform. 
 Against Spinoza and Hegel we may put Kant and 
 Lotze, both of whom use the word substantially in the 
 theological sense. The balance of convenience, then, 
 seems greatly in favour of confining the name of God 
 to a being who is personal, supreme, and good. 
 
 154. Such a definition, of course, leaves many ques- 
 tions still open. Two of these may be regarded as of 
 fundamental importance from a religious point of view. 
 The first of these is whether God is Omnipotent in the 
 strict sense of the word — whether he could do anything 
 whatever, or whether there are things which he could 
 not do. The second is whether all existent reality 
 other than God has been created by him, or whether 
 there are other beings — personal or impersonal — whose 
 existence is as ultimate and uncaused as God's own. 
 
 If God is not creative, he cannot be omnipotent. 
 If there are beings whose existence is as much an ulti- 
 mate fact as his own existence, then he could not have 
 prevented their existence, and therefore his power 
 would be limited. But even a creative God need not 
 
GOD AS OMNIPOTENT 189 
 
 be omnipotent. It is possible that he should have 
 enough power to create this universOj or one of a rather 
 different nature, or perhaps to have abstained from 
 creating a universe at all, and yet that in creating he 
 acted under limitations which would prevent him from 
 doing certain things. We have therefore three possi- 
 bilities, a creative and omnipotent God, a God who is 
 creative but not omnipotent, and a God who is neither 
 creative nor omnipotent. In this chapter I shall con- 
 sider the first of these. 
 
 This theory scarcely requires further explanation, 
 since it is familiar to every one. It holds that there 
 is a reahty other than God, but that this reality is 
 entirely dependent on God. God, on the contrary, is 
 not in any way dependent on this other reahty. He 
 is held to have created it, and to have been purely self- 
 determined in that act of creation. The nature of what 
 he created, and the character of whatever changes have 
 befallen it, are held to be entirely due to his will, 
 which is not believed to be acting under any Hmitations 
 which can hinder its fulfilment. 
 
 155. What reasons are generally given for holding 
 such a theory ? We are often told by its supporters 
 that they believe it because they have an instinctive 
 and irresistible conviction of its truth. It is not neces- 
 sary to recapitulate here what has been said in 
 Chapter II on this ground of belief. 
 
 It is also frequently said that there must be a God, 
 and that God must have a particular nature, because 
 such disastrous consequences would follow from the 
 non-existence of a God, or from the existence of a God 
 with a different nature. The use of such considera- 
 tions as a ground of belief has also been discussed in 
 Chapter II. I shall, however, consider later on the 
 practical effects of behef or disbelief in a God, not as 
 
190 SOME DOGMAS OF RELIGION 
 
 bearing on the truth of the belief, but as a matter of 
 practical interest. 
 
 We have Likewise discussed in Chapter II the possi- 
 bility of basing behef in the existence of God on 
 miracles. It is to be noted, however, that this is not 
 a very common ground for the belief. It is frequently 
 admitted by those persons who accept certain dogmas 
 on the evidence of miracles that the evidence of 
 miracles depends on the belief that there is a God, 
 and a God whose nature is incompatible with deceit 
 or error. From this, of course, it follows that the 
 miracles cannot be used to prove the existence of God. 
 
 Passing to more direct arguments, we find that Kant 
 enumerates three, which he names Ontological, Cosmo- 
 logical, and Physico-Theological. I do not propose to 
 discuss the Ontological argument. It had very little 
 influence at any time except among professed meta- 
 physicians, and, since Kant's destructive criticism, it 
 has been abandoned by all schools of philosophy. 
 (Hegel, indeed, asserts that Kant has not disproved it. 
 But what Hegel calls the Ontological argument is not 
 what was previously called so, and is not what Kant 
 intended to disprove. Nor would what Hegel calls 
 the Ontological argument sufl&ce to prove a personal 
 God.) 
 
 Kant's Cosmological argument resembles very closely 
 the argument from the necessity of a first cause, while 
 his Physico-Theological argument is what is more 
 commonly called the argument from design. Both 
 these arguments are in frequent use at the present 
 day, and both will require our consideration. We 
 shall also have to consider the argument that the truth 
 of Idealism would involve the existence of an omni- 
 scient being, and that such a being could only be God. 
 
 156. The argument for a first cause may, I think, be 
 
GOD AS OMNIPOTENT 191 
 
 fairly stated as follows. Events continually happen 
 in the world, and every event requires a cause. And 
 the non-divine substances which exist in the world 
 cannot have existed eternally. The commencement 
 of the existence of each of these substances is also an 
 event, and will require a cause. Now an event may 
 be caused by another event, but then this earher event 
 will, like every other, require a cause. If we recognize 
 no causes but events, every cause will itself require 
 a cause, and the series will never be completed. We 
 must, therefore, believe that the ultimate cause of all 
 events is not an event but a being, who never began to 
 exist, and who therefore needs no cause. And this 
 being is said to be God. 
 
 This argument, even if valid, would not be sufficient 
 to prove the existence of God. Even if it proved the 
 existence of a being who was a cause without being 
 itself causally determined, it could not prove that such 
 a being was either personal or good. The nature of 
 the effects that it produced might prove its personality 
 and goodness, but the mere fact that it was a cause 
 would be consistent with no personality or with an 
 evil personahty. Thus the argument could not reach 
 the desired conclusion without calling in the aid of the 
 argument from design. 
 
 157. But why, to go further, is a cause which is not an 
 event wanted at all ? (To speak accurately, we should 
 say, a cause which does not include an event. For the 
 full cause of any event always includes many condi- 
 tions and circumstances which are not events.) Why 
 should we not derive one event from another, and this 
 again from another, backwards through unending time, 
 without any first cause at all ? As for the finite sub- 
 stances, they might either be conceived as existing 
 through unending past time, or as brought into exist- 
 
192 SOME DOGMAS OF EELIGION 
 
 ence at some time. In the latter case the event of their 
 commencing to existwould, like all others, be dependent 
 on an unending series of previous events. 
 
 The supporters of the necessity of a first cause 
 would argue against this as follows. The event -4, 
 with which we start, has a definite nature — certain 
 predicates are true of it, others are false. The law of 
 Causality tells us that the nature of A is entirely 
 dependent on its cause. If (as the supporters of the 
 unending regress assert) A is to be explained by an 
 unending regress of events, then its immediate cause 
 wiU contain an event, which we may call B. The 
 nature of A is then dependent, among other things, 
 on the nature of B. It cannot be determined unless 
 B is determined. But B is an event, and, on the same 
 principle, will have a cause containing an event C, 
 Then B is not determined unless C is determined. 
 And, consequently, A is not determined unless C is 
 determined. And^ similarly, A is not determined 
 unless D, jEJ, and so on are determined. The conclu- 
 sion drawn is that A cannot be determined unless 
 the earhest event in the chain is determined. But 
 there is no earhest event in the endless series. There- 
 fore A is not determined at all. And this is impossible. 
 We must therefore either reject the law of Causality, 
 which makes the determination of an event dependent 
 on the determination of its cause, or we must assert 
 that the series of causes is not endless. And this can 
 only be done by asserting the existence of a first cause 
 which is not an event. 
 
 158. I shall not inquire whether this conclusion is 
 justifiable, nor is the inquiry necessary for our present 
 purpose. Even if we grant, to the believers in a first 
 cause, that an unending regress of causes is impos- 
 sible, their position wiU be untenable. For the 
 
GOD AS OMNIPOTENT 193 
 
 hypothesis of a first cause involves us in hopeless 
 difficulties. 
 
 The argument conceives God's existence as having 
 no beginning in time. But this leaves two possibilities 
 open. Either God's existence is in time, and has no 
 beginning, because he has existed through unending 
 time in the past. Or else God's existence is timeless, 
 in which case of course there could be no question of 
 a beginning. 
 
 On the first alternative — that God exists in time — 
 we have a substance which has persisted through an 
 infinite past time. Now if one siubstance is admitted 
 to exist in time without being caused, why should not 
 other substances do so too? And, if any substance 
 other than God can be uncreated, then the necessity of 
 assuming the existence of a God to create them has 
 disappeared. I cannot see why it should be said, of 
 three substances existing in time, that God did not 
 need a creator, but that a man and a pebble did. If 
 God is held to be timeless, indeed, it might possibly 
 be maintained that all substances existing in time re- 
 quired a creator, while God, who was out of time, did 
 not require one. But we are considering at present 
 the hypothesis that God's existence is in time. 
 
 The universe, however, does not consist merely of 
 quiescent substances. It contains events. And the 
 argument before us says that, apart from the creation 
 of substances, God is required as the first cause of the 
 events which occur to these substances. This conten- 
 tion depends upon the principle that every event must 
 finally be derived from a cause which is not an event, 
 in order to avoid the infinite regress which would 
 ensue if every cause was an event, and therefore 
 required a cause. 
 
 But how does God cause an event to happen at 
 
194 SOME DOGMAS OF EELIGION 
 
 a particular time which did not happen before ? Is it 
 by an act of volition which occurred at a particular 
 time, and did not occur before? Then that act of 
 volition is an event, and itself requires a cause. And 
 if that cause is found in a previous event in God's 
 mind, and so on, we should get back to the unending 
 causal regress which the argument started by declaring 
 impossible. We must therefore suppose that God 
 causes changes without changing. Either he directly 
 causes an event without forming a vohtion which is an 
 event, or, if he does form such a volition, then the 
 cause of the vohtion is not an event.^ 
 
 But is this possible ? How can that which is 
 changeless be the sole cause of any event ? A change- 
 less state can, no doubt, he part of the cause of an event. 
 But it would contradict the law of Causahty to ascribe 
 an event to a cause which contained no change. For 
 in that case the cause would first exist without pro- 
 ducing the effect, and would then produce it. And this 
 change in the action of the cause would be itself an 
 event which would have occurred without a cause. 
 
 159. If we pass to the second alternative about 
 God's nature — namely, that it is timeless, it is clear 
 that it is also incapable of change. And thus we get 
 back the difficulties which we have just considered. 
 An event happens, and makes the state of the universe 
 different from what it had been before. The cause is 
 said to be God's timeless nature. That nature is the 
 same, however, before and after the event. (In itself, 
 of course, there is no question of before or after. 
 I mean that a human being who was judging of God's 
 
 * We are not yet discussing, it will be remembered, the 
 hypothesis that God's nature is timeless, but the hypothesis that, 
 while itself remaining unchanged in time, it can be the cause of 
 an event — that is, a change in time. 
 
GOD AS OMNIPOTENT 195 
 
 nature before the event, and one who was judging of 
 it after the event, would be right if they made the 
 same judgement.) Then there is nothing in that 
 nature which accounts for the change ; and it cannot 
 be the cause. If, while the so-called cause remains 
 the same, the effect varies, it is clear that the variation 
 of the effect — that is, the event — is uncaused. 
 
 As to the substances, other than God, if they are 
 held to have come into existence at a particular moment 
 of time, the event cannot be explained by a timeless 
 nature in God. If, on the other hand, they are held 
 to have existed through all past time, they have lost 
 that characteristic — their commencement — on the 
 strength of which it was asserted that a creator was 
 necessary. 
 
 It has been suggested that the series of events in 
 time will appear, to a timeless being, as a timeless 
 reality, and may thus be due to an eternal and un- 
 changing volition of that being. But, - if the true 
 nature of what appears as temporal is timeless, it is 
 not really a series of events, and therefore the law of 
 Causality does not apply to it. It needs a cause no 
 more than God himself. And thus the argument 
 breaks down. 
 
 It may perhaps be replied that these objections are 
 valid as far as they go, but that God's nature is beyond 
 our comprehension, and that in some way which we 
 do not understand he may be the first cause of changes, 
 in spite of his own changelessness, I have discussed 
 in Chapter II the validity of such appeals to believe 
 in something, which our reason tells us is impossible, 
 on the ground that our reason is certainly hmited, and 
 may be wrong. With regard to this particular argu- 
 ment, however, it may be specially noticed that it 
 requires us to be convinced, not only that we do not 
 
 2 
 
196 SOME DOGMAS OF KELIGION 
 
 know the nature of God, but that we do not know the 
 nature of a cause. The position we are discussing 
 maintains that God is changeless and a cause. Our 
 objections were directed to show that a (complete) 
 cause could not be changeless. If this is to be doubted 
 on the ground that the changelessness may be possible 
 in some way which we do not understand, then what 
 may be possible in this mysterious way is not merely 
 a changeless God but a changeless cause. This 
 means that a cause may be what our reason says it 
 cannot be. 
 
 Now, if we do not understand the nature of a cause 
 sufficiently to trust what our reason says about it, 
 the whole argument for a first cause breaks down. If 
 we are to be so sceptical about causes, we shall have 
 no right to believe that every event must have a cause, 
 or that an endless regress of causes is impossible, 
 since these conclusions rest on what our reason tells 
 us about causes. And the argument for a first cause 
 depends, as we have already seen, on the exclusion 
 of the alternatives of an uncaused change, and of an 
 endless regress of causes. 
 
 160. Thus the hypothesis of a first cause is useless 
 for the purpose of extricating us from the difficulties, 
 whatever they may be, involved in an endless causal 
 regress. If we are to avoid complete scepticism, it 
 would be necessary for philosophy to show, either 
 that such an endless regress is not impossible, or 
 else that some other alternative has been overlooked. 
 This question, however, does not concern us here, 
 since our discussion of Causality has merely related 
 to the asserted necessity of a first cause. 
 
 161. We must now consider the argument from 
 design. It cannot, I think, be better stated than in 
 the words of Kant, who takes a very sympathetic view 
 
GOD AS OMNIPOTENT 197 
 
 of this argument, although he does not accept its 
 vaUdity. Kant, however, speaks only of the wisdom 
 of the arrangement, and of the consequent wisdom 
 inferred in God. The more ordinary form of the 
 argument includes goodness among the qualities ob- 
 served in the world, and inferred in God. I have 
 therefore added goodness to Kant's statement. 
 
 'The principal points', he says 'of the physico- 
 theological proof are the following. 
 
 * First. There are everywhere in the world clear 
 indications of an intentional arrangement carried out 
 with great wisdom ' (and goodness) * and forming a 
 whole indescribably varied in its contents and infinite 
 in extent. 
 
 ' Secondly. The fitness of this arrangement is 
 entirely foreign to the things existing in the world, 
 and belongs to them contingently only ; that is, the 
 nature of different things could never spontaneously, 
 by the combination of so many means, co-operate 
 towards definite aims, if these means had not been 
 selected and arranged on purpose by a rational dis- 
 posing principle, according to certain fundamental 
 ideas. 
 
 * Thirdly. There exists, therefore, a subhme and 
 wise ' (and good) ' cause (or many), which must be the 
 cause of the world, not only as a blind and aU-powerful 
 nature, by means of unconscious fecundity^ but as an 
 intelligence, by freedom. 
 
 ' Fourthly. The unity of that cause may be inferred 
 with certainty from the unity of the reciprocal relations 
 of the parts of the world, as portions of a skilful 
 edifice, so far as our experience reaches, and beyond 
 it, with plausibility, according to the rules of analogy.'^ 
 
 * Critique of Pure Beason, first edition, p. 625. I quote from 
 Max Muller's translation. 
 
198 SOME DOGMAS OF EELIGION 
 
 162. It is to be noticed that this argument does 
 not assert, that all the part of the universe which we 
 can observe is of such a nature that it could reasonably 
 be supposed to be an end in itself for a good being. 
 The argument is that what we observe is of such 
 a nature that much of it suggests itself as being 
 suitable means to an end such as a wise and good 
 being would propose to himself ; and, further, that it 
 is of such a nature that we cannot believe it to have 
 come into its present state except as arranged by such 
 a being for the promotion of such an end. 
 
 It is clear from Kant's second paragraph that this 
 is what he means. And this is the position generally 
 adopted when the argument from design is used. It 
 is never contended that all the facts which are brought 
 forward in support of it are in every respect such 
 as God might wiU for their own sake, but that in 
 many cases they appear adapted as means to a divine 
 purpose. The eye, for example, is not treated as 
 something of intrinsic value, but as a means to sight. 
 And sight again is held to be a divine means to 
 a further purpose. 
 
 It would be impossible to base any valid argument 
 on the assertion that all we observe is worthy to be 
 itself a divine end. For such a conclusion would be 
 absolutely unsupported by the facts. Very much of 
 what we observe in the universe is of such a nature 
 that we cannot conceive that it should have intrinsic 
 worth in the sight of a wise and good being. How 
 could we suppose that such a being should find the 
 existence of one hair more or less on a man's head 
 to have intrinsic worth? Can we suppose that the 
 universe would be less good for such a being, because 
 a particular man had one hair less ? And then there 
 is much in the universe which is positively bad, and 
 
GOD AS OMNIPOTENT 199 
 
 this certainly cannot have any intrinsic worth for him. 
 A being for whom sin or misery had intrinsic worth, 
 could not be wise or good. 
 
 The force of these considerations cannot be avoided 
 by the suggestion that perhaps all existing things have 
 an intrinsic worth for a perfectly wise and good being, 
 though our minds are too feeble to perceive it in all 
 cases, and that, consequently, all things can be justified 
 otherwise than as means. In the first place, the 
 view that sin and misery can have intrinsic worth for 
 a wise and good being seems absolutely untenable. 
 There could be no justification for calling a being 
 good, who approved of sin and misery for their own 
 sakes, (To this point we shall return later.) And, in 
 the second place, if such a view were possible, it 
 would be quite irrelevant for our present purpose. 
 For the relation of the facts of experience to ideals 
 having intrinsic worth is not the conclusion of the 
 argument from design, but the premise. And it is 
 clear that we can only take as a premise the relations 
 which we know to exist, and not those which may 
 possibly exist beyond our knowledge. 
 
 Thus it is impossible to take everything as in itself 
 a divine end. On the other hand, things which 
 are intrinsically indifferent, or even intrinsically evil, 
 may have worth as means to some end other than 
 themselves. 
 
 163. It seems to me that, whatever worth the argu- 
 ment from design may have to prove the existence of 
 a God who is not omnipotent, it is quite useless as 
 a proof of the existence of an omnipotent God. If it 
 proved the existence of a God at all, it would also offer 
 a positive disproof of his omnipotence. 
 
 Kant points out that the argument could not 
 prove the existence of a creative God. As the design 
 
200 SOME DOGMAS OF EELIGION 
 
 has to be contingent to the material in which it is 
 carried out, it gives us no reason to suppose that the 
 being who carried out the design created the material. 
 * The utmost ', he says, * that could be established by 
 such a proof would be an architect of the world, always 
 very much hampered by the quality of the material 
 with which he has to work, not a creator to whose idea 
 everything is subject.' ^ 
 
 And, he goes on, we should have no reason to 
 consider the power and wisdom of such an architect 
 as infinite. We cannot logically infer to more power 
 or wisdom in the cause than is necessary to produce 
 the effect. There may be more, of course, but we 
 have no reason to believe that there is more. Now 
 the amount of the universe that we can observe, or 
 infer, to exist, and the perfection of the arrangement 
 that we see in it^ may be very great, but they are 
 certainly not infinite. It is not in our power to make 
 empirical observations which shall be infinite. Ac- 
 cordingly, we may possibly infer the existence of an 
 architect of the universe who is very much more 
 powerful and more wise that we are, but not the 
 existence in him of infinite power and wisdom. 
 
 The same, I may add, is the case with goodness. 
 Quantitatively, however, the criticism has much more 
 force with respect to goodness. If there is an architect 
 of the world, our observation of the world will tell us 
 that his power and wisdom must be enormously greater 
 than ours. But it seems to me that a being of very 
 moderate goodness might easily have designed a world 
 with as large a proportion of good in it as we are able 
 to observe in the facts around us. This point, how- 
 ever, will be treated more appropriately in the next 
 chapter. 
 
 ^ Op. cit., first edition, p. 627. 
 
GOD AS OMNIPOTENT 201 
 
 164. We may go further than this. If a wise and 
 good being has used means to an end, this is a positive 
 proof that he is not omnipotent. For means are those 
 things which have no worth in themselves, but which 
 it is right to use because, vdthout using them, some 
 end which has worth in itself cannot be attained. 
 Now there is nothing which an omnipotent God can- 
 not do — otherwise he would not be omnipotent. He 
 could get the ends without the means, if he chose to 
 do so. And therefore it would be inconsistent with 
 his wisdom to use them, since they are of no value 
 except to get an end which he could get as well with- 
 out them. In so far, therefore, as the nature of any 
 fact in the universe suggests that it owes its existence 
 to its utihty as means for a divine purpose, it suggests, 
 with just the same force, that the divine designer of 
 the universe is not omnipotent. 
 
 But this is not all. The God whose existence it is 
 attempted to prove is not only omnipotent, but good. 
 Now there are many things in the universe which are 
 not only intrinsically indifferent, but intrinsically bad. 
 Such, for example, is pain. The argument from design 
 justifies the existence of these things as means to some 
 greater good. And when we are dealing with human 
 beings, or with any being of limited power, the use 
 of such means may be completely justified. A surgeon, 
 for example, is often justified in inflicting very great 
 pain. But this is because his end — an end whose 
 good overbalances the evil of the pain — cannot be 
 attained without the use of these means. To use an 
 intrinsically evil thing as a means when the end could 
 be obtained as well without it, would deprive the 
 agent of all claim to goodness as well to wisdom. 
 And this would be the position of an omnipotent God 
 who used such means. 
 
202 SOME DOGMAS OF KELIGION 
 
 We may conclude, then, that, whatever force the 
 argument from design may have in proving the exist- 
 ence of a God of limited power, it is worse than useless 
 as a proof of the existence of an omnipotent God, 
 
 165, We need not consider here the argument that 
 the truth of Idealism would involve the existence of 
 an omniscient being, and that such a being could only 
 be God. There seems no reason to conclude that 
 omniscience implies omnipotence, and in that case the 
 argument could not prove the existence of an omni- 
 potent God, It is true that there is nothing in the 
 argument to prevent the God, whose existence should 
 be demonstrated in this way, from being omnipotent. 
 But, since omnipotence is not a quality which can be 
 assumed in any person without being demonstrated, it 
 will be better to treat this in connexion with the 
 arguments for a God who is not omnipotent. In the 
 next chapter I shall endeavour to show that it cannot 
 be used to prove the existence of a God of any nature. 
 
 166. We have now considered the chief arguments 
 put forward to prove the existence of an omnipotent 
 God. It remains to consider what objections there 
 are to our belief in his existence of a more positive 
 nature than the insufficiency of those arguments, God 
 has been defined as personal, and as good. I maintain 
 that omnipotence is incompatible with personahty, 
 and that, since the world around us is what it is, 
 omnipotence is incompatible with goodness. 
 
 An omnipotent person is one who can do anything. 
 The believers in an omnipotent God do not always 
 hold, however, that nothing happens in the universe 
 which is not in the long run completely determined 
 by God's will. For the believers in free will hold that 
 human volitions are not completely determined at all, 
 
GOD AS OMNIPOTENT 203 
 
 and, in that case, they are not completely determined 
 by God. But, if they also believe in an omnipotent 
 God, they believe that these human vohtions could 
 have been prevented by God, if he had willed to 
 prevent them. And they also believe that it is due 
 to God's will that human beings were created and 
 endowed with the power to will in this manner. If 
 this were not so, God would not be omnipotent. For, 
 if their existence was independent of him, then he 
 could not have destroyed them. 
 
 Now suppose that God had willed to create a uni- 
 verse, and had not willed that the law of Identity 
 should be vahd. It seems that we have no alterna- 
 tive but to be inconsistent or to be completely 
 unmeaning. To suppose that the universe would not 
 have been created, although God had willed that it 
 should, would be inconsistent with his omnipotence. 
 But the assertion that the universe could be created 
 without being a universe, and without being created, 
 is surely unmeaning. And yet how can the universe 
 be the universe, or creation be creation, unless the law 
 of Identity is true ? 
 
 Again, is there any meaning in the supposition that 
 God could create a man who was not a man, or that 
 he coiold create a being who was neither man nor not- 
 man ? But, if he could not, then he is bound by the 
 law of Contradiction and the law of Excluded Middle, 
 and, once more, he is not omnipotent. 
 
 The assertion that God is omnipotent imphes that 
 God could still be God, and omnipotent, whatever else 
 was the case. Any other statement might, if God 
 willed it to be so, be either true or false, and yet these 
 two statements would remain true. But these two 
 statements are not independent of the truth of all 
 other statements — we have just seen that they are nqt 
 
204 SOME DOGMAS OF RELIGION 
 
 independent of the truth of the laws of Identity, and 
 so there are statements which God cannot make false. 
 Consequently he is not omnipotent. 
 
 To take a less abstract instance, we may ask the old 
 question : Could God create a being of such a nature 
 that he could not subsequently destroy it ? Whatever 
 answer we make to this question is fatal to God's 
 omnipotence. K we say that he could not create such 
 a being, then there is something that he cannot do. If 
 we say that he can create such a being, then there is 
 still something that he cannot do — to follow such an 
 act of creation by an act of destruction. 
 
 167. Another point must be considered in reference 
 to omnipotent personality. Human personality is 
 never found to exist without a recognition of the 
 existence of something not itself. (We may follow 
 Hegel's example in calling this the Other of the 
 person. Other is a better term than Non-ego, since 
 that may suggest that what is recognized by one person 
 as not himself must not be any other person, but some- 
 thing impersonal. And this suggestion would be 
 wrong, for what I recognize as not myself may quite 
 weU be another person.) We only reahze our person- 
 ality in so far as our consciousness has a content — 
 a manifold to which the centre is formed by that 
 I, awareness of which constitutes personality. And 
 this content of consciousness involves for us the 
 recognition of an Other. This may be direct, as when 
 I know something other than myself, or have some 
 volition regarding it, or some emotion towards it. But 
 even when the Other is not involved directly, it is 
 involved indirectly. It may be that that which directly 
 occupies my consciousness is some part of my own 
 nature, as when I think of past events in my life, or 
 wiU to correct a fault in my disposition. But when 
 
GOD AS OMNIPOTENT 205 
 
 we inquire into the nature of those events, or of that 
 fault, we find that they include, or in the long run 
 involve, the recognition of the existence of an Other, 
 
 Nor is this recognition, for finite personality, a 
 limitation or imperfection, which it is impossible to 
 remove altogether, but which hampers the fullness of 
 self-consciousness. On the contrary, the more vivid, 
 definite, and extensive is our recognition of the Other, 
 the more vivid and definite becomes our self-conscious- 
 ness. As consciousness of an Other becomes vague 
 and indefinite, consciousness of self becomes vague 
 and indefinite too. As we fall asleep the two become 
 gradually faint together, and as we awake the two 
 gradually revive together. 
 
 168. An omnipotent person, if one can exist, must 
 be capable of being in a different position from this. He 
 must be capable of existing out of relation to anything 
 except himself. This does not, I think, follow from 
 the conception of a personal creator, even if the 
 universe is supposed to have been created at a 
 particular moment of time. For the creator, although 
 he would have existed before the creation as the only 
 existent reahty, might perhaps have found his Other 
 by contemplating the universe which he proposed to 
 create in the future. It is perhaps not necessary that 
 the Other to be recognized should be co-existent with 
 the recognition. 
 
 But it is necessary that he should be capable of 
 existing out of all relation to an Other, if he is to be 
 omnipotent. For, if he were not, then he could only 
 be a person on condition that a universe had arisen or 
 would some day arise. That is to say, it would be 
 impossible for him to prevent the existence, some time 
 or other, of a universe. And a person who cannot 
 prevent something from taking place is clearly not 
 
206 SOME DOGMAS OF EELIGION 
 
 omnipotent. There is perhaps no contradiction in the 
 idea of a creator who must create, but it is impossible 
 that there should be an omnipotent creator who must 
 create. 
 
 169. It may be said that reality is wider than 
 existence. The law of Contradiction and the multi- 
 plication table do not exist, but they are valid. And 
 so, it may be said, they are real. Now perhaps an 
 omnipotent God could find an Other, which was real 
 though not existent, in eternal truths which he con- 
 templated. And perhaps all that is essential for self- 
 consciousness is the recognition of a real Other, whether 
 it is existent or not. 
 
 Before this could be accepted, it would be necessary 
 to discuss a very fundamental and a very controversial 
 question in Metaphysics — how far the valid can be real 
 independently of the existent. But without entering 
 into this question, we can see, I think, that the sugges- 
 tion will not help us here. For what could the eternal 
 truths be, which the omnipotent God is to contemplate? 
 They cannot be the laws of the universe, for that 
 would imply that there was to be a universe, and so 
 limit God's omnipotence. Nothing seems to remain 
 but the possibilities to create a universe or not to 
 create it, or to create a universe of one kind or a 
 universe of another. That all this should be possible, 
 it may be said, does not depend on God's willing it to 
 be possible — on the contrary, the possibilities constitute 
 his omnipotence. And so his omnipotence is not 
 limited if we say that he inevitably recognizes these 
 possibilities as real. 
 
 But when we look closer we see that for an omni- 
 potent God, just because he is omnipotent, there are 
 no possibilities to recognize. For me, who am not 
 omnipotent, there are definite possibilities, and these 
 
GOD AS OMNIPOTENT 207 
 
 possibilities are real, and are not myself. I can draw 
 a triangle or not draw one. If I draw it, I can give it 
 a right angle or not. But the existence of these 
 definite possibilities for me to recognize depends on 
 the fact that there is something, other than my own 
 will, which prevents my doing other things. I cannot 
 both draw a triangle and abstain from drawing it. 
 Nor can I draw a triangle which is courageous, or 
 which has two right angles. 
 
 Now with an omnipotent God it would be different. 
 Nothing is impossible to him, and therefore there are 
 no definite possibilities which he recognizes because 
 his wiU must submit to their limitations. If we ask 
 what exists outside an omnipotent God, the answer — 
 not only about existence but about all reality — must 
 be : What he wills. If we ask what exists outside an 
 omnipotent God, independently of his will, the answer 
 must be : Nothing whatever.^ There are not even 
 possibilities. There is nothing at all. 
 
 Since an omnipotent God, then, could exist even if 
 he willed that nothing else should exist and that no 
 laws should be true, and since his omnipotence leaves 
 no possibilities outside him, but only nonentity, it 
 follows that he could exist as the only reality, and 
 consequently that he could exist without recognizing 
 any other reality. His knowledge might be only of 
 his own qualities and states, his volitions might con- 
 cern nothing but those quaHties and states, his emotions 
 be directed towards himself only. And from these 
 
 ^ This is tiie case, even if we suppose, with the indeterminists, 
 that human volitions, not being completely determined by any- 
 thing, are not completely determined by God's will. For our 
 wills are held to act in this way only because God has endowed 
 them with the power to do so, and because he does not withdraw 
 it. The volitions, although not completely determined by God's 
 will, are still completely dependent on it, since they only take 
 place because he wills them to do so. 
 
208 SOME DOGMAS OF EELIGION 
 
 qualities and states must be deducted all those that are 
 determined by his will— for if these were essential to 
 him, his will would be under limitations, since he 
 could not dispense with them. 
 
 170. It would be too much to say that this gives us 
 a second proof of the impossibility of an omnipotent 
 person. Self-consciousness is not the same as con- 
 sciousness of an Other, and we are not entitled to 
 assert that they must inevitably be found together. 
 But we have no knowledge of the first without the 
 second, and it seems rather unwise to build on the 
 abstract possibility that a being could exist, beyond 
 the Hmits of our experience, who could have the first, 
 without having the second. 
 
 And this will appear still more imprudent when we 
 consider that, as was said above, the recognition of an 
 Other is, for human beings, not a hindrance from 
 which they cannot escape, but the condition of their 
 development. If an omnipotent person can exist it 
 follows that the recognition of the Other, while it 
 becomes more and more definite and extensive as 
 personality develops within our experience, becomes 
 quite unessential to the most developed personahty 
 of all. 
 
 171. We now come to the relation of omnipotence 
 to goodness. There is evil in the universe. It is not 
 necessary to inquire how great or how small the 
 amount of evil may be. All that is important for the 
 present discussion is that there is some evil, and this 
 is beyond doubt. A single pang of toothache, a single 
 ungenerous thought, in the midst of a universe other- 
 wise perfectly good, would prove the existence of evil. 
 
 The existence of evil is beyond doubt in the sense 
 that no one denies the existence of pain and sin in 
 experience, and that no one denies that pain and sin 
 
GOD AS OMNIPOTENT 209 
 
 are, from the point of view of ordinary life, to be con- 
 sidered evil. But it has been asserted that the universe, 
 when looked at rightly, may be completely good. 
 Sometimes the standard is challenged, and it is sug- 
 gested that pain and sin are really good, though we 
 think them evil. Sometimes our comprehension of 
 the facts is challenged ; it is admitted that pain and 
 sin, if they existed, would be bad, but it is maintained 
 that they do not really exist. 
 
 The first of these alternatives means complete 
 ethical scepticism. There is no judgement about the 
 good of whose truth we are more certain than the 
 judgement that what is painful or sinful cannot be 
 perfectly good. If we distrust this judgement, we 
 have no reason to put any trust in any judgement of 
 good or evil. In that case we should have no right to 
 call anything or anybody good, and therefore it wotild 
 be impossible to justify any belief in God, whose defi- 
 nition includes goodness. This objection, therefore, 
 cannot consistently be used, by the beHevers in an 
 omnipotent God, against the existence of evil. 
 
 The second alternative is one which can only be 
 supported by metaphysical arguments of a somewhat 
 abstruse and elaborate nature. To expomid and 
 examine these arguments in detail would take us too 
 far from our subject, I will only say briefly that the 
 theory of the unreality of evil now seems to me 
 untenable. Supposing that it could be proved that 
 all that we think evil was in reality good, the fact 
 would still remain that we think it evil. This may 
 be called a delusion or a mistake. But a delusion or 
 mistake is as real as anything else. A savage's errone- 
 ous belief that the earth is stationary is just as real 
 a fact as an astronomer's correct belief that it moves. 
 The delusion that evil exists, then, is real. But then. 
 
210 SOME DOGMAS OF EELIGION 
 
 to me at least, it seems certain that a delusion or an 
 error which hid from us the goodness of the universe 
 would itself be evil. And so there would be real 
 evil after all. If, again, the existence of the delusion 
 is pronounced to be a delusion, then this second de- 
 lusion, which would be admitted to be real, must be 
 pronounced evil, since it is now this delusion which 
 deceives us about the true natm-e of reality, and hides 
 its goodness from us. And so on indefinitely. How- 
 ever many times we pronounce evil unreal, we always 
 leave a reality behind, which in its turn is to be 
 pronounced evil. 
 
 172. An omnipotent God is conceived as creating 
 the universe. In that case it seems a natural inference 
 that he is the cause of all the evil in the universe. 
 But some people, who maintain the existence of a 
 creative omnipotent God, maintain also that the choice 
 of the human will between motives has no cause, and, 
 therefore, is not ultimately caused by the creator. 
 They admit, however, that God could have dispensed 
 with the freedom of the human will, if he had chosen 
 to do so.^ 
 
 We may therefore say that an omnipotent God 
 could have prevented aU the evil in the universe if he 
 had willed to do so. It is impossible to deny this, if 
 omnipotence is to have any meaning, for to deny it 
 would be to assert that there was something that God 
 could not do if he willed to do it. 
 
 173. What bearing has this on the question of God's 
 goodness? It is clear that man may act rightly in 
 permitting evil, and even in directly causing it. It is 
 
 ^ It seems curious that believers in human free will should 
 often accept the argument for God's existence from the necessity 
 of a first cause. If human volition is not completely determined, 
 the law of causality is not universally valid. And, in that case, 
 what force remains in the argument for a first cause ? 
 
GOD AS OMNIPOTENT 211 
 
 evil that a child should lose his leg, for the loss de- 
 prives him of much happiness, and causes him much 
 pain. But the surgeon who performs the operation, 
 and the parent who allows it to be performed, may be 
 perfectly justified. For amputation may be the only 
 alternative to evils much greater than those it produces. 
 
 And, again, the production of sin may under certain 
 circumstances be justified. Supposing that it were 
 true — fortunately there is no reason to believe that it 
 is true — that employment as an executioner tended to 
 degrade morally a large proportion of those who were 
 employed, it would by no means follow that men 
 ought not to be induced to act as executioners. The 
 evil results which might follow from having no hang- 
 man might far outweigh the evil done to morality by 
 having one. 
 
 But the justification in these cases depends entirely 
 on the limited powers of the agents. The father and 
 the surgeon, for example, are justified because it is 
 only through the evils of amputation that worse evils 
 can be avoided. If they could have avoided those 
 worse evils by some other course that would not have 
 been evil at all, they would not have been justified in 
 deciding on the amputation. 
 
 174. Now the power of an omnipotent God is not 
 limited. He can effect whatever he wills. If he wiUs 
 to have A without B, he can have A without B, however 
 closely A and B may be connected in the present 
 scheme of the universe. For that scheme also is de- 
 pendent on his will. It thus appears that his action 
 cannot be justified as 'the amputation was. It rather 
 resembles that of a father who should first gratuitously 
 break his son's leg, or permit it to be broken, and 
 should then decide for amputation, although a complete 
 cure was possible. 
 
 P 2 
 
212 SOME DOGMAS OF EELIGION 
 
 If a man did this we should call him wicked. We 
 do not wait to call a man wicked till he does more 
 evil than good. If a man should, at the risk of his 
 life, save all the crew of a sinking ship but one, and 
 should then, from mere caprice, leave that man to 
 sink, whom he could easily have saved, we should 
 say that he had acted wickedly. Nor is it necessary 
 that a man should do evil for the sake of evil. To 
 desire to attend a concert is not a desire for evil as 
 such, but if I kiUed a man in order to acquire his 
 ticket, I should have acted wickedly. 
 
 Now in what way would the conduct of an omnipo- 
 tent God, who permitted the existence of evil, differ 
 from the conduct of such men, except for the worse ? 
 There are palliations of men's guilt, but what pallia- 
 tions could there be for such a God? A man may 
 have lived a long life of virtue before he fell into sin, 
 or, again, we may have reason to hope that he will 
 repent and amend. But could we have any reason for 
 hoping that the omnipotent God would repent and 
 amend? It seems difficult to imagine such a reason. 
 Again, a man may be excused to some extent for his 
 sin, if ignorance or folly prevents him from realizing 
 the full meaning of his action. But if an omnipotent 
 God is not omniscient (and it seems most natural to 
 suppose that he is), at any rate he could be so if he 
 chose. Or again, a man may have a genuine repug- 
 nance for his sin, and only commit it under extreme 
 temptation. A man who betrays his country under 
 torture is less wicked than if he had betrayed it 
 for money. But an omnipotent God can be forced to 
 nothing, and can therefore not be forced to choose 
 between wickedness and suffering. 
 
 175. Such conduct, then, as we must attribute to an 
 omnipotent God, would be called wicked in men, 
 
GOD AS OMNIPOTENT 213 
 
 although the amount of evil for which any man is 
 responsible is insignificant as compared with the sum 
 of all evilj and although men have in most cases ex- 
 cuses which would not apply to an omnipotent God. 
 Yet this being is still called God, by people who admit 
 that goodness is part of the definition of God. Why 
 is God called good, when his action is asserted to be 
 such as would prove a man to be a monster of wicked- 
 ness? Two lines of defence have been tried. The first is, 
 in substance, that the omnipotent and good God is not 
 really good, the second that he is not really omnipotent. 
 
 The form in which the first is put by its supporters 
 is that goodness in God is of a different nature from 
 what it is in man. Thus Mansel says that ' the inflic- 
 tion of physical suffering, the permission of moral evil ' 
 and various other things ' are facts which no doubt are 
 reconcilable, we know not how, with the infinite Good- 
 ness of God, but which certainly are not to be explained 
 on the supposition that its sole and sufficient type is 
 to be found in the finite goodness of man '.^ And he 
 goes on to say that the difference is not one of degree 
 only, but of kind. Pascal is still more plain-spoken : 
 ' What can be more opposed to our wretched rules of 
 justice than the eternal damnation of a child without 
 any will of its own for a sin in which it seems to have 
 had so little share that it was committed six thousand 
 years before the said child came into existence.' ^ 
 Nevertheless Pascal continued to call good, and to 
 worship, a God whom he believed to have done this. 
 
 176. But why should the word good be used in two 
 senses absolutely opposed to one another ? The senses 
 are not merely different, as they would be if, for example, 
 it was proposed to use the word good to indicate what 
 
 * Limits ofEeligious TJiouglit, fourth edition, Preface, p. xiii. 
 2 Works, ed. Brunschvicg, ii, p. 348. 
 
2U SOME DOGMAS OF RELIGION 
 
 is generally meant by the word scarlet. For what 
 is called good in God would be called wicked in men, 
 and good and wicked are predicates directly contrary 
 to one another. 
 
 Is the alteration to be considered as one of mere 
 caprice ? Are the people who say that God is good, 
 while Nero was wicked, in the same position as a man 
 who should call Everest a valley, while he called 
 Snowdon a mountain ? It seems to me that there is 
 more than this involved, and that the real ground of 
 the alteration is that good is a word of praise, and that 
 wicked is a word of blame, and that it is felt to be 
 desirable to praise God rather than to blame him. 
 
 But why is it desirable to praise him ? Certainly 
 not for the reasons which make us praise Socrates and 
 blame Nero. For the conduct which in God we call 
 good is conduct for some faint and imperfect approxi- 
 mation to which we blame Nero. What other reason 
 is left ? I can only see one — that an omnipotent God 
 is, and will remain, infinitely more powerful than Nero 
 ever was. 
 
 On this subject MiU has spoken,^ and it is unneces- 
 sary to quote words which form one of the great turn- 
 ing-points in the religious development of the world. 
 Yet when Mill says that rather than worship such 
 a God he would go to hell, it is possible to raise a 
 doubt. To call such a being good, and to worship him, 
 is to lie and to be degraded. But it is not certain that 
 nothing could be a greater evil than to lie and to be 
 degraded. It is not impossible that God's goodness, as 
 explained by Pascal and Mansel, should include the 
 infliction of such tortures, physical and mental, on one 
 who refused to worship him that they would be a 
 greater evil than lying and degradation. Unless it is 
 ^ Examination ofHamilton^ chap. vii. 
 
GOD AS OMNIPOTENT 215 
 
 said that moral degradation is absolutely incommensur- 
 able with suffering — and I doubt if this can be main- 
 tained — the case does not seem impossible. Nor need 
 the motive of the worshipper be selfish. The goodness 
 of God, like the wickedness of some men, might include 
 the torture of the culprit's friends as well as of himself. 
 
 177. We may doubt, then, whether we should be 
 bound, or justified, in refusing to misapply the predi- 
 cate good to such an omnipotent being, if the use of 
 the word would diminish our chances of unending 
 torture. But it seems just as likely to increase them. 
 There are, no doubt, men who are prepared to inflict 
 suffering on all who do not flatter them, even when they 
 know that the flattery is empty and undeserved. But, 
 granted that God has some quahties which would be 
 called wicked in men, it does not follow that he has all 
 qualitieswhich would be called wicked in men, and there 
 is no reason tosuppose that he has this particularquality. 
 Many men, bad as weU as good, are not appeased by 
 such flattery, but rather irritated by it, especially if 
 they know it to be insincere, or to have been insincere 
 when it began. God may resemble these men rather 
 than the others. Indeed, the probabUity seems to be 
 that he would do so, since pleasure in such flattery is 
 generally a mark of a weak intellect, and even if God's 
 goodness is like our wickedness, it can scarcely be 
 suggested that his wisdom is hke our foUy. Or, again, 
 God's goodness may induce him to damn us whatever 
 we do, in which case we shall gain nothing by lying. 
 
 When everything is so doubtful there does not seem 
 to be the least prudence in flattery. Nor can we rest 
 our actions on any statement made by God as to the 
 conduct which he will pursue. For, if goodness in 
 God is different from goodness in us, we should have 
 no reason to believe a statement to be true rather than 
 
216 SOME DOGMAS OF RELIGION 
 
 false, even if it were certain that it came from God. 
 Divine goodness may not exclude the desire to destroy 
 our happiness by false statements. 
 
 178. There remains the attempt to save the goodness 
 of an omnipotent God by giving up the reality of his 
 omnipotence, while retaining the name. Various 
 elements in the universe have been taken either as 
 good or as inevitable, and the evil in the universe 
 explained as the necessary consequence of the reality 
 of these elements. Thus, for example, the sin of the 
 universe has been accounted for by the free will of the 
 sinners, and the suffering explained as the necessary 
 consequence, in some way, of the sin. Thus all the 
 evil in the universe, it is asserted, is a necessary 
 consequence of free will, and it is said that free will 
 is so good that God was justified in choosing a 
 universe with all the present evil in it, rather than 
 surrender free will. 
 
 Or, again, it is said that it is impossible that there 
 should not be some evil in a universe which was 
 governed according to general laws, and that to be 
 governed according to general laws is so great a 
 perfection in the universe that God did well to choose 
 it with all the evil that it involves. 
 
 It seems to me rather difficult to see such supreme 
 value in free will that it would be worth more than 
 the absence of all the present evil of the universe. 
 It might be doubted, even, whether the advantage of 
 unbroken general laws is so great that the evil of the 
 universe would not be cheaply removed at the cost 
 of frequent miracles \ But we need not discuss this. 
 
 ^ The supporters of this view of the supreme value of unbroken 
 general laws, have often, it may be noted, been men who believed 
 that God did well in permitting the human will to be undeter- 
 mined, and in working occasional miracles. 
 
GOD AS OMNIPOTENT 217 
 
 For it is quite evident that a God who cannot create 
 a universe in which all men have free will, and which 
 is at the same time free from all evil, is not an omni- 
 potent God, since there is one thing which he cannot 
 do. In the same way, a God who cannot ordain a 
 series of general laws, the uniform working of which 
 would exclude all evil from the universe, is not an 
 omnipotent God. 
 
 Or, once more, it is said that a universe without 
 evil would involve in some way the violation of such 
 laws as the law of Contradiction or of Excluded 
 Middle, and that these laws are so fundamental that 
 the existence of evil in the universe is inevitable. 
 
 Even if there were any ground for believing that 
 the absence of evil from the universe would violate 
 such laws as these, it is clear that a God who is bound 
 by any laws is not omnipotent, since he cannot alter 
 them. If it is said — as it may very reasonably be said 
 — that these laws are so fundamental that it is un- 
 meaning to speak of a being who is not bound by 
 them, the proper conclusion is not that an omnipotent 
 God is bound by them, but that, if there is a God, he 
 is not omnipotent. 
 
 It is necessary to emphasize this point because, 
 remarkable as it may appear, it is not an unusual 
 position to maintain that God is absolutely omnipotent, 
 and, at the same time, to believe that there are certain 
 things he cannot do, and even to be quite certain what 
 those things are,^ As against such a view as this it 
 seems necessary to emphasize the tolerably obvious 
 fact that, if there is anything which God could not do 
 if he wished, he is not omnipotent. 
 
 ^ Cp., for example, Flint's Theism. The omnipotence of God is 
 asserted : I. 2, III. 1, IX. 2. In VL 1 and VIII. 1 we find state- 
 ments of some of the things which God cannot do. 
 
218 SOME DOGMAS OF RELIGION 
 
 179. It may be said that we are attaching too much 
 importance to a slight inaccuracy of language. If 
 people say that there are certain things which God 
 could not do, then they do not believe him to be 
 omnipotent, and they are simply using the wrong 
 word when they say that they do believe him to be 
 omnipotent. 
 
 But then why do they use the word ? It seems to 
 me that the confusion of language covers a confusion 
 of thought. Many people are unwilling to accept the 
 idea that God is not omnipotent. It is held to detract 
 from his perfection, and to render it difficult to regard 
 him as the creator of the universe. 
 
 And there is another point of grave importance. 
 If God is not omnipotent, the fact that God exists 
 and is good gives us no guarantee that the universe 
 is more good than bad, or even that it is not very 
 bad. If God exists and is good, the universe will of 
 course be as good as he can make it. But, if there 
 are some things that he cannot do, how can we tell 
 that among these impossibilities may not be the 
 impossibility of preventing the world from being more 
 bad than good, or of preventing it from being very 
 bad ? If it could be shown that God's power, though 
 limited, was strong enough to prevent this, it could 
 only be by a determination of the precise limits of 
 his power, and, if this could be done at all, it could 
 only be done by an elaborate metaphysical investiga- 
 tion. Such investigations are open to few, and their 
 results are frequently highly controversial. It is not 
 strange that popular theology is unwilling to accept 
 the position that the goodness of the universe can 
 only be proved in such a way. 
 
 And thus popular theology has two conflicting im- 
 pulses. It desires, among other things, to show that 
 
GOD AS OMNIPOTENT 219 
 
 the universe is more good than bad — at any rate in 
 the long run. The only means at its disposal for 
 showing this — if it is to remain popular — is its behef 
 in the existence of a God to whose will all evil is 
 repugnant, and who is powerful enough to effect the 
 predominance of good. But if God is to be taken as 
 onmipotent, it is certain that all evil is not repugnant 
 to his will, and if he is to be taken as not omnipotent, 
 it is not certain that he is powerful enough to effect 
 the predominance of good. 
 
 The inaccurate use of the word omnipotence hides 
 this dilemma. When popular theology is pressed to 
 reconcile the present existence of evil with the good- 
 ness of God, then it pleads that omnipotent does not 
 mean omnipotent, but only very powerful. But when 
 the sceptic has been crushed, and what is wanted is 
 a belief in the future extinction of evil, then omni- 
 potence slides back into its strict meaning, and it 
 is triumphantly asserted that the cause which has 
 an omnipotent God on its side must certainly win. 
 The confusion is unintentional, no doubt, but it is 
 dangerous. 
 
 It seems to me that when believers in God save his 
 goodness by saying that he is not really omnipotent, 
 they are taking the best course open to them, since 
 both the personality and the goodness of God present 
 much fewer diflBculties if he is not conceived as omni- 
 potent. But then they must accept the consequences 
 of their choice, and realize that the efforts of a non- 
 omnipotent God in favour of good may, for anything 
 they have yet shown, be doomed to almost total defeat. 
 It is not a very cheerful creed, unless it can be supple- 
 mented by some other dogmas which can assure us 
 of God's eventual victory. But it is less depressing 
 and less revolting than the belief that the destinies 
 
220 SOME DOGMAS OF EELIGION 
 
 of the universe are at the mercy of a being who, with 
 the resources of omnipotence at his disposal, decided 
 to make a universe no better than this. 
 
 180. In this chapter I have only discussed the view 
 which makes God a being separate from the universe, 
 though all-powerful over it. I have not thought it 
 necessary to consider the view — which has been main- 
 tained — according to which the whole of the universe 
 is one omnipotent person, so that God is the sole 
 reality, and we are not his creatures, but parts of 
 himself. This view has not had much influence on 
 religious thought. (Pantheism, indeed, has been and 
 is very powerful, but the prevailing type is that which 
 denies personality to the unity of all things.) And 
 it is so little known, and of so obscure a nature, that 
 I could scarcely have done it justice without expound- 
 ing the whole philosophical system of each of its 
 principal adherents. It is clear that the arguments 
 given above to prove the inconsistency of omnipotence 
 with personahty and goodness would apply to God 
 equally, if this theory of his nature were true. In 
 addition to these there would be the further difficulty 
 that it would require us to regard ourselves as parts 
 of God. The belief that one person could be part of 
 another would, I think, be found very difficult to 
 defend. 
 
CHAPTEK VII 
 
 A NON-OMNIPOTENT GOD 
 
 181. We must now proceed to consider the theory 
 that God exists without being omnipotent, so that 
 there are results which his volitions could not produce. 
 On this hypothesis it is possible to hold either that 
 God created all existent reality except himself, or that 
 the other existent reality exists co-eternally with God, 
 and independent of his wiU. In the latter case God 
 would act on his environment in the same way that 
 men act on theirs. His claim to be considered as 
 God would rest on the very great excess of his good- 
 ness and power over that of all other beings.^ 
 
 What grounds have we for believing in the existence 
 of a non-omnipotent God ? His existence is sometimes 
 maintained, as in the case of an omnipotent God, on 
 the groxmd of an instinctive and irresistible conviction, 
 on the ground of the disastrous consequences which 
 would result from his non-existence, or on the evidence 
 of miracles. These grounds of behef were discussed 
 in Chapter II, and need not be considered here. 
 
 It will also be unnecessary to consider here the 
 Ontological argument, both for the reasons mentioned 
 in the last chapter, and also because it has always, so 
 
 ^ The belief in a creative God who is not omnipotent has been 
 recently urged with great force by Dr. Eashdall in his essay on 
 Personality Human and Divine in the collection entitled Personal 
 Idealism, The conception of a non-creative God has been 
 brilliantly expounded and defended in The Limits of Evolution 
 by Dr. Howison. 
 
222 SOME DOGMAS OF RELIGION 
 
 far as I know, been employed to prove the existence 
 of a God who is omnipotent. 
 
 The objections brought in the last chapter against 
 the validity of the argument for the necessity of a first 
 cause, will apply here also, since none of them 
 depended on the omnipotence of the first cause. In 
 the case of a non-creative God, indeed, there will 
 also be additional objections to the argument. A first 
 cause was said to be required, both for non-divine 
 substances and for events. But the hypothesis of 
 a non-creative God implies that non-divine substances 
 can exist eternally and in their own right, so that no 
 creator is required for them. And, as for events, the 
 argument for a first cause asserts that it is a sufficient 
 explanation of them to refer them to an eternal and 
 uncreated being for their first cause. But if this is so, 
 then the existence of a God is not required to explain 
 them, for, by the hypothesis, the non-divine beings 
 are also eternal and uncreated, and therefore could be 
 the causes required. 
 
 The position of the argument from design is different. 
 The objections made to that argimient in the last 
 chapter rested on the fact that its truth would be 
 compatible with the existence of a God who was not 
 omnipotent, and, indeed, incompatible with the ex- 
 istence of a God who was omnipotent. Such an 
 objection cannot, of course, apply to the attempt to 
 prove, by this argument, the existence of a non-omni- 
 potent God, But this argument will not give us any 
 help in determining whether God is or is not creative. 
 A creative God without unlimited power might find it 
 necessary to use means to carry out his ends, since the 
 limitations on his power might preclude the possibiUty 
 of attaining his ends directly. In this case he would 
 produce a universe such as is asserted to exist by the 
 
A NON-OMKIPOTENT GOD 223 
 
 argument from design. On the other hand it is quite 
 possible that the universe might have been organized 
 as means for an end by a being who did not create the 
 reahty which he dealt with, but arranged and modified 
 it as a watchmaker arranges and modifies the metal 
 which he cannot create. 
 
 In the same way the argument that a God is 
 necessary because all reaUty must be known to some 
 being tells us nothing as to whether that God is a 
 creator. Such an omniscient being might also be the 
 creator of all the existent reality, other than himself, 
 which he knows. But we can know things which we 
 have not created, and, if there is an omniscient being 
 at all, there seems no reason why he should not know 
 various existent realities which do not owe their 
 existence to his will, 
 
 182, Since these arguments give us no help in 
 deciding between the two alternative forms of a belief 
 in a non-omnipotent God, let us consider whether 
 there are any other reasons which should lead us to 
 prefer either alternative to the other. I shall endea- 
 vour to show that the hypothesis of a creative God, 
 even when he is conceived as non-omnipotent, labours 
 under serious difficulties which do not affect the 
 hj^othesis of a non-creative God. 
 
 Before considering these difficulties, however, we 
 must make a further distinction. A creative, but 
 non-omnipotent, God may be conceived either as free 
 not to create, or as not free not to create. It may be 
 held, that is, that the limitations of his power left 
 him free to remain permanently the sole reality of the 
 universe, should he choose to do so. In this case the 
 limitation of his power would consist in the fact that, 
 if he did choose to create, he could not create any- 
 
224 SOME DOGMAS OF EELIGION 
 
 thing that he chose, but had only the choice between 
 certain alternatives. Or it may be held that he had 
 no power to abstain from creating reality other than 
 himself, although he had power to decide, within 
 certain limits, what sort of reality he would create. 
 The second objection to God's personality mentioned 
 in the last chapter would also apply to a God who was 
 free to abstain from creating,^ But it need not apply 
 to a God who was compelled to create, since the 
 created universe might afford him his necessary 
 Other. 
 
 183. The difficulties which, as I have said, seem to 
 me to arise from the hypothesis that the non-omni- 
 potent God is creative, arise from the goodness which 
 we decided to be part of the definition of God. We 
 saw in the last chapter that it was impossible that an 
 omnipotent being should be good, in view of the fact 
 that there is evil in the universe. The behevers in the 
 hmitation of God's power assert that they have saved 
 the possibihty of his goodness, because it is possible 
 that a non-omnipotent God might wish to make the 
 universe much better than it is, and yet be unable to 
 do so. That this is possible with a non-creative God, 
 is, I think, beyond doubt. But I am not so clear that 
 it is possible with a creative God. It is quite possible, 
 no doubt, that there are some things that the creator 
 of the universe— if there is a creator — cannot do. But 
 I cannot satisfy myself that it is possible that there 
 could be anything which he willed to do, and which 
 yet he could not do. And he would not be good — and 
 consequently would not be God — unless he did will to 
 remove the evil which he could not remove. 
 
 ^ Sections 167-70. These objections, however, would not 
 apply to a non-omnipotent God, if it were held — as some philo- 
 sophers hold — that it is possible for a truth to be real without 
 reference to any existence. 
 
A NON-OMNIPOTENT GOD 225 
 
 184. In one sense, indeed, it is clear that nothing 
 exists except as God wills it, if God is creative. For 
 nothing exists unless he decides to create it — unless, 
 that is, he prefers its existence to its non-existence. 
 But we often will to produce a result, and yet say that 
 we do so unwillingly, meaning that the result is the 
 one most desirable under the circumstances, but is at 
 the same time one which, in many respects, is unde- 
 sirable. 
 
 The old distinction between antecedent and con- 
 sequent volition is useful here. We will antecedently 
 that which we desire in all respects. We will con- 
 sequently that which, under the circumstances, we 
 prefer to any other alternative which the circumstances 
 leave possible. The two would be coincident for an 
 omnipotent being, but for others they may diverge, 
 and very frequently do so. A good man does not 
 antecedently will to produce anything which is not 
 good. But he often wills consequently to produce 
 much that is bad. He wiUs, for example, to inflict 
 pain in order to avoid worse evils. 
 
 With the exception of the undetermined volitions 
 of non-divine beings — if such undetermined vohtions 
 exist — everything in the universe must be held to be 
 caused by God, if God is conceived as creative.^ And 
 what is caused by God must be in accordance with his 
 consequent will. But need it be in accordance with 
 his antecedent wiU ? 
 
 The will of any being is said to be thwarted in so 
 far as he cannot act in conformity with his antecedent 
 will. If I wish to enter a certain field, and abstain 
 from doing so because there is a savage bull in it, my 
 
 ^ It would be due to the will of God, also, that any beings did 
 form undetermined volitions, though the character of the volitions 
 they formed would be due to nothing. 
 
 Q 
 
226 SOME DOGMAS OP EELIGION 
 
 action is in accordance with my consequent will. The 
 bull cannot prevent me from entering the field, and 
 I only abstain from entering because I prefer being 
 disappointed to running a serious risk of death. But 
 my will has, nevertheless, been thwarted. I did not 
 will to enter the field at great risk, but I did will to 
 enter the field. It is this latter volition, the ante- 
 cedent volition, which has been thwarted. 
 
 185. If the evil of the universe thwarts God's 
 antecedent will, the fact that it has all arisen by his 
 consequent will is no more reflection on his goodness, 
 than the pain caused by amputation is on the 
 benevolence of the surgeon. But it is essential that 
 it should thwart his antecedent will, since God has 
 been defined as good. God must feel the same im- 
 potent hatred of the evil which he cannot remove as 
 is felt by a good man. Indeed, he must feel more. 
 For whether God is conceived as perfectly good or 
 not, he is always conceived as appreciably better 
 than the best man. From this it would follow that 
 evil must be more repugnant to him than it is to the 
 best man. 
 
 This does not involve that God is more dissatisfied 
 with the state of the universe than a good man would 
 be. If God is much better than man, he must not 
 only revolt more intensely against the evil, but he 
 must rejoice more intensely in the good. Again, it is 
 possible that a divine knowledge, which embraced the 
 whole universe, might be able to see that the proportion 
 of good to evil in the universe was much greater than 
 it appears to us who see such a small corner of the 
 whole. And thus it is possible, though only possible, 
 that God's satisfaction might enormously outweigh his 
 dissatisfaction. But, however much the good in the 
 universe may exceed the evil, there is evil, and this 
 
A NON-OMNIPOTENT GOD 227 
 
 must be more repugnant to God than it can be to 
 any one else. 
 
 The only way to escape from this conclusion would 
 be to deny that any evil did exist in the universe, or to 
 maintain that God's goodness was so different from 
 ours that his antecedent will could be in harmony with 
 a state of things with which the antecedent will of 
 a good man could not be in harmony. These sugges- 
 tions were considered in the last chapter. 
 
 If, indeed, we were to say that the universe was 
 created by a being who was not good, it would not 
 follow from the existence of evil in the universe that 
 the will of such a being must be thwarted. Such a 
 being, however, would not be a God, of whose defi- 
 nition, as we have seen, goodness is a part. Nor 
 would the discussion of such a view be of any practical 
 interest, since, as far as I know, the theory of a 
 personal creator who is not good is not one which 
 commands any support. It has been maintained that 
 the creator's goodness was of a different nature to 
 man's, but goodness has always been attributed to him. 
 
 186. Why is it that any person wills consequently 
 that which is contrary to his antecedent will ? It is 
 because things are so connected that, if he did not 
 will this consequently, he would produce a result still 
 more repugnant to his antecedent will. I remain out- 
 side the field, in the previous instance, although it is 
 repugnant to my antecedent will to do so, because 
 otherwise I should put my life in great peril, and 
 because that is more repugnant to my antecedent will 
 than staying outside the field would be. In the same 
 way, we are told, God creates evil things, which is an 
 act repugnant to his antecedent volition, because if 
 he did not do so, he would have to create stiU worse 
 things, or abstain from creating things with an amount 
 
 Q 2 
 
228 SOME DOGMAS OF KELIGION 
 
 of good which more than outweighs the evil. And 
 either of these courses would be still more repugnant 
 to his antecedent volition than the course which he 
 actually takes. 
 
 In every case where the consequent volition has an 
 element repugnant to antecedent volition the explana- 
 tion is the limited power of the agent. If I could 
 pacify the bull by a look, I should enter the field safely, 
 and thus gratify both my antecedent volitions. If 
 God could have abstained from creating the evil he 
 did create, and at the same time abstained from ci^eat- 
 ing any greater evil, and created all the good he did 
 create, he would have created a universe which was all 
 good and no evil, and was therefore in complete har- 
 mony with his antecedent volition. 
 
 There is often considerable unwillingness among 
 theists, as was said at the end of the last chapter, to 
 admit that God is not omnipotent. But when that 
 has once been admitted, it seems to be generally taken 
 as a matter of course, that God might will consequently 
 to create, and so create, that which was repugnant to 
 his antecedent will. We have continual examples in 
 daily life of men willing consequently what is repugnant 
 to their antecedent will. It seems to be thought that 
 so soon as we are in a position to say that God is like 
 man in not being omnipotent, we are entitled to beheve 
 that he may act repugnantly to his antecedent volition. 
 
 187. But there is another very important difference 
 between man and a creative God, even if that God is 
 not omnipotent. No characteristic, and therefore no 
 impotence, of a man can be explained entirely from 
 his own nature. He is part of a universe, all the parts 
 of which are connected. Nothing can happen in the 
 rest of the universe which does not affect him, nothing 
 can happen in him which does not affect the rest of the 
 
A NON-OMNIPOTENT GOD 229 
 
 universe. And, therefore, when a man -wills to do 
 something and cannot do it, his impotence is never 
 due entirely to his own nature. 
 
 This is clearest when the thing which I wish to do, 
 and cannot do, is to produce some effect on some being 
 other than myself. Then it is clear that success or 
 failure depends not only on my own nature, but also on 
 the nature of the being I am trying to affect. In 
 ordinary language it would sometimes be said that 
 a failure was due entirely to my own nature, if the 
 circumstances were such that an average person, in my 
 place, would have succeeded. But it is obvious that 
 this leaves out an essential element. However feeble 
 my effort, its failure cannot be explained merely from 
 its feebleness. For we cannot know that it was too 
 feeble to succeed, until we know the nature of the 
 being it was attempting to influence. 
 
 But the proposition is also true when the effect 
 which I vainly endeavour to produce is a change in 
 my own nature. My nature, in all its characteristics, 
 is dependent, not on itself only, but on the rest of the 
 universe. How far it will be changed, in any case, by 
 the influence of my will to change it, will depend on 
 the result produced by my wiU in combination with all 
 the other factors which are influencing it. If my will 
 fails to change it, the failure cannot be explained by my 
 nature alone, but will depend also on the strength of 
 the other factors which opposed the change which was 
 willed, and on the feebleness of the other factors 
 which supported the change. 
 
 188. With a creative God the matter is different. At 
 the moment at which he creates nothing exists except 
 himself. He is the only existent reality. Whatever 
 happens must be explained from his nature, and from 
 his natm-e alone, for there is nothing else anywhere. 
 
230 SOME DOGMAS OF EELIGION 
 
 It is from his own nature, then, that we must ex- 
 plain the limitation of his power, which, according to 
 the theory we are considering, accounts for the exist- 
 ence of evil in the universe in spite of his antecedent 
 volition for something different and better. 
 
 But God endeavours to produce whatever he wills 
 antecedently. The defeat of an antecedent volition 
 means a defeated effort. And what I am unable to 
 see is the possibility of explaining the defeat of the 
 effort solely from the nature of the being who made it- 
 He only acts by his will. And if his will is directed 
 to a certain end, can there be anything in his nature 
 which can hinder its execution ? 
 
 The result of answering this question in the negative 
 would not be to make the creator omnipotent. We 
 could still say that there were things which he could 
 not do, but we should hold that it was also impossible 
 for him to will to do them, since the impossibility 
 must be found entirely in his own nature, and could 
 only act through his will. We should not be compelled 
 to believe — as the upholders of his omnipotence must 
 beheve — that he could have created a universe in which 
 the laws of Identity and Contradiction did not apply. 
 But, since we should beheve that whatever he willed 
 would find nothing to hinder it, we should conclude 
 that the impossibihty of creating such a universe proves 
 the impossibility of his willing to create it. For the 
 impossibility of creating it rests entirely on his nature, 
 and his nature in acting is expressed simply by his will. 
 
 But if this is the case, the creator of the universe 
 cannot be a God. For then the cause of the evil in the 
 universe is not that the creator could not do what he 
 antecedently willed to do. Whatever he antecedently 
 willed to do has been done. But a person whose 
 antecedent will includes the production of evil cannot 
 
A NON-OMNIPOTENT GOD 231 
 
 be good, although, as we have seen, a being may be good 
 whose consequent will includes the production of evil. 
 And a being who is not good is not God. Even if we 
 say that the limitations on the creator's nature rendered 
 it impossible for him to will anything better, that 
 would not make him good, but only assert that his 
 badness was inevitable. Under no circumstances can 
 a person be called good, whose will is not thwarted 
 by the production of evil. 
 
 I do not know that I can make this important and 
 difficult point any plainer, except by clearing away 
 a few objections. It is not fair to argue that, because 
 men can will all sorts of impossible things, it must be 
 possible for God to will things which it is impossible 
 for him to carry out. A non-creative God could doubt- 
 less do so, because the impossibility of his carrying them 
 out would not depend exclusively on himself. His 
 efforts to do so might be thwarted by the nature of some 
 other being co-existent with him. It is only when 
 everything is completely dependent on God's nature — 
 as it must be in the case of a creative God — ^that we 
 can argue from the non-existence of a result to the 
 non-existence of any will on God's part to produce it. 
 
 189. We must also be on our guard against methods 
 of expression which treat God's nature as if it consisted 
 of different beings, capable of working in opposite 
 directions and neutralizing one another in the same 
 way that the wiU of one diplomatist may be opposed 
 by the will of another, or the will of a carver by the 
 nature of the marble. Thus it is said, for example, 
 that God is prevented from completely realizing his 
 antecedent volitions by the law of his own nature. 
 The will of God and the law which prevents its realiza- 
 tion are conceived of as so separate that their opposition 
 appears to present no difficulty. 
 
232 SOME DOGMAS OF EELIGION 
 
 But this is unjustifiable. A law is not another 
 existent thing, apart from the thing of which it is 
 a law, and capable of acting on it, as a wall may check 
 the course of a bullet. The law is simply the state- 
 ment of how the thing will act under certain circum- 
 stances. In other words, the law is not something 
 which controls the thing's nature from without. We 
 are often tempted to give it this significance — probably 
 from confusion with the laws of jurisprudence, which 
 do control our acts from without. But this is to con- 
 fuse two quite different conceptions, which have nothing 
 but the name in common. The laws of which we speak 
 here are not commands, but statements of facts. The 
 law of a thing's nature does not control it, but 
 expresses it. When we say that the law of the nature 
 of wax is to melt at a certain temperature, this is not 
 an outside authority to which the nature of the wax 
 submits. It is the nature of the wax. 
 
 And so the difiiculty remains unsolved — how can 
 God's nature at once impel him towards an end and 
 yet be the sole obstacle to his realizing that end ? 
 
 190. Or, again, God's will and God's power are 
 taken almost as separate beings.^ The tendency to 
 realize the antecedent volitions is ascribed to the 
 nature of the will, the failure to realize them completely 
 is ascribed to the nature of the power — that is to say 
 to its limitation. And, as we have seen, there is no 
 difficulty in maintaining that the antecedent volitions 
 which are due to the nature of one being may be 
 hindered from complete realization by the nature of 
 another being. 
 
 But this view involves a disruption of God's nature 
 which is indefensible. If there is a God, he is a person, 
 
 ^ Cp. Dr. Eashdall's Personality Human and Divine, quoted 
 above. 
 
A NON-OMNIPOTENT GOD 233 
 
 and not an abstract quality.^ Still less is God to be 
 resolved into a couple of abstract qualities which can 
 be treated as opposing one another. God's power — the 
 power of a creative God for whom there are no external 
 obstacles— can only mean the extent to which his own 
 nature does or does not permit him to perform things. 
 And we are no further towards a solution of the diffi- 
 culty how his nature can stop him from doing anything 
 except by stopping his wilhng it. 
 
 Whatever a creator does, he is completely self- 
 determined in doing it, for there is nothing but his 
 own nature to determine him. Therefore he is wholly 
 self-determined in producing, among other things, 
 much that is evil. Now when a being who is com- 
 pletely self-determined produces evil, knowing that it 
 is evil, can we say that such a being is good ? I do 
 not see how we can do so. A person who consciously 
 produces evil can be good, if he is not completely self- 
 determined, for then the evil may be due to the 
 external forces which determine him, as when the 
 surgeon is forced to give pain to prevent greater pain. 
 But a person whose nature is such that he spontaneously 
 produces what he knows to be evil — what is this but 
 the definition of a wicked person ? 
 
 191. If we combine this result with that attained in 
 the last chapter it would lead us to reject the idea 
 of a creative person who is either omnipotent or 
 thwarted in his volitions. This does not absolutely 
 exclude the possibility of a creative person. It would 
 be possible that a non-omnipotent person should exist 
 for whom the present universe would be in every 
 
 ^ No one has realized this truth more clearly than Dr. Rashdall, 
 or expressed it more forcibly. But I confess that it seems to me 
 that, in his treatment of the relation of God's will to his power, 
 he has temporarily lost sight of his own principle. 
 
234 SOME DOGMAS OF RELIGION 
 
 respect desirable for its own sake, and for whom 
 nothing which is excluded from that universe would 
 be desirable for its own sake. In that case his ante- 
 cedent vohtion would be for this universe, and for 
 nothing else, and it would therefore not be in the least 
 thwarted by his consequent volition to produce this 
 universe and nothing else. But such a person could 
 not be a God, because he would not be good. It is, as 
 we have seen, impossible that any person should be 
 good to whom some features of the present universe 
 were not intrinsically repugnant ^. 
 
 192. I have spoken so far of a creative and non- 
 omnipotent God who is distinct from that which he 
 creates. But it is also possible to beheve in a creative 
 and non-omnipotent God who is the sum of aU ex- 
 istence, and whose acts of creation consist in pro- 
 ducing modifications within his own nature, and not 
 beings outside it. Lotze, for example, appears to take 
 this view. This conception, I think, does not require 
 separate discussion. It is obvious that the difficulties, 
 already considered, of regarding a person as self- 
 thwarted, will be just as great if his creation operates 
 within himself as it would be if his creation produced 
 separate realities. And then, as was said in the last 
 chapter, the conception of God as the sum of all 
 existent reahty involves the additional difficulty — 
 which seems to me wholly insuperable — of regarding 
 one person as part of another person. 
 
 ^ Another argument against the possibility of a creative God 
 would be supplied if we found reason to believe that all existent 
 reality was eternal and ultimate. In that case there could not 
 be a creative God, or a creator of any sort, because there would 
 be nothing for him to create. It seems to me that by following 
 out the lines of Hegel's philosophy we are led to this conclusion, 
 as I have endeavoured to show in my Studies in Hegelian Cosmology, 
 chap, ii. 
 
A NON-OMNIPOTENT GOD 235 
 
 193, We now pass to the theory that God is neither 
 omnipotent nor creative. According to this theory- 
 all the non-divine existent beings of the universe are 
 co-eternal with God, and have not been called into 
 being by his will. Their existence is an ultimate fact, 
 and a fact which God has simply to accept, as we all 
 have to accept it about one another. God can affect the 
 condition of these other beings, in the same way that 
 they can affect the condition of one another, and, pre- 
 sumably, of God. So far there is no difference between 
 God and other persons. The difference is quantitative. 
 God is conceived to be so much more perfect in good- 
 ness than his fellow persons, that the due attitude of 
 all of them, even the highest, towards him is that of 
 reverence and adoration. And he is conceived to 
 excel them so much in wisdom and power that his 
 efforts are capable of producing important effects, not 
 only in one small comer of the universe, but in every 
 part of it. His position towards us is that of a school- 
 master towards his scholars. He does not create us. 
 He cannot destroy us. His power over us is limited. 
 And we can resist his power, and in some cases our 
 resistance is effectual — at any rate for a time, perhaps 
 permanently. But, on the other hand, his power is 
 greater than the power of any one of us, and is so 
 great that it can do much, though not all, of what 
 he wishes to do throughout the imiverse. Indepen- 
 dently of his exertions the universe would not be 
 completely bad, since beings who are capable of im- 
 provement cannot be completely bad. In spite of all 
 his exertions, he has not yet succeeded in making the 
 universe completely good. It is uncertain what his 
 eventual success wiU be. But it is at any rate certain 
 that the universe is better because he is working in it, 
 
 194, Of the three theories of God's nature which 
 
236 SOME DOGMAS OF EELIGION 
 
 we have to consider, this seems to me by far the 
 most tenable. God has been defined as good and as 
 a person. We have seen that, if God is held to be 
 omnipotent, it is, at the least, very difficult to conceive 
 him as a person, and absolutely impossible to conceive 
 him as good. The creative God who is not omni- 
 potent presents less difficulties. But it is almost as 
 difficult to conceive him as a person, unless he is 
 conceived as not free to abstain from creating. And, 
 in any case, we have seen that he can only be con- 
 ceived as good if he is conceived as self-thwarted, and 
 that the possibility of this is, at any rate, very doubtful. 
 
 Neither of these difficulties applies to the conception 
 of God which we shall now discuss. There is no 
 more difficulty in believing such a God to be a person 
 than in believing myself to be a person, since he, like 
 myself, is one member in a universe, none of the other 
 elements in which are dependent on hira for their 
 existence. These other members, therefore, may form 
 the Other of his personality. Nor does the existence 
 of evil in the universe reflect, of necessity, on his 
 goodness, since it may possibly all be due to defects 
 in the constitution of the other beings co-eternal with 
 him. In that case, of course, we should know that 
 God's power was not sufficient to remove these evils, 
 but this need not imply defective goodness in him, 
 any more than it would in a man. He is only re- 
 sponsible for making the evil as small as he can. And 
 the existence of evil does not prove that he has not 
 dqne this. 
 
 We may admit that there is nothing antecedently 
 impossible in the existence of such a God as this. 
 Persons do exist. And, of these persons, some excel 
 others in virtue, some in wisdom, and some in power. 
 It happens not infrequently that one person surpasses 
 
A NON-OMNIPOTENT GOD 237 
 
 another in all three. There seems no reasons why 
 one person should not surpass all others in all three 
 to such an extent that his goodness would fit him for 
 a universal worship, and his wisdom and power would 
 enable him to dominate the universe as much as an 
 efficient schoolmaster dominates his school. 
 
 195. But this is not enough. There is nothing 
 antecedently impossible in the supposition that the 
 next six dice I throw will come up 4, 1, 4, 8, 5, 5, 
 Indeed it may be said that, in a sense, there is no 
 antecedent improbability, for they must come up 
 somehow, and this series is just as likely as any other, 
 if the dice ai^e honest. Yet, if the dice are honest, 
 the odds required for a prudent bet on this series 
 would be enormous. There is nothing antecedently 
 impossible, again, in the supposition that at this 
 moment a play equal to Hamlet is being written in 
 Gower Street. The existence of Hamlet proves that 
 such a feat is possible to mankind. If one such play 
 has been written, another may be, and it is just as 
 likely to be written now in Gower Street as at any 
 other time and place. But we should want a good deal 
 more than this negative evidence before we should 
 believe that the great play was actually being written. 
 
 We must now inquire what evidence can be brought 
 in favour of the existence of such a God. Immediate 
 conviction, the disastrous consequences of his non-exist- 
 ence, revelation, and the ontological argument present 
 no new features when God is conceived in this way. 
 And such a God is not a first cause. But the argument 
 from the necessity of aU reality being known to some 
 one has not been considered at all, and the argument 
 from design has not been considered as applied to 
 a non-omnipotent God. Let us first consider the 
 argument from design. 
 
238 SOME DOGMAS OF KELIGION 
 
 196. It is doubtless the case that much of what we 
 observe around us in the universe could be easily 
 explained as the means for reahzing some end which 
 is good in itself, or for realizing some end which may 
 in its turn serve as means to a further end which 
 is good in itself. Now if reality is so arranged that 
 it becomes means to ends, it is a simple and natural 
 explanation to suppose that it has been so arranged 
 by some conscious being who wills the end, and 
 brings about the means for its sake. Much of what 
 appears to be so arranged would require much more 
 than human wisdom and power to arrange it. From 
 this it is a natural inference — though not strictly 
 logical, as we shall see later on — that it was arranged 
 by a being of sufficient pre-eminence to deserve the 
 name of God. Since omnipotence precludes the 
 employment of means, such a God could not be 
 omnipotent. 
 
 I have thought it convenient to discuss this argu- 
 ment while speaking of a non-creative God, because 
 we found reason for thinking that, if there was a God 
 at all, he was not a creator, and therefore the alternar 
 tive of the non-creative God was the only one which 
 remained. But the argument from design itself tells 
 us nothing as to whether the directing God creates 
 things so as to form certain arrangements, or whether 
 he arranges what exists independently of him, as the 
 watchmaker arranges the metal of the watch, and 
 does not create it. I shall endeavour to deal with the 
 argument in such a way that my remarks will be 
 applicable to either alternative, although, for the 
 reasons mentioned above, we have come to the con- 
 clusion that, if any God is proved by this argument, 
 he will not be a creator. 
 
 The facts of the case would be met, as I have 
 
A NON-OMNIPOTENT GOD 239 
 
 admitted, by the hypothesis that the order and good- 
 ness to be observed around us are due to the influence 
 of a person who antecedently willed the good, and 
 who, though not omnipotent, was very powerful. 
 Such a person would be entitled to be called God. 
 But before we can legitimately infer from this that 
 God does exist, we must inquire whether any other 
 hj^othesis would explain the facts as well. Would 
 the order, and the tendency to produce good, which 
 we can observe, in certain parts of our experience, 
 and to a certain degree, be explicable on the sup- 
 position that the reality around us was not under the 
 influence of a God ? 
 
 197. Much of what we observe in experience appears, 
 at any rate, not to be spirit but matter. And the 
 traces of order and goodness are to be observed in this 
 part of our experience as well as in others. Now 
 suppose that this appearance is correct, and that those 
 parts of the universe, which present themselves prima 
 facie as matter, really were matter. In that case we 
 should have existent reality which was unconscious, 
 which was not actuated by final causes, since it could 
 desire nothing and judge nothing good, and which was 
 actuated only by causes of a mechanical nature. More- 
 over, it would either be infinitely divisible, or divided 
 mto a great number of very small parts, and its parts 
 would have no intrinsic connexions which should 
 arrange them in a definite order, but would enter into 
 various combinations according to the external forces 
 which actuate them. 
 
 If such matter shows an order tending to the good, 
 which does not arise from the action of a conscious 
 being with an intention to promote the good, such bh 
 order can only arise, as is commonly said, by chance. 
 This does not mean that the events which brought 
 
240 SOME DOGMAS OF RELIGION 
 
 about the order had no cause, for every event has a 
 cause. (Even the systems which deny this of human 
 volitions would not deny it of matter.) "What then is 
 meant by chance ? 
 
 198. If a roulette table were constructed in which 
 letters were substituted for numbers, and if the letters 
 indicated by consecutive settlements of the baU spelled 
 out the received text of Hamlet, we should say, pro- 
 vided we were certain that the table was an honest 
 one, that this was a most extraordinary chance. Yet 
 the result was absolutely determined. With sufficient 
 knowledge, and sufficient power of reasoning, the 
 result could have been inferred with absolute certainty 
 before it happened. 
 
 There is no contingency in the connexion of the 
 particular cause with the particular effect. We can- 
 not get contingency unless we describe the cause and 
 effect respectively by some general class to which each 
 of them belongs. And then we get contingency if the 
 nature of the general class to which the cause belongs 
 has no tendency to ensure that the effect shall be of 
 the general class to which, in point of fact, it does 
 belong. Thus it is not contingent that a roulette table 
 of the sort I have described should produce an effect 
 which is a succession of letters. The nature of such 
 roulette tables ensures that they should produce such 
 an effect. It could not, for example, produce a succes- 
 sion of numbers, because the holes are not marked 
 with numbers, but letters. But if we bring the effect 
 under another general class to which it belongs, and 
 call it not merely a succession of letters (which it is), 
 but also a copy of the play of Hamlet (which it also 
 is), then the connexion is contingent. For there is 
 nothing in the nature of such roulette tables as a class 
 which makes it necessary or probable that its effect 
 
A NON-OMNIPOTENT GOD 241 
 
 should come under the general class of copies of Hamlet. 
 Thus we do not say that it is a chance that the table 
 produces a sequence of letters, but w^e do say that it is 
 a chance that it produces Hamlet 
 
 If then, matter uninfluenced by mind should assume 
 a form which contained ordered means towards a 
 good end, we should say that this was a chance, not 
 meaning that the particular cause could have produced 
 anything but what it did produce, but meaning that 
 there is no necessity or probability that a cause which 
 answers to the description of matter uninfluenced by 
 mind should produce such an effect as answers to the 
 description of ordered means towards a good end, 
 
 199. If we are perfectly sure that no cause can be 
 found for the effect which would have a tendency to 
 produce an effect of this nature, we must accept the 
 view that it has happened by chance, as explained 
 above. But when the chance would be antecedently 
 a very small one, we are unwilling to accept this view. 
 I suppose that no evidence which our imperfect human 
 knowledge admits would convince us that a roulette 
 table spelled out Hamlet by chance. We should always 
 prefer the belief that it was influenced, in some way 
 unknown to us, and perhaps by some law as yet 
 unknown to science, by a mind which had for some 
 reason resolved that it should spell out Hamlet. Even 
 supposing — which is a closer analogy to the state of the 
 world as we perceive it — that the table should at inter- 
 vals, interspersed with long series of letters without 
 significance or coherence, give us a sentence, now of 
 Hamlet, now of Paradise Lost, mutilated and distorted, 
 but still to be recognized, we should probably insist on 
 attributing this to the action of a mind, however unable 
 we were to explain how a mind could act on the 
 table. 
 
242 SOME DOGMAS OF RELIGION 
 
 The reason of this unwillingness to accept chance 
 is, it would seem, that if we accept as a cause a mind 
 which wills to produce the result, there is a reason 
 why a significant result should be produced rather 
 than any other, while, if we hold that all that we 
 know about the cause is that it is matter uninfluenced 
 by mind, we have to assume that a result has taken 
 place against which there was an enormous antecedent 
 probability. The uninfluenced roulette table is just as 
 likely to produce any one combination of letters as any 
 other, and the number of combinations of letters which 
 have no significance are enormously greater than those 
 which have a significance. This is true even if we 
 take a series of a hundred letters, much more if 
 we take so long a series as that of the letters in 
 Hamlet. 
 
 200. On this principle the argument from design 
 urges that the traces of order directed towards good 
 which we perceive in the universe render the hypo- 
 thesis that matter is directed by mind more probable 
 than the hypothesis that it is not directed by mind. 
 In the first case the order is accounted for at once. In 
 the second case we have to make the additional 
 hypothesis that it has been brought about by causes 
 which were just as likely to produce any one of 
 many results which excluded such an order, and 
 therefore were more likely to exclude the order than 
 to produce it. 
 
 There is certainly force in this contention, but it 
 is not as strong as it is sometimes supposed to be. It 
 is sometimes put as if the antecedent improbability 
 was that of precisely the actual arrangement of the 
 universe occurring among all the possible arrange- 
 ments which undirected matter could fall into. But 
 this is wrong. For many other arrangements of the 
 
A NON-OMNIPOTENT GOD 243 
 
 universe would have given us as good reason to trace 
 order directed towards good as the present arrange- 
 ment can. Consequently, to account for the present 
 state of the world, we should have, after assuming 
 a directing mind, to assume that it had chanced to 
 produce this particular arrangement, when it might 
 just as well, as far as we can see, have produced any 
 one of many others, and was therefore less likely to 
 have produced the existing one than some one of the 
 others. 
 
 Of course, if the present order is produced by a 
 directing mind there must be a sufficient cause why it 
 produced this order rather than another. But then, 
 if the present order were produced by undirected 
 matter, there must be sufficient cause why it produced 
 this order rather than any other arrangement. The 
 argument was based on the fact that all we know 
 about the nature of undirected matter was equally 
 compatible with an enormous number of other arrange- 
 ments. And it must accept as relevant the corre- 
 sponding fact that all we know about the nature of 
 a mind which wills the good would be equally com- 
 patible with a number of different orders. 
 
 Thus the greater improbability of the hypothesis 
 which excludes a directing mind is not to be expressed, 
 as is sometimes maintained, by the proportion of a 
 large number to unity, but of one large number to 
 another. But the number of arrangements which 
 would show some traces of such an order is very small 
 as compared with the total number of arrangements of 
 which undirected matter is capable. And thus the 
 argument, if not as strong as has been sometimes 
 supposed, retains considerable strength. 
 
 201. But there is another point to be considered. 
 The improbability of a result arising by chance de- 
 
 R 2 
 
244 SOME DOGMAS OF EELIGION 
 
 pends on the number of times that it does happen 
 as compared with the number of times when it 
 might have happened and did not. If I cast a die 
 ten times, and threw a six each time, I should have 
 Httle doubt that it had been loaded. But if I cast it 
 sixty million times, and got only one run of ten sixes, 
 I should find no difficulty in supposing that it came 
 by chance. 
 
 Now it is possible that all the traces of order in the 
 universe are in the position of the run of ten sixes 
 among sixty miUion throws. Our present discussion 
 is based on the hypothesis that matter does exist, and 
 that its existence is not necessarily connected with 
 spirit. In that case it would be possible that by far 
 the greater amount of the universe at this moment 
 does not exhibit the least traces of order, since it is 
 quite possible that by far the greater amount of the 
 universe is beyond the range of our observation. 
 And it would be possible that by far the greater 
 amount of past time did not exhibit, and that by far 
 the greater amount of future time will not exhibit, 
 any traces of order in any part of the universe. The 
 proportion of the amount of the universe which shows 
 such traces to the amount which does not may be so 
 small as to make the order which we see as explicable 
 by chance as the run of ten sixes in sixty million 
 throws. 
 
 In answer to this it has sometimes been said that 
 the times and spaces which would correspond to the 
 sixty million throws would be so great that even to 
 write the figures expressing them would be fatiguing. 
 But this would only be a valid objection if there were 
 any reason to suppose the world of matter to be limited 
 in time and space in such a way as to make it smaller 
 than this. And, on the hypothesis that there is a 
 
A NON-OMNIPOTENT GOD 245 
 
 world of matter at all, there seems no reason to 
 suppose that there is any finite duration or magnitude 
 which is too great to be possible for it. 
 
 The improbability that the traces of order should be 
 due to chance is thus lessened, but it is not removed. 
 For the existence of vast areais of reality in which no 
 trace of order is to be observed, while certainly possible, 
 is not more than possible. 
 
 The conclusion so far seems to be that, if any reality 
 is rightly conceived as matter, then there is a consider- 
 able probability — though by no means a certainty — 
 that any traces of order in it are due to the action of 
 a directing person. Such a person would be greatly 
 superior in wisdom and power to ourselves, and might 
 be — though this would not necessarily follow — also 
 superior in goodness. His superiority in all three 
 over all other beings might be sufficient to entitle him 
 to be called God. 
 
 202. But the existence of matter is not universally 
 admitted. It has often been maintained that notliing 
 exists except selves, and that the reahty which lies 
 behind the appearance of matter is in truth spiritual. 
 In that case all traces of order are to be found in spirit 
 only, because nothing else exists. How does this 
 affect the probability of a directing God ? 
 
 We must distinguish two varieties of the doctrine. 
 According to the first, the selves, taken separately, are 
 spiritual, but they do not form a spiritual unity. They 
 form a unity, no doubt, for they form the universe, 
 and the universe is certainly more or less of a unity. 
 But the arrangement of that unity is determined by 
 laws of a purely mechanical nature. The relations in 
 which selves find themselves with other selves are 
 determined by laws analogous to those on which, 
 according to the theory which accepts the existence 
 
246 SOME DOGMAS OF EELIGION 
 
 of matter, one particle of such matter is brought into 
 relation with other particles. 
 
 203. If we hold this view to be the truth, it will 
 not be so difficult, as it would have been on the 
 hypothesis that matter existed, to account for traces 
 of order without the agency of a directing God. For 
 every unit in the whole universe will be a self. It is 
 absurd to suppose that matter, if it is really matter, 
 should will the good, or take steps to carry out that 
 will. But it is by no means absurd to suppose that 
 selves should do so. However much the forces which 
 combined and separated selves were blind forces 
 regardless of the good, still if all or many of the selves 
 which they combined and separated strove towards the 
 good in whatever circumstances they found them- 
 selves, some traces of order would naturally arise in 
 the imiverse. 
 
 In such a case the universe might be compared to a 
 space into which himaan beings had been flung by 
 some natural force — an earthquake or a cyclone. That 
 they were in that place at all, and all their original 
 arrangement in it, and much of their future arrange- 
 ment, would be due to this blind natural force. But 
 if they kept any control of their actions, their state 
 would soon show some efforts towards ends. They 
 would try to escape, or to alleviate their position. 
 Some would sacrifice the advantages of others to their 
 own, some would sacrifice their own advantage for the 
 sake of others. An observer would soon be assured 
 that they were thinking and wilhng persons, and not 
 statues or automata. 
 
 But while a directing mind could thus be dispensed 
 with, on this hypothesis, with less improbability than 
 if matter really existed, there would still be some 
 difficulty in dispensing with it. We perceive traces 
 
A NON-OMNIPOTENT GOD 247 
 
 of order which extend over such large areas of the 
 universe, and which are so minute and exact in their 
 arrangements, that it is impossible to regard them as 
 due to the efforts of a single self no higher than we 
 are. And it seems very difficult to suppose them due 
 to the unconscious co-operation of many such selves, if 
 the relations and juxtapositions of those selves are due 
 to bhnd forces. On the other hand, if we suppose it 
 to be due to the conscious co-operation of selves, we 
 cannot suppose that those selves are no higher than 
 our own, since the knowledge required in order to 
 plan and carry out such a work would require far more 
 than human capabilities. It is simpler and more 
 probable to fall back on the supposition of a directing 
 person — or possibly a pluraHty of such persons — of 
 a much higher nature, 
 
 204. But there is a third hypothesis. We may 
 deny that matter exists, and we may deny that selves 
 are only connected by blind forces. We may hold 
 that reality consists of a system of selves, and we may 
 hold that the nature of that system can to some extent 
 be determined. We may hold further that what we 
 can determine about that system is such that it shows 
 it to exhibit an order directed towards the production 
 of something of spiritual significance and value — of 
 something which is either completely good, or, at any 
 rate, more good than bad. Such a view as this is the 
 basis of the systems of Fichte, of Hegel, of Lotze. 
 
 Now, if we hold this view, it seems to me that the 
 directing mind is not wanted at all to account for the 
 traces of order in the universe. The natui'e of reality 
 has been seen to be such that it inevitably manifests 
 itself in order directed towards the good. It is a 
 harmonious system — no longer a mere aggregate, 
 whether of atoms or souls— and so it must manifest 
 
248 SOME DOGMAS OF EELIGION 
 
 itself in order. And it is a system of such a nature 
 that it is directed towards some end which is, on the 
 whole, good. And therefore the order in which it 
 manifests itself must be an order directed, to some 
 degree at any rate, towards what is good.^ 
 
 205. We may be asked how such a harmonious 
 system of selves came into existence, and it may be 
 said that its production would require a directing 
 mind. But this is not valid. For, if this theory is 
 true, the harmonious system of selves is the funda- 
 mental fact of the universe. And, on any theory, 
 there must always be a fundamental fact which is 
 accounted for by nothing, and which accounts for 
 everything. If it were a fatal objection to this theory 
 that it takes a harmonious system of selves as an 
 ultimate fact, which cannot be explained any further, 
 it would be just as fatal an objection to the opposite 
 theory that it takes the existence and nature of God 
 as an ultimate fact, which cannot be explained any 
 further. If it is asked, 'Who made the system of 
 selves, and secured that it should be harmonious 
 rather than chaotic ? ' it may as well be asked, ' Who 
 made God^ and secured that he should be good rather 
 than bad?' The fact is that both questions are un- 
 justifiable, since the system of selves, on the one hand, 
 and God on the other hand, are both put forward as 
 ultimate facts. 
 
 If matter, or selves mechanically arranged, are 
 accepted as real, the probability of a directing mind 
 arises from the fact that nothing that we know about 
 
 ^ The point is not merely that we know that the nature of 
 reality is such as to produce the traces of order. We knew that 
 before, since we knew that it did produce them. The point is 
 that the qualities which we have determined reality to possess are 
 such that they are seen to involve the production of such an order, 
 and there is no longer any chance or contingency in its production. 
 
A NON-OMNIPOTENT GOD 249 
 
 the nature of the matter, or the nature of the selves, 
 would explain the traces of order, and that therefore 
 we had a tenden-cy to postulate an external cause 
 whose nature would explain them^ But here the 
 nature of the system of selves does explain the traces 
 of order, and there is no need to go further. 
 
 206. Such a harmonious system of selves, if it 
 existed, would be the fundamental reality of the 
 universe. But its existence is, of course, by no means 
 self-evident, and we must prove its truth before we 
 are entitled to believe in it. Such a proof cannot be 
 attempted here, since it would involve a discussion 
 of a complete system of metaphysics. The theory is, 
 at any rate, one which is widely held. And the onus 
 of disproving it lies on the supporters of the argument 
 from design. Their position is that the hypothesis of 
 a directing God is the only one which, without great 
 improbability, can be used to explain the traces of 
 order in the universe. They are bound, therefore, to 
 disprove an hypothesis, which, if true, would explain 
 the traces of order in the universe without any im- 
 probability at all. 
 
 The existence of such a harmonious system of selves 
 would not disprove the existence of a directing God. 
 It might well be that among those selves was one 
 who, in goodness, wisdom, and power, so much ex- 
 celled all the others that he might appropriately be 
 called God, And so powerful a being would doubtless 
 have a large share in carrying out the order of the 
 universe. But, although the existence of a harmonious 
 system of selves is not incompatible with the existence 
 of a directing God, it is, as we have seen, incompatible 
 with any possibility of proving his existence by the 
 argument from design. The order in the universe 
 would be there, on this hypothesis, whether there was 
 
260 SOME DOGMAS OF EELIGION 
 
 a God or not, and cannot, therefore, be used to prove 
 his existence. 
 
 207. Idealism and Theism are generally held to be 
 theories which have a natural affinity. But they are 
 perfectly separable. It is quite possible to believe 
 that there is no reality except spirit, and yet to reject 
 the conception of a God. On the other side, most 
 people who accept the existence of God have no doubt 
 that jnatter is as real as spirit. And we see now that 
 under certain circumstances the interests of Theism 
 may lie in disproving Idealism. If there is matter, 
 there is much in the arrangement of matter which 
 seems to require a God to direct it. But if all reality 
 is a harmonious system of selves, it is perhaps itself 
 sufficiently godlike to dispense with a God. 
 
 The weight which MiU attaches to the argument 
 from design must in any case be an interesting and 
 significant fact.^ But considerations which affected 
 him would have no force for a thinker who believed, 
 as Mill did not believe, that reality, of its own nature, 
 and without any direction, formed a harmonious 
 spiritual system. 
 
 208. We must now proceed from the argument 
 from design to the argument from the necessity of 
 everything being known to some one. This argument 
 has been stated by Dr. Eashdall in a criticism with 
 which he honoured a previous book of mine. 'Mr. 
 M^Taggart (whatever we may say of the " Pluralists ") 
 feels that the world must be a Unity, that it consists 
 not merely of souls but of related and inter-connected 
 souls which form a system. But a system for whom ? 
 The idea of a system which is not " for " any mind 
 at all is not open to an Idealist ; and the idea of a 
 world each part of which is known to some mind 
 
 ^ Essays on Eeligion, III, i. 
 
A NON-OMNIPOTENT GOD 261 
 
 but is not known as a whole to any mind is almost 
 equally difficult. Where then, in his view, is the 
 Mind that knows the whole? i.e. the whole system 
 of souls with the content of each.'^ 
 
 If there is no omniscient person there is doubtless 
 a possibility that some things may exist which are not 
 known to anybody — or, at least, not fuUy known to 
 anybody. And in that case there can be nobody who 
 knows everything — no one, that is, who knows the 
 system of selves in its entirety. And, again, ^n 
 omniscient person might very probably be— though it 
 is not certain that he would be— a person of such 
 power and goodness as to be rightly called God. 
 
 209. But I cannot see that it is at all necessary 
 for an idealist to admit that nothing can exist except 
 that which is for a mind, in other words, except that 
 which is known. There is, no doubt, a school of 
 Idealism which maintains this. It has been main- 
 tained that to Be is to be Perceived, or that to Be 
 is to be Thought. To such Ideahsm, certainly, Dr. 
 Eashdall's argument applies. If all reality is a system, 
 and if only that has being which is known, then 
 some person must know the system, and so know all 
 reality. 
 
 There is, however, another form of Ideahsm — the 
 form which seems to me to be true — which is not liable 
 to these criticisms. This form of Idealism does not 
 say that nothing can be real except what is known. It 
 says that nothing can exist but persons — conscious 
 beings, who know, will, and feel. To the traditional 
 expression of the first-mentioned school, Esm est 
 percipi, the adherents of the second view might, for 
 the sake of antithesis, oppose the maxim Esse est 
 percipere. But it must always be remembered that 
 
 ^ Ferso'iial Idealism, Essay VIII, p. 393. 
 
252 SOME DOGMAS OF RELIGION 
 
 such a formula sacrifices accuracy to antithesis, since 
 persons have other activities as fundamental as know- 
 ledge. 
 
 Now if we take this view, there seems no difficulty 
 at aU in saying that certain aspects of reality are 
 unknown to every one. The theory maintains that 
 nothing exists but persons connected in a unity. Ac- 
 cordingly, whatever exists must either be a conscious 
 person, or a quality belonging to him, or an event 
 happening to him, or else it must be one of those 
 relations which connect these persons, and make up 
 their unity. In the latter case, while it does not 
 actually fall within any one person, it involves a quality 
 which does. For if A and B are in relation, then 
 A has the quality of being related to B, and B has 
 the quality of being related to A, Thus there is no 
 reality which cannot be expressed in a proposition 
 about conscious persons. 
 
 But it does not follow from this that whatever 
 exists must be known to some one. A proposition 
 about one conscious person stands in no different 
 relation to any other conscious person than a pro- 
 position which was not about a person at all would 
 do. If a true proposition about a conscious person 
 cannot remain unknown, and the reason of this is 
 that it is about a conscious person, it can only be 
 because such a proposition must be known to the 
 person himself. But this is certainly not the case. 
 It is often a true proposition about a person that he 
 is vain, humble, or the like, without his knowing it. 
 Again, it must be true of every conscious person either 
 that he was created by a God or that he was not. 
 But many persons do not know either of these pro- 
 positions to be true — either because they are in doubt, 
 or because they are too young or ignorant to have 
 
A NOK-OMNIPOTENT GOD 263 
 
 thought about them. In these cases something must 
 be true about a person which he does not himself know, 
 and in this event the form of Ideahsm which we are 
 now discussing can offer no opposition to its being true 
 without any one knowing it. 
 
 If I am right, then, we must reject Dr. Eashdall's 
 argument that Ideahsm involves an omniscient person. 
 My counter-argument has, of course, not been directed 
 against Dr. EashdaU's view that one form of Idealism 
 does involve an omniscient person, but has only pointed 
 out that another form of Idealism 'exists which does not. 
 
 210. Another point remains for consideration. If 
 it were proved that there was a person in the universe 
 who greatly excelled all others both in wisdom and 
 power, yet this would not by itself prove the existence 
 of a God. For God has not only to be wise and 
 powerful, but also good. How are we to prove that 
 this director of the universe is good enough to be 
 called God? 
 
 In the first place, he could not, as it seems to me, 
 be perfectly good. He is one self in a universe which 
 contains other selves. He is continually acting on 
 them. They, in their turn, are continually acting on 
 him, sometimes helping his plans, sometimes thwart- 
 ing them, sometimes rejoicing him, sometimes griev- 
 ing him. Selves in such a union as this form a 
 society. We have to consider, then, the director of 
 the universe as one of a society of selves. 
 
 Is it possible that one member of a society should 
 be completely iperSect while others are not? We see, 
 of course, in everyday life, that people of very different 
 degrees of perfection may be closely united. But 
 I do not see how there can be any unity at all if each 
 is not helped by the perfection, and hindered by the 
 imperfection, of every other. Any hindrance must 
 
254 SOME DOGMAS OF EELIGION 
 
 prevent the person hindered from being quite perfect, 
 and this seems to render God's perfection impossible. 
 
 ' No doubt ', it might be rephed, * God is not com- 
 pletely perfect — at any rate, as yet. To be perfect he 
 would have to be in perfect harmony with his environ- 
 ment, to be perfectly acquiescent and happy. And 
 this he cannot be in a world which contains sin and 
 pain. But this does not prevent him from being com- 
 pletely perfect morally. He can be completely good 
 without being completely happy.' 
 
 It is surely, however, a false abstraction to maintain 
 that any cause can work an effect on one aspect of 
 a person's life and leave the others untouched, A per- 
 son is not a mere aggregate of water-tight compartments. 
 He is essentially a unity, although that unity is mani- 
 fested in a plurality of activities. And, consequently, 
 nothing can be a hindrance to the perfection of any 
 side of his nature without affecting that unity, and, 
 through it, all the other sides of his nature. One 
 cause, no doubt, may have much more effect on one 
 side of his nature than on other sides, and so the sides 
 may develop unevenly. Thus three people, if arranged 
 in the order of their intellectual perfection, might be 
 placed — A, B, C; if in the order of their moral perfec- 
 tion, B, Aj G; and if in the order of their happiness, 
 (7, A, B, But every cause must have some effect on 
 each aspect of each nature, and so, if any cause exists 
 which causes some imperfection in any aspect of God's 
 nature, it will be impossible to regard him as completely 
 perfect in any respect. 
 
 211. The possibility would remain that the director 
 of the universe, although not perfectly good, was very 
 much better than any other person. This would 
 justify us in calling him God. Such a comparative 
 perfection would suffice for worship. If worship is to 
 
A NON-OMNIPOTENT GOD 256 
 
 be anything higher than selfish propitiation of the 
 powerful, it requires that the being worshipped shall 
 be conceived as morally better than the devotee. But, 
 if this is secured, it does not seem essential that it 
 should be accompanied by an assertion of the absolute 
 perfection of the object of worship. Thus the director 
 of the universe, if he exists, may be a God. But it is 
 also possible that he may not be a God. For that 
 would require that he should be distinctly better than 
 the best man. And I do not see that it is impossible 
 that he should be even worse than the worst man. 
 Absolutely bad, indeed, he could not be. Evil, 
 though not a mere negation, is nevertheless of a dis- 
 tinctly negative nature, and could only exist as limit- 
 ing and confining the good. An absolutely bad person 
 ■ — one without any goodness in him — would be an 
 impossibility. 
 
 But it is not impossible that the director of the 
 universe should be worse than the worst man. Our 
 only ground of inference as to his moral nature is the 
 present condition of the universe, which would be due 
 largely to his influence. It was admitted that much 
 of what we observe around us in the imiverse could 
 be easily explained as the means for realizing some 
 end which is good in itself. And if it is to be explained 
 this way, we might argue that the director of the uni- 
 verse must be good. But it seems to me that the facts 
 which we observe could just as easily be explained as 
 the means for realizing some end which is bad in itself. 
 Many of the ends to which facts of the universe appear 
 to be adapted as means are partially good and partially 
 bad. Some of the ends to which the facts of the 
 universe appear to be adapted as means seem to be 
 purely bad. If we believe the director of the universe 
 to be good we say, in the first case, that he brings such 
 
256 SOME DOGMAS OF EELIGION 
 
 ends about for the sake of the good in them, and 
 because the good outweighs the evil with which it is 
 necessarily connected. In the second case, when the 
 ends achieved seem to be purely bad, it is said that in 
 some way which we do not know they may in their 
 turn be indispensable means to some end whose good 
 outweighs their evil. 
 
 And all this is quite possible. But then, it seems 
 to me, the contrary hypothesis is also quite possible — ■ 
 that the director of the universe has proposed to him- 
 self an end which is distinctly bad. The existence of 
 the good could then be explained in the same way as 
 the existence of evil was by the more cheerful theory. 
 It would be something which the director of the uni- 
 verse unwillingly brought about, because the constitu- 
 tion of the universe prevented him from producing 
 the evil without producing this amount of good also. 
 It would no more prove his goodness than the refresh- 
 ments administered in the intervals of tortures proved 
 the humanity of the torturers. 
 
 Nor would this possibility be removed, even if we 
 could prove that good far outweighs evil in the uni- 
 verse. It would still be possible that the aim of the 
 director of the universe was to produce a much worse 
 result, and that the excess of good merely proved that 
 the conditions under which he worked were unfavour- 
 able to his purpose. 
 
 212. I cannot see, therefore, that any reason has been 
 given for supposing a director of the universe to be 
 good rather than bad. But even if we assume that he 
 is more good than bad, it would not follow from this 
 that he would be a God. He might be no better than 
 the average good man. The size and complexity of 
 the universe — even of that little part of it which we 
 imperfectly observe — is so enormous that a being who 
 
A NON-OMNIPOTENT GOD 257 
 
 directed it must very greatly excel huraan beings in 
 wisdom and power. But it is not equally necessary 
 that he should excel them in goodness. For it is not 
 certain that the universe excels the works of men in 
 goodness at all, and still less certain that it excels 
 them very greatly in goodness. Indeed most men 
 would, I imagine, be distinctly relieved if they were 
 certain that good bore as large a proportion to evil in 
 the universe as a whole as it did in the work of such 
 men as Buddha, Aquinas, or Mill, Of course, a uni- 
 verse which was proportionately much more evil than 
 Mill's work might have been produced by a person 
 much better than Mill, for the evil might have been 
 due almost exclusively to the limitations of his power. 
 But it is also possible that the conditions might have 
 allowed of a better state of things, and that some of 
 the existing evil is due to the defective virtue of the 
 agent. 
 
 It is possible, for example, that, while good is 
 attractive to the director of the universe, other things 
 are also attractive. And, when the two attractions 
 lead to different courses, he may sometimes be 
 tempted, like less wise and less powerful persons, 
 and choose the worse alternative. Or, again, to 
 produce the best possible result may well cause him 
 fatigue and pain, and it is possible that the amount of 
 these which he is prepared to encounter is limited. 
 The universe might have been better if his self- 
 sacrifice had been greater. 
 
 213. And, once more, if the director of the universe 
 is finite, why should we be certain that there is only 
 one ? Many of the facts of experience, while they are 
 compatible with the theory of a single director working 
 under limitations, suggest at least as strongly the idea 
 of several such beings, working in opposition, or 
 
 s 
 
268 SOME DOGMAS OF EELIGION 
 
 possibly — as patriotic Tories work with patriotic 
 Liberals — partly in harmony and partly in oppo- 
 sition. The opposition might be direct, between 
 powers whose ends are intrinsically contrary, or 
 indirect, between powers which had different ends, 
 not intrinsically opposed, but capable of clashing 
 under certain circumstances. 
 
 Or, supposing that only one director is at work in 
 the part of the universe which we know, still that 
 part may be very small compared with the whole. 
 How shall we tell that there are not other regions — 
 perhaps separated from ours by Vast ungoverned 
 intervals — in which other beings, higher or lower 
 than he whose work we perceive here, are working 
 out other independent and isolated purposes? 
 
 There is nothing, perhaps, which should prevent us 
 from giving the name of God to each of several beings 
 simultaneously existing, or to one such being, existing 
 simultaneously with others, who equal him in wisdom 
 and power, but not in goodness. It may not be 
 impossible to revert to polytheism, or to conceive 
 God as striving against other persons who equal him 
 in everything but goodness. But the name of God 
 seems to imply that the person to whom it is applied 
 is of appreciable importance when measured against 
 the whole universe* A person who was only one 
 among millions of similar beings would scarcely be 
 allowed the name. And yet this may be the case 
 with the person, if there is one, to whom we owe all 
 the order and purpose which we can observe in the 
 universe. 
 
 214. If we make God to be less than a creator, we 
 make it possible that he should be a person, and that 
 he should be good. And it is sufficiently certain that 
 his wisdom and power would excel our own. But 
 
A NON-OMNIPOTENT GOD 259 
 
 when we come to his relation to the universe, Hume's 
 suggestions are more easily ignored than disproved. 
 * In a word, Cleanthes, a man who follows your 
 hypothesis, is able, perhaps, to assert, or conjecture, 
 that the universe, sometirae, arose from something 
 like design : but beyond that position he cannot 
 ascertain one single circumstance, and is left after- 
 wards to fix every point of his theology by the 
 utmost licence of fancy and hypothesis. This world, 
 for aught he knows, is very faulty and imperfect, 
 compared to a superior standard ; and was only the 
 first rude essay of some infant deity, who afterwards 
 abandoned it, ashamed of his lame performance : it is 
 the work only of some dependent, inferior deity ; and 
 is the object of derision to his superiors : it is the 
 production of old age and dotage in some super- 
 annuated deity ; and ever since his death, has run on 
 at adventures, from the first impulse and active force, 
 which it received from him. You justly give signs of 
 horror, Demea, at these strange suppositions : but 
 these, and a thousand more of the same kind, are 
 Cleanthes's suppositions, not mine. From the moment 
 the attributes of the deity are supposed finite, all 
 these have place/ ^ 
 
 215. Finally, it is necessary once more to emphasize 
 the fact that, if God's moral character is saved by 
 limiting his power, we have no right to be confident 
 as to the eventual victory of those ends in which God 
 is interested. We know that he will work for them, 
 and we know that they will be the more triumphant 
 or the less defeated because of his efforts. But we do 
 
 ^ Dialogues concerning Natmal Eeligion, Part V, ad fin, I may 
 remark that it seems to me that Hume's most important contri- 
 bution to the philosophy of religion is to be found in these 
 Dialogues, and not in the more famous discussion on Miracles 
 in the Enquiry, 
 
 S 2 
 
260 SOME DOGMAS OF RELIGION 
 
 not know that they will be completely triumphant. 
 Nor is this all. We do not even know that they will 
 not be almost completely defeated. The fundamental 
 conditions of the universe may prevent it from being 
 an3i;hing but very bad, and all that God's utmost 
 efforts may be able to do is to make the inevitable 
 calamity a little less calamitous. 
 
 That is all that the doctrine of a non-omnipotent 
 God can give us — a person who fights for the good 
 and who may be victorious. But it is at any rate 
 better than the doctrine of an omnipotent person to 
 whom good and evil are equally pleasing. And it is 
 fortunate that, as we have seen, the more attractive 
 of the two ideas is also the more probable. Indeed, 
 when the non-omnipotent God is also taken as non- 
 creative, there seems to me, as I have said, only one 
 reason why we should not beheve in his existence — 
 namely, that there is no reason why we should 
 believe in it. 
 
CHAPTEE VIII 
 
 THEISM AND HAPPINESS 
 
 216. If the results which I have reached in the last 
 two chapters are vahd, it would seem that we have no 
 reason to beUeve in the existence of a God. It remains 
 to discuss the effect which such a conclusion would 
 have on the happiness of those who hold it. The fact 
 that the belief in a proposition would render us happy 
 or miserable gives us no reason whatever for accepting 
 or rejecting the proposition. But this does not prevent 
 the inquiry being important. Indeed, except for those 
 few students who pursue knowledge for its own sake, 
 the main interest of the inquiry into the existence of 
 God lies in its bearing on our happiness and on that 
 of our fellow creatures. 
 
 Men's desires on this subject are very diverse. 
 Some people passionately desire the existence of a 
 God; some are indifferent ; some are desirous that God 
 should not exist. Now a thing may be desired either 
 for its own sake, or as a means to something else 
 which is desired for its own sake. And, when a thing 
 is desired for its own sake, any difference of opinion 
 on the subject may prove to be an ultimate dif- 
 ference not susceptible of alteration by argument. 
 It is sometimes possible to alter a desire for an end 
 by argument. For example, a man who desired the 
 existence of an omnipotent God for its own sake, 
 might be convinced by argument that such an omni- 
 potent God would be wicked, and might then cease to 
 
262 SOME DOGMAS OF RELIGION 
 
 desire his existence. But it is also possible that of 
 two men who held exactly the same idea of what the 
 nature of an omnipotent God would be, one would 
 desire that he should exist, and one would desire that 
 he should not exist. In this case, it would be impos- 
 sible to alter the opinions of either of them by argu- 
 ment. In the same way, if one man desires to drink 
 claret rather than burgundy, because he prefers the taste 
 of claret, and another desires to drink burgundy rather 
 than claret, because he prefers the taste of burgundy, 
 it is useless to try to alter the opinion of either by 
 argument. In the latter case the futility would be 
 generally recognized. It is admitted that there is no 
 disputing about tastes. But with those desires, which 
 relate to our deepest needs and highest aspirations, the 
 uselessness of argument is often ignored where it 
 should be admitted. 
 
 When a thing is desired as a means, however, the 
 desire is never so ultimate as to make argument irre- 
 levant. If a man desires A only for the sake of Z, it 
 would always be relevant to prove to him that A will 
 not produce Z, or that B will produce it better, or as 
 well. In the first case he will cease to desire A at all ; 
 in the second case he will only desire it if B is im- 
 possible ; in the third case he will be indifferent 
 whether A ov B exists, so long as one of them does. 
 
 The interests which men conceive themselves to 
 have in the existence of God may be divided into two 
 classes. In the first place, there are the desirable 
 results which we may expect will be produced in the 
 universe by a person who is pre-eminent in goodness, 
 wisdom, and power. We may call this our interest 
 in God's works. In the second place, the very exist- 
 ence of such a being, apart from the beneficent results 
 which his action may produce, may be considered 
 
THEISM AND HAPPINESS 263 
 
 desirable, This may be called, to distinguish it from 
 the other, interest in God's person, 
 
 217. If God is desired through our interest in his 
 works, he is desired as a means. The universe — the 
 sum total of existence other than God — wiU be more 
 desirable, it is said, if he exists than if he does not 
 exist. And for this reason God's existence is desir- 
 able. About such a desire as this it is not useless 
 to argue. We can inquire how far the existence of 
 God would guarantee a desirable universe, and whether 
 anything else could guarantee it if God did not exist. 
 
 It is clear, from what has been said in Chapter VI, 
 that the existence of an omnipotent creator would 
 afford no guarantee as to the goodness of the universe. 
 Such an omnipotent creator, we saw, could not be 
 supposed to be attracted by goodness as such, nor to 
 be repelled by badness as such. For if he were, every- 
 thing that existed would be good, and nothing would 
 be bad. And this is notoriously not the case. 
 
 Some things in the universe are good^ and some are 
 bad. And, since everything would be exactly as an 
 omnipotent creator would like best to have it, such 
 a creator must sometimes like things which are good 
 and sometimes things which are bad. Therefore we 
 have no reason at all for anticipating what the nature 
 will be of those parts of the universe which are so far 
 removed from us, either in space or time, that we do 
 not know them directly. They may be all good, so 
 that evil is to be fouijd nowhere except where we 
 already know that it exists. Again, they may be all 
 evil, so that there is nothing good except what we 
 already know to be good. Or anji^hing between these 
 extremes is possible. 
 
 The knowledge that the universe was created by 
 
264 SOME DOGMAS OF EELIGION 
 
 an omnipotent creator can tell us nothing about the 
 character of the universe that we did not know already. 
 If we can see in the universe traces of a divine design, 
 we might be able to infer what God wanted to produce 
 everywhere. But omnipotence, as we have seen, can 
 have no designs, for it needs no means. All we could 
 infer would be that everything which we know to 
 exist was such as to please the creator for its own sake. 
 And as what we know to exist consists of good and 
 evil, in proportions which vary from place to place, 
 and from time to time, it is impossible to predict how 
 much good and how much evil will be pleasing to the 
 creator in places and times which we do not know. 
 
 Thus the existence of an omnipotent creator — to 
 whom, as we have seen, the name of God would be 
 inappropriate — would give us no reason to expect any 
 goodness in the universe which we should not have 
 expected otherwise. And therefore, from the point 
 of view of our interest in his works, we lose nothing 
 by the non-existence of such a creator. 
 
 218. The existence of an omnipotent creator has, 
 no doubt, sometimes been made the basis of an 
 optimistic view of the universe. Some people have 
 been able to believe that the creator of such a universe 
 as this could be both omnipotent and good. And they 
 have also believed that, while a certain amount of evil 
 in the universe would not be incompatible with his 
 goodness, a greater amount would be. In particular, 
 it is often held that, although temporary evil is com- 
 patible with God's goodness, permanent evil is not.^ 
 
 ^ Cp. Lowell (Biglow Papers^ Second Series, VII) on the conflict 
 between freedom and slavery ; 
 
 * Set the two forces foot to foot, 
 An' every man knows who'll be winner, 
 Whose faith in God hez ary root 
 Thet goes down deeper than his dinner,' 
 
THEISM AND HAPPINESS 265 
 
 Such a belief would bring happiness to those who 
 held it. It would enable them to believe that the evil 
 in the universe would not exceed a certain amount. 
 It might also enable them to believe that the evil 
 diminished as time went on, and perhaps died away 
 entirely. And a man who lost such beliefs as these 
 would lose much happiness. 
 
 But the position rests on a basis wliich is untenable. 
 It is not possible that the creator can be both omni- 
 potent and good. And, if this difficulty was removed, 
 and he could be both, then his goodness would be 
 completely untarnished by the existence, in the present 
 and the past, of some evil in the universe. And, if 
 any evil can exist without tarnishing it, it is impossible 
 to prove that it would be tarnished by any amount of 
 evil, however permanent. 
 
 Optimism which rests on such a basis would not 
 need for its destruction the rejection of the belief in 
 an omnipotent creator. It would destroy itself as soon 
 as the believer in an omnipotent creator realized what 
 his own belief meant. A believer in an omnipotent 
 creator, if he were not hopelessly illogical, could get no 
 optimism out of his belief, and would lose no optimism 
 by ceasing to believe in such a being. 
 
 219. Let us pass to the conception of a non-omnipo- 
 tent being who was either the creator or the director 
 of the universe. If non-omnipotent, he may be good, 
 though we found reasons in the last chapter for sup- 
 posing that, if he is good, he would not be creator, but 
 only director. If he is good, then he may be called 
 
 But since slavery existed in 1862, when the poem was written, 
 God had clearly been either unable or unwilling to abolish it up 
 to that date. If the recognition of that fact is compatible with 
 faith in God, I can see no reason why a man should be thought 
 faithless who doubted as to God's power or intention in the 
 future. 
 
266 SOME DOGMAS OF KELIGION 
 
 God. From the point of view of our interest in his 
 works, do we gain much by believing in such a God, 
 or lose much by disbelieving in him ? It seems to me 
 that we gain very little, whether we take him as direc- 
 tor only, or whether we waive the difficulties which 
 stand in the way of taking him both as good and as 
 creator. 
 
 Let us consider first the non-creative God. If such 
 a God exists, and if (as would be perhaps more 
 probable) he never dies, and never deserts the cause of 
 the good, then the imiverse is better than it might 
 possibly have been. This, so far as it goes, is a gain. 
 But it goes a very little way. The amount that God 
 may be able to effect in making the universe better 
 may be almost indefinitely small. It is true that we 
 should not call him God unless his power were very 
 great in comparison with ours, and unless it were so 
 great as to have some appreciable effect on tliB universe 
 as a whole. But the conditions under which his power 
 is to be exerted — conditions which he cannot change — 
 might be such that, however great be his power, how- 
 ever wisely it be exercised, it could only be said of 
 him, as of the general of a hopelessly beaten army, that 
 his efforts had made the result slightly less disastrous 
 than it would have been without him. 
 
 And, again, even if we suppose that God's exertions 
 in the cause of the good have produced a result which 
 is very greatly better than anything which would have 
 happened without his exertions, still what is greatly 
 better may yet be very bad. Those fundamental 
 conditions of all existence which he cannot change 
 may have such evil effects that, even after God has 
 succeeded in introducing very much good into the 
 universe, it may still be very much more bad than 
 good. 
 
THEISM AND HAPPINESS 267 
 
 If we put these two considerations together — the 
 uncertainty of how much God may be able to do, and 
 the uncertainty of how much there is to be done — it 
 would seem that the hypothesis of a non-creative God 
 can give us no appreciable help towards a cheerful 
 view of the universe. For such a view we require 
 evidence of how good or bad the universe is. The 
 hypothesis gives us no help here. And for our pre- 
 sent purpose it is useless to be told that circumstances 
 could be imagined under which the universe would be 
 worse than it is. 
 
 220. The same principles apply in the case of a 
 non-omnipotent God who is a creator. He will make 
 the best universe he can, but we cannot teU how bad 
 the best possible universe may be. The conditions 
 under which he works may be such that the best 
 possible may be very bad indeed. 
 
 'At any rate,' it may be said, 'if God creates the 
 universe, we have a certainty that it is not, on the 
 whole, more bad than good. For, if it were more bad 
 than good, it would have been better to have created 
 nothing at all, than to have created it. And so the 
 fact that God has created it proves that the evil does 
 not outweigh the good.' 
 
 But this implies that God was free to abstain from 
 creating at all. Of course this is possible. Much 
 would depend on the will of a creator, even if he 
 were not omnipotent. And among the things which 
 depended on it might be whether any created world 
 should ever exist, or whether he should remain in 
 eternal isolation. But it is also possible that among 
 the things which the constitution of the universe 
 imposed as necessities on God was that he should 
 create something. In that case, the existence of the 
 universe only proves that it was the best thing that 
 
268 SOME DOGMAS OF EELIGION 
 
 God was able to create. It gives us no ground for 
 believing that the universe is good on the whole or 
 even that it is not, on the whole, very bad, 
 
 221. Is there any way by which we can escape from 
 the shadow of such dreary possibilities ? It seems to 
 me that there is only one way. If we can so deter- 
 mine the fundamental nature of reality as to see 
 what limitations it imposes on the accomplishment 
 of God's volitions, we might find that, while it made 
 it impossible that the universe should be entirely 
 devoid of evil, it did not make it impossible that the 
 good should always exceed the evil. We might also 
 find that the question of whether the good should 
 exceed the evil depended on God's volition. Now 
 God has been defined as good, and if we have any 
 reason to beheve that a God exists, we may be sure 
 that he would will that the good should exceed the 
 evil. In that case we might be sure that it did always 
 exceed it. 
 
 Or we might find that the limitations on God's 
 power, while they did not admit of a universe originally 
 more good than evil, yet did admit of a progressive 
 amelioration of the universe in time, if God willed 
 it. And, as God certainly would will it, we should 
 then be certain that this progressive amelioration 
 would take place, and would be carried as far as was 
 permitted by the constitution of reality — which might 
 be to the point of the complete extirpation of evil. 
 
 Both these results, again, might be combined. We 
 might find reason to believe that it was in God's 
 power to ensure that good should always predominate 
 over evil, and also that the predominance of good over 
 evil should increase till only good remained. And 
 we might be certain in that case that God would bring 
 about these desirable results. 
 
THEISM AND HAPPINESS 269 
 
 222. Thus, although the belief in a God, taken by 
 itself, cannot lead us to optimistic conclusions about 
 the universe, yet it might be possible to establish a 
 metaphysical position, which included the belief in 
 a non-omnipotent God and led to optimistic con- 
 clusions. But it might also be possible to establish 
 a metaphysical position, which would lead to equally 
 optimistic conclusions, and which did not include the 
 beHef in a God. Even with a God, we are not entitled 
 to be optimistic unless the nature of reality, inde- 
 pendent of God's vohtions and logically prior to them, 
 is such as to permit the predominance of good. Now 
 it is quite possible to maintain that the nature of 
 reality is such as to ensure the predominance of good, 
 independently of the volitions of a God. In that case 
 our optimism would not be affected if we denied God's 
 existence. 
 
 I cannot here examine whether the nature of reality 
 is such as to bring about the predominance of the 
 good, whether by means of God's volitions or inde- 
 pendently of them. The only way of answering this 
 question would be by the establishment of a complete 
 system of metaphysics. I merely wish to point out 
 that the existence of a God striving for the good is 
 neither sufficient nor necessary for an optimistic view 
 of the universe. There might be a God, and yet the 
 universe might be, on the whole, bad. There might 
 be no God, and yet the universe might be, on the 
 whole, good. 
 
 The statement that the existence of God is not 
 necessary for an optimistic view of the universe re- 
 quires some explanation. I only mean by it that it 
 is possible to conceive a theory of the nature of reality 
 which excludes the existence of a God, and yet leads 
 to an optimistic view of the universe. Such a theory 
 
270 SOME DOGMAS OF EELIGION 
 
 can be conceived, and, in fact, theoi'ies of this sort 
 have been held by great philosophers. ^ But, of 
 course, it might be contended that all such theories 
 break down on further examination, while the theory 
 which bases optimism partly on the belief in a God 
 does not break down. I do not agree with this con- 
 tention ; but I certainly do not wish to assume its 
 falsity. My position is merely that the belief in 
 God will not justify optimism, unless it is supported 
 by other metaphysical conclusions (which, as far as 
 our present investigation goes, may or may not be 
 true), and that there are metaphysical conclusions 
 (which, as far as our present investigation goes, may 
 or may not be true) which would justify optimism 
 without a beUef in God. 
 
 223. It follows from this that we must reject the 
 very usual assumption that theism, as such, is an 
 adequate basis for optimism, and the only adequate 
 basis. A man who believes in God, it is said, is 
 logically bound to take an optimistic view of the 
 universe, while a man who does not believe in God 
 has no right to take such a view. In fact, however, 
 the theist requires much more than his theism before 
 he is entitled to pronounce the universe good, and the 
 atheist requires much more than his atheism before 
 he can reject optimism as untenable. 
 
 What is the cause which leads so many people to 
 the view that optimism and theism are so closely 
 connected ? I believe it is the same as that which 
 makes the argument from design plausible — the dif- 
 ficulty of apprehending the position of Idealism. ^ 
 
 We may hold that all existent being — other than 
 
 ^ E. g. Fichte^ in his earlier system ; Mr. Bradley ; and, as it 
 seems to me, Hegel. 
 2 Cp. Sections 196-207. 
 
THEISM AND HAPPINESS 271 
 
 God, if there is one — is matter, or that some of it is 
 matter and some of it spirit, or that all of it is spirit. 
 If we hold that it is aU matter, we should have no 
 reason to take an optimistic view of the universe, 
 unless we supposed a God who directed and controlled 
 it. The nature of matter has no tendency to bring 
 about the good more than the bad. It may do so if 
 it is controlled by a person who desires to produce 
 the good. But, if it is left to itself, it can only realize 
 a good result by chance, and that it should do this 
 seems very improbable, since, so far as we can judge, 
 the arrangements df matter which would thwart the 
 good are much more numerous than the arrangements 
 which would promote it. 
 
 A similar result follows if we hold that, apart from 
 God, both spirit and matter exist. In this case spirit 
 exists in the midst of matter. It is continually affected 
 by matter, and it cannot carry out its designs — in many 
 cases, perhaps in all — except by employing the agency 
 of matter. Even if Spirit should aim at the good, 
 then, its success in attaining it will be conditional on 
 matter. And the extent to which such spirits as ours 
 can affect matter is extremely limited. We have experi- 
 ence every day sufficient to convince us that our efforts 
 cannot always make matter subservient to our pur- 
 poses. Unless the matter is controlled by a spirit 
 much more powerful than ours, the predominance of 
 the good will be largely dependent on the undirected 
 action of matter. And it will only be by chance — 
 and that an improbable chance — that the undirected 
 action of matter will realize the good, or permit it to 
 be realized. 
 
 Thus, if we accept the existence of matter at aU, 
 optimism, it may reasonably be held, does find the 
 existence of a God essential to it. Even if a God did 
 
272 SOME DOGMAS OF EELIGION 
 
 exist, it would not be certain that he could realize the 
 good, but without him it seems almost certain that the 
 good would not be realized. And the same result 
 would happen if we substituted for matter anything 
 else which .was not spii'itual. 
 
 Even if we take all the existent beings of the uni- 
 verse as spiritual, the question arises whether the 
 individual spirits are a mere aggregate, whose juxta- 
 positions and relations are decided by laws resembling 
 those of matter. If this is so, the chance of a predo- 
 minance of good in the universe, without a directing 
 God, is still not very large. It is larger than it would 
 be if matter existed, for the reasons which we con- 
 sidered in the last chapter, when speaking of the argu- 
 ment from design. But, for the reasons which were 
 also considered then, the probability is not great. In 
 order that a system of individuals shall be good, it is 
 necessary that the relations between them shall be 
 satisfactory. If these relations are of such a nature 
 that they are as likely to thwart the good as to promote 
 it, they will only promote it by chance. They may, 
 indeed, be modified by the efforts of the related indi- 
 viduals. But we do not know how far such efforts 
 will meet with insuperable obstacles. 
 
 224. Thus the existence of a God seems to be neces- 
 sary for optimism unless we are able to hold that all 
 existence is spiritual, and that all spirits form a system 
 such that it has a greater tendency to produce good 
 than to produce evil. And this is the metaphysical 
 view which is most removed from the attitude of every- 
 day hfe. 
 
 A civilized man who does not reflect on meta- 
 physical subjects almost always accepts both matter 
 and spirit as existent realities. He conceives himself 
 and his friends as spiritual, while their bodies and 
 
THEISM AND HAPPINESS 273 
 
 clothes and tables are material. Matter is lower and 
 less important than spirit, in his view, but it is equally 
 real. If he beheves, as he generally does, that the 
 universe is on the whole good, he believes it because 
 he holds that the universe is created and controlled by 
 a God. The intrinsic nature of the universe — or at 
 any rate of one side of it — is no more calculated to 
 produce the good, than the intrinsic nature of canvas 
 and paint is calculated to produce a beautiful picture. 
 
 K a man does not hold this view, he is more likely 
 to hold that the universe is completely material than 
 to hold that it is completely spiritual. The arguments 
 — such as they are — for the first view can be learned 
 and appreciated without the systematic study of meta- 
 physics, and without using any ideas but those 
 which are known to every educated man. The argu- 
 ments for the second view, though they are much 
 stronger, almost inevitably require a systematic study 
 of metaphysics, and introduce a way of looking at 
 things which is very unlike our everyday way of 
 regarding them, though it may be much more correct. 
 And if the entire universe is material, there is no 
 ground for optimism unless all this matter was 
 controlled and directed by a God behind and 
 above it. 
 
 Finally, of the different forms which can be assumed 
 by the belief that all existence is spiritual, the form 
 which makes it an aggregate of mechanically con- 
 nected selves is better known, and appears at first 
 sight simpler and more probable than the view which 
 makes it a really spiritual unity. The former seems 
 more probable because it has a greater resemblance to 
 the views of relation which are adopted in the physical 
 sciences, while the latter involves ideas which are 
 scarcely known except in metaphysics. As we have 
 
 T 
 
274 SOME DOGMAS OF EELIGION 
 
 seen, the former would almost inevitably require the 
 existence of a God to secure a desirable result. 
 
 It is not wonderful then that it should be so gener- 
 ally believed that optimism requires theism as a basis. 
 But it is quite possible to conceive the universe as 
 a spiritual unity of such a nature that it necessarily 
 manifests itself in more good than evil, or in continu- 
 ally increasing good and in continually diminishing 
 evil. In such a case an optimistic view of the uni- 
 verse would be justifiable, which did not depend on 
 the existence of a God. Such a theory must necessarily 
 take us far away from the standpoint of ordinary life. 
 It must make us look on matter as only an appearance 
 of spirit, and it must make us look on unity as equally 
 real with differentiation. It will not be easily under- 
 stood, or easily proved. But it may be true. 
 
 225. Before leaving our interest in God's works, it 
 may be worth while to discuss the relation of the belief 
 in God to the belief in immortality. These beliefs are 
 often supposed to be logically joined. If the existence 
 of God is accepted, it is often held that from this alone 
 — independent of any alleged revelation on the subject 
 — the immortality of man may be either proved, or, at 
 least, shown to be highly probable. And, on the other 
 hand, it is held that, if the existence of God is rejected, 
 not only has one argument for immortality gone, but 
 there is no chance of proving it at all, or rendering it 
 in the least probable. Atheism must necessarily 
 involve disbelief in immortality. 
 
 Let us consider the first of these assertions. Would 
 the certainty of God's existence support the belief in 
 human immortality ? If it does so, it must be on 
 account of God's goodness. His wisdom and power 
 might possibly be arguments for proving that he could 
 make man immortal if he wished to do so, but they 
 
THEISM AJJ^D HAPPINESS 275 
 
 could throw no light on the question whether he 
 would will to do so or not. The only way of deciding 
 this — and it is a way which is often taken — is to say 
 that immortality is very good, that its absence would 
 deprive the universe of almost aU its value and signifi- 
 cance, and then to argue that God, who is good, could 
 not refuse us this gift. 
 
 226. But if immortality is to be proved in this way, 
 it is necessary to be certain that it is good. And this 
 would be difficult. Many people do not desire immor- 
 tality. There are some who desire no sort of immor- 
 tality — who would rather cease to exist at the death of 
 their bodies than continue to exist in any way what- 
 ever. There are others who would not be unwilling 
 to encounter an immortal life of rest and tranquillity, 
 but who would prefer extinction to an immortality 
 which involved a continuance beyond death of the 
 pain and struggle of this life, even if that struggle 
 should be comparatively short, and should lead to 
 eventual repose. In the far East, again, we find two 
 of the greatest religions of the world teaching that 
 personal immortality is an evil, and that the highest 
 consummation of a wise and virtuous life will lead us 
 to lay down life altogether. 
 
 Even if we were able to put these difficulties on one 
 side, and assume that human immortahty was certainly 
 good, should we be entitled to base a belief in immor- 
 tality on a belief in God ? I do not see that we are 
 justified in this, whether God is held to be omnipotent 
 or not. 
 
 If God is omnipotent, then it is clear, as we saw in 
 Chapter VI, that there are some good things which he 
 does not antecedently will to exist, and some bad 
 things which he does antecedently will to exist. For 
 some good things do not exist, and some bad things 
 
 T 2 
 
276 SOME DOGMAS OF RELIGION 
 
 do. Now how can we tell that immortality is not one 
 of the good things which he does not will to exist, and 
 the annihilation of persons one of the bad things which 
 he does will to exist ? To reply that immortality is 
 better than most things, and annihilation worse than 
 most things may be true, but is irrelevant. For we 
 saw above that, if any evil in the universe was con- 
 sistent with the goodness of God, it was impossible 
 for us to determine any hmits to its amount. 
 
 But if God is not omnipotent, then his goodness is 
 no security for immortaUty. There are some good 
 things which we know that he cannot realize, since 
 we know that they do not exist. How can we be sure 
 that human immortality is not among such things? 
 There are some bad things which we know that he 
 cannot prevent from existing, since they do exist. 
 And among such things may be our annihilation at 
 the death of our bodies. 
 
 If immortality can be proved at all, it can only be 
 proved by arguments of a different sort. If we are to 
 do anything, we must be able so to determine the 
 nature of reahty as to show that every self is an 
 eternal part of the eternal reality, and that its eternity 
 necessarily manifests itself in an existence throughout 
 all future time, or we must, in some similar way, 
 prove the immortality of the self as a direct conse- 
 quence of its own nature. Such a line of argument, 
 if possible at all — I believe, as I have said earlier, that 
 it is possible — is difficult and arduous. But nothing 
 less will suffice. 
 
 Now it seems clear that such a deduction as this 
 would not be helped by a demonstration of God's 
 existence. If God is taken as director, and not as 
 creator, then my existence is independent of his will. 
 And if my immortality is a consequence of my nature, 
 
THEISM AND HAPPINESS 277 
 
 then my immortality is involved in my existence, and 
 is likewise independent of God's will. And in no other 
 way than by his will could the existence of a directing 
 God be supposed to bear on the question. The mere 
 fact that the universe contained a being who greatly 
 excelled me and all others in perfection could not be 
 a ground for believing that my existence would be 
 unending. 
 
 227. If we go back to the hypothesis that God is 
 a creator, and that, among other things, he creates 
 ourselves, then, so far from the existence of God 
 rendering my immortality more probable, it seems to 
 me that it makes it less so. The idea of creation is 
 generally held in a form which makes the created 
 being begin to exist at a point in time, so that there 
 was a time, however distant, when that being had not 
 yet begun to exist. Whether this form is essential to 
 the idea of creation or not, it is the form in which it 
 is almost always found. Indeed the hypothesis of 
 creation is often supported by the assertion that it is 
 impossible for finite beings to have existed through all 
 past time. Now if for every man there was a time 
 when he did not yet exist, I cannot see how his 
 immortaUty is to be proved. If his relation to the 
 universe is compatible with his non-existence in past 
 time, how shall we be able to prove that his relation 
 with the universe is not compatible with his non- 
 existence in future time ? Metaphysics have not yet 
 given us the right to discriminate in this manner 
 between the past and the future. 
 
 Of course the difference between the past and the 
 future is very considerable from the point of view of 
 our own interests. We are much more interested in 
 what is about to happen to us than we are in what 
 has happened to us. And this might be important, if 
 
278 SOME DOGMAS OF EELIGION 
 
 the proof of immortality rested on the goodness of the 
 universe or of its creator. To annihilate existing 
 persons might well be a much more serious blemish 
 on the universe than to postpone the creation of 
 persons not yet existent. And so, if immortality 
 could be based on this ground at all, we might prove 
 that a being would never cease to exist in the future, 
 although he had formerly not existed. But we have 
 seen that immortality cannot be based on this ground, 
 and on any other it seems impossible to give any 
 reason why that which was once created should not 
 be again annihilated.* 
 
 228. Thus we see that there is no logical connexion 
 between the belief in God and the belief in human 
 immortality. And there has not always been a histori- 
 cal connexion. Fichte, in his earlier system at any 
 rate, believed in immortality without believing in God. 
 The same may be said, in my opinion, about Hegel, 
 though this is disputed. Buddhism, again, which 
 has no God, holds immortality to be the natural state 
 of man, from which only the most perfect can escape. 
 And, in modern times, Schopenhauer is in the same 
 position. On the other side we find Lotze. Of all the 
 theists of the nineteenth century he is philosophically 
 the most important. And he regards immortality as 
 quite undemonstrable and as very doubtful.^ 
 
 What is the cause of the opinion that a belief in 
 immortality requires a belief in theism? Partly, 
 perhaps, it is the fact that the majority of theists do 
 believe in immortahty, and that the majority of 
 Westei^n believers in immortality are theists. But, 
 
 ^ Cp. chap, iv, 
 
 ^ Lotze's defence of theism will be found in his Microcosmus 
 (Book IX, chaps, iv and v) and his doctrine of immortality in 
 his Metaphysics (Section 245). Both these works are translated 
 into English. 
 
THEISM AND HAPPINESS 279 
 
 in addition to this, it must be remembered that 
 materialism would make any belief in immortahty 
 perfectly unreasonable, and that scepticism makes all 
 beliefs unreasonable.^ Now there is a very common 
 idea that an atheist must either be a materialist or 
 a sceptic, and, therefore, that it is unreasonable for 
 him to believe in immortahty. But this, Hke many 
 other common ideas, is erroneous. 
 
 229. We must now pass from our interest in God's 
 works to our interest in his person. This interest may 
 be due, in the first place, to the guidance which, it is 
 maintained, we can procure by imitating a character 
 so superior to our own. In this case argument is still 
 possible. For the existence of God is here desired 
 as a means to an end which we should all admit to 
 be good — namely, our guidance to right action. The 
 question is only how far the contemplation of the 
 character of an existent God is a necessary means 
 to that purpose. And this is a question wliich admits 
 of argument. 
 
 If the belief in God is rejected, we lose along with 
 it the behef in the present existence of any person 
 who is, from a moral point of view, absolutely perfect. 
 The behef in a being who was morally absolutely 
 perfect, but whose wisdom and power were too limited 
 to allow him to be called God, is perhaps not absolutely 
 impossible. But it seems impossible to see how the 
 existence of such a being, as distinguished from one 
 whose perfection was very great, could be proved. And, 
 historically, absolute moral perfection has never been 
 attributed to any person except God, unless that person 
 
 ^ Materialism — the belief that all existent reality is material — 
 must be distinguished from the belief mentioned above (Section 
 223) that all non-divine existent reality is material. 
 
280 SOME DOGMAS OF EELIGION 
 
 was conceived as the incarnation of God, or as his 
 specially favoured minister, (In the following discus- 
 sion I will, for the sake of brevity, use 'perfect' to 
 mean morally perfect.) 
 
 230. But, as we have already seen,^ this loss, if 
 it is one, does not depend on the denial of God's 
 existence. Even if a God does exist, he could not be 
 absolutely perfect so long as the society of selves of 
 which he forms a part is by no means perfect. It is 
 impossible that any one member of the society should 
 be unaffected by the defects of his fellows. And, 
 if this is so, the disbelief in God's existence loses us 
 in this respect nothing which ought not to have been 
 lost already. 
 
 God might be looked at, no doubt, as destined to 
 become eventually perfect. But he could only become 
 so if all other persons were perfect too. And, if they 
 were perfect, they would need no moral guidance. It 
 would be superfluous for them to look to God for an 
 example of moral perfection if they were equally 
 examples of it. 
 
 StiU God, though not at present perfect, would at 
 any rate be much better than any other person. If we 
 were certain that God existed, we should be certain of 
 the existence of a being whose character, so far as it 
 was known to us, and so far as its lessons were applic- 
 able in our circumstances, would be a far safer ideal 
 for our actions than the character of any man could be. 
 If we lose the belief in God's existence, we lose this 
 ideal. Let us inquire how serious the loss would 
 prove. 
 
 231. There are cases, to begin with, in which it is 
 clear that it is possible to see what is right, without 
 having an ideal in a God conceived as existent. Many 
 
 1 Section 210. 
 
THEISM AND HAPPINESS 281 
 
 pagans — and many other people — practise virtues of 
 which no trace is found in the conduct which they 
 attribute to their Gods, and would shrink with horror 
 from vices to which they believe their Gods addicted. 
 Atheists, again, who believe in no God at all, can distin- 
 guish between right and wrong, and their judgements 
 would often be accepted as correct by theists. The 
 possession of a divine ideal, then, is not always essen- 
 tial for the knowledge of the right. 
 
 In many cases, however, the knowledge of a person 
 good enough to serve as an ideal is of the greatest 
 value, not only as a stimulus, but as a guide. It may 
 be so, I conceive, in two ways. In the first place, it 
 may suggest to us that a quality is desirable which we 
 should never have thought desirable if we had not 
 seen a person who possessed it. It is quite possible, 
 for example, that the goodness of self-control in grief 
 might not occur to some people until they saw a man 
 who practised self-control, and that it would be 
 acknowledged by them when it did occur to them. 
 In the second place, when something has been re- 
 cognized as desirable, a person who exhibits that 
 quality may indicate to us by his example shades of 
 that quality, too delicate to be determined by any 
 abstract rule. It does not need the contemplation of 
 a courteous person to teach us that courtesy, in the 
 abstract, is good. But for most people there is only 
 one way of learning in detail what courtesy is — to 
 live with those who are courteous already. 
 
 232. It must be admitted then that a personal ideal 
 — by which I mean a person believed to exist, not 
 merely imagined as possible — may be of great use as 
 a guide to action. But then such an ideal need not 
 be God. It can be another man. And it very frequently 
 is another man. We are all of us guided, more or less, 
 
282 SOME DOGMAS OF EELIGION 
 
 by our knowledge of the characters of other men, whom 
 we recognize as our superiors in certain qualities, and 
 whom we endeavour to imitate in respect of those 
 qualities. 
 
 Thus to lose God would not involve losing a personal 
 ideal. It might be objected that a man who had 
 reached higher than any of those known to him in any 
 particular direction could not have a personal ideal in 
 that direction — unless he mistakenly followed a man 
 lower than himself, which might lead him wrong. To 
 this it may fairly be answered that the imitation of 
 another person, though often useful, is not indispensa- 
 ble to right action, and that a man who had got so high 
 as this would probably be able to see what was right 
 without waiting for any one else to show him the way. 
 Nor would it follow that such a man had not the advan- 
 tage of feeling reverence and veneration. Different 
 good qualities are distributed in different ways, and 
 a man who would be mistaken if he thought that he 
 knew of any one braver than himself might perhaps 
 find a man worthy of his imitation in respect of 
 industry or unselfishness. 
 
 At the risk of appearing fanciful, I will also suggest 
 that it is not impossible that two men might each make 
 an ideal of the other, with respect to the same quality, 
 and yet each might have found a true ideal. A would 
 think B excelled him in that quahty. B would think 
 he was excelled by A. Both could not be right. But, 
 if they happened to love each other, each might see in 
 the other the eternal perfection of which the temporal 
 imperfection — which is all a man can see of himself — 
 is only an imperfect manifestation. To pursue this 
 possibility, however, would take us too far into the 
 metaphysics of personality. 
 
 233. But non-divine beings, it may be said, must 
 
THEISM AND HAPPINESS 283 
 
 always be defective as personal ideals, unless supple- 
 mented by an ideal of a divine person. I may take 
 the wrong man as my ideal. Or, even if I take the 
 right man, the best man whom I know, still he may 
 be mistaken just on the point on which I am imitating 
 him, and so I may be misled. 
 
 Doubtless this may be so. But we are no better off 
 with a divine ideal. It is true that there cannot be 
 a better ideal than God, if he exists, and that the 
 chance that he is acting wrongly in any particular case, 
 though it exists, is much smaller than it is in the case 
 of any particular man. But, on the other hand, we 
 are much more likely to make mistakes about what 
 God's character is than we are in the case of a man. 
 We can only gather God's character from his acts. 
 And as to those acts there is much doubt. It has been, 
 for example, a much disputed point whether he has or 
 has not predestined certain of his creatures to eternal 
 damnation. Even if we know what he has done, we may 
 not be certain whether he has willed it for its own 
 sake, or as a means ; and, if as a means, we may not 
 be certain for what end. Now this makes all the 
 difference as to his character. If God has caused pain 
 to a wicked man, for example, we can tell nothing from 
 this as to his character, if we do not know whether 
 he did so simply because the man was wicked, or to 
 reform him, or to deter others, or to spare him greater 
 pain, or for some other reason. 
 
 If we believe that there is a God, we shall certainly 
 beheve that he has acted well, since a goodness ex- 
 ceeding that of all other persons was part of the defini- 
 tion of God. But this belief, by itself, will give us no 
 guidance. If we know that God has acted well, and 
 also know why he has acted, we shall know what 
 good action is under certain circumstances, and this 
 
284 SOME DOGMAS OF EELIGION 
 
 may afford us guidance. But if we merely know that 
 God acted well, without knowing what he did or 
 why he did it, we shall get no more guidance from the 
 nature of God than we could have got from our abstract 
 notion of the good, without reference to God at all, 
 
 234. And there is another difficulty which meets us 
 if we try to take God as the ideal which is to guide 
 our actions. His circumstances, and his character in 
 non-moral aspects (such as his wisdom and his power) 
 are so different to ours that it is very difficult to apply 
 his example so as to guide us in our difficulties. The 
 example, no doubt, is not altogether irrelevant. The 
 moral nature and the moral problems of one individual 
 have always some analogy to those of any other. And 
 if God is conceived as not omnipotent, and not abso- 
 lutely perfect, though much more powerful and more 
 good than any other person, then we can conceive that 
 his duties, his perfections, his imperfections, his temp- 
 tations, would have, so far as we could know them, 
 many lessons for our guidance. But it is obvious that 
 they would have far less application than those of a 
 person whose conditions resembled our own more 
 closely. If there is a God, he would be morally much 
 superior to Socrates. But Socrates was sufficiently 
 superior to most of us to afford us an ideal. And 
 Socrates acted under conditions so much more like 
 our own than those under which God would act, that 
 it seems to me we should derive more guidance from 
 his example than from God's — even if we knew as 
 much about the character of God as we do about the 
 character of Socrates. 
 
 235. This difficulty is, to a great extent, removed if 
 God is considered to be incarnate in some man or 
 men whose history is known. The conditions under 
 which an incarnate God acted would be far more like 
 
THEISM AND HAPPINESS 285 
 
 the conditions under which an ordinary man acts than 
 the conditions under which an unincai'nate God acted 
 could be. Thus we could obtain more guidance. In 
 the Christian reb'gion, for example, we see that moral 
 guidance is more frequently sought from the earthly 
 life of Jesus than from the action of God in creating 
 and governing the world. And this is, doubtless, one 
 of the reasons — I do not say the chief reason — for the 
 the singular fascination of the idea of an incarnate 
 God. 
 
 But what guidance shall we lose if we hold that 
 a person who has been regarded as an incarnation of 
 God was in truth an ordinary man? This opinion 
 will not prove that such a man did not exist, or that he 
 was not a good man. All that goodness, which history 
 tells the believers in his divinity that he exhibited, 
 may still be held — if there is sufficient evidence- — to 
 be his. The believers in his divinity, indeed, may 
 infer from his divinity the existence of greater good- 
 ness in him, of which no manifestations are recorded 
 in his history. And those who reject his divinity 
 will have no grounds to infer such greater goodness. 
 But goodness which is merely inferred cannot give 
 us moral guidance, for we cannot infer it unless we 
 already know it to be good. It is only goodness which 
 is observed — either directly or through history — which 
 we can recognize as good without having previously 
 known of its goodness. And so it is only goodness 
 which is observed which can give us moral guidance. 
 Whatever goodness is observed in a person who is 
 believed to be an incarnation of God will not cease 
 to be there, or to be good, if he is believed to be 
 a man of the ordinary kind. And thus he will not 
 be a less satisfactory ideal for moral guidance. 
 
 * But,' it may be said, ' if you no longer believe this 
 
286 SOME DOGMAS OF EELIGION 
 
 man to be an incarnation of God, you cannot know 
 that his actions are good, and so you cannot be safe 
 in taking them as your ideal,' I conceive that this 
 objection may be validly answered by a dilemma. 
 When we observe an action, either we are competent 
 to judge of its goodness or we are not. In so far as 
 we are competent, then we can pronounce it good, 
 even if its author is not an incarnation of God. But 
 in so far as we are not competent, then it is impossible 
 for us to tell whether the being of whom he is as- 
 serted to be an incarnation is properly called God. 
 For no being is to be called God unless he is good, 
 and what evidence can we have of the goodness of the 
 being incarnated, except that the action of the being 
 who incarnates is good ? 
 
 And if the incarnation is not the incarnation of a 
 good being it can give us no moral guidance. How- 
 ever wise and powerful the being who was incarnated 
 might be, we should have no reason to follow his 
 example if he was not good. To follow the example 
 of any person merely because he was wise would be 
 foolish, since he might be using his wisdom as a means 
 to bring about undesirable ends. And to follow the 
 example of any person merely because he was power- 
 ful would be contemptible. It would not even be 
 prudent, unless we were reasonably certain that he 
 would be pleased with flattery and servility. 
 
 236. Moreover, the example of an incarnate God, 
 although much more useful for our guidance than that 
 of an unincarnate God, is much less applicable than 
 the example of an ordinary man. For the position of 
 an ordinary man is much more like ours, who are 
 ordinary men, than is the position of a being who is 
 the incarnation of God. We have reason, for example, 
 to believe that Jesus behaved with courage and dignity 
 
THEISM AND HAPPINESS 287 
 
 at his trial and execution. This fact has much more 
 value for those who do not believe in his divinity than 
 for those who do. For the former class it shows — as, 
 of course, many other examples show — that man can 
 rise above pain and the approach of death. And what 
 men have done, we, who are men, may hope, if the 
 need comes, to do also. At any rate, the fact that 
 it has been done by a man raises humanity. But it 
 by no means follows that, because these things can be 
 done by a being who is both ' perfect God and perfect 
 man' that they can be done by us, to whom this 
 description does not apply. 
 
 Again, most people would now be inclined to admit 
 that Jesus was right in decUning to be bound by 
 various rules of his national church — as to the strict 
 observance of the Sabbath, for example. And, if he 
 is considered as simply a man, we may derive guidance 
 for ourselves from our perception that he was right as 
 against his contemporaries in these matters. But, if 
 we accept his divinity, I do not see that we have any 
 guidance. For it raight well be that a being who was 
 both God and man might be entitled to dispense with 
 rules, while beings who were only men were not. (I 
 am speaking here, of course, only of what may be 
 gained from example. The question of explicit teaching 
 is therefore irrelevant.) 
 
 God could, no doubt, be held to be incarnate without 
 adopting the extreme view that the being in whom it 
 took place was perfect God and perfect man. And, 
 in so far as that being resembled an ordinary man, in 
 so far his example would be more applicable for our 
 guidance. But then, in so far as he resembled an 
 ordinary man, in so far his example would only help 
 us in the way that the example of an ordinary man 
 woxild help us, and so we should lose nothing by 
 
288 SOME DOGMAS OF EELIGION 
 
 denying that he was an incarnation of God, and con- 
 sidering him merely as a man. 
 
 Some systems of philosophy, again, treat eveiy man 
 as an incarnation of God. In this case it is clear that 
 whether the doctrine is true or not, we can hope for 
 no special guidance from it. Every possible example 
 offered to us will be that of an incarnation of God, and 
 many of them will lead us in different directions. 
 The question will still remain which of them we ought 
 to follow, and which to avoid. 
 
 237. It would seem then that we should lose little, 
 if anything, in the way of moral guidance by rejecting 
 a belief in God. But moral encouragement is another 
 matter. If any man finds that he is encouraged to do 
 right by a belief in God's existence, taken in itself 
 and not as a means of producing the good, or guiding 
 men's actions, this is ultimate — in his case, and while 
 it lasts. If a man feels morally helped by hearing 
 music of a particular sort, and that sort only, it would 
 be foolish to maintain that such a man lost nothing 
 when he could hear no music, or only music of 
 another sort. And, in the same way, a man who feels 
 discouraged in right action when he rejects the behef 
 in a God, does lose something by his rejection. 
 
 Not all men would feel this, but some would. At 
 the same time it must be noticed that it does not 
 follow that the loss would be permanent in the cases 
 in which it is real. A man who has been accustomed 
 to do right in the name of God may feel a loss when 
 he ceases to beheve in God. So a man who has been 
 accustomed to be loyal to his country in the name of 
 a king may feel his loyalty shaken if his country 
 becomes a republic. But loyalty, as the history of 
 republics shows us, can do very well without a king, 
 when the king has once gone. And so it may be — 
 
THEISM AND HAPPINESS 289 
 
 I do not say that it always will be — with morality and 
 God, 
 
 238. And the rejection of the belief in God may 
 bring discouragement of a different sort. Without 
 depriving us of stimulus, it may deprive us of comfort. 
 It may well be that some people find some consolation 
 in the thought that the struggle for the good has a 
 leader who can, at any rate, survey the whole field, 
 of which each of us sees but a corner. We have 
 seen that the mere existence of such a leader tells us 
 nothing about the results of the struggle. It may be 
 defeated with such a leader, or it may be victorious 
 without one. But if a man finds some comfort in the 
 existence of a leader for its own sake, and without 
 reference to the result, then his comfort will be 
 diminished by the rejection of the belief in a God. 
 Here, as in the last case, it is possible that a desire 
 which cannot be removed by argument may be re- 
 moved by time. And here, as in the last case, it is 
 also possible that it may not be removed. 
 
 239. In a universe without a God there wiU be no 
 one to worship, and there will be one person less to 
 love. And the loss — if any — which results from this 
 change is again a thing which each man must estimate 
 for himself. With regard to worship, however, it is 
 to be remembered that no metaphysical conclusion can 
 deprive us of the power to feel reverence, and to feel 
 it justly, because it is certain upon any metaphysical 
 theory that there are many men who are worthy 
 of reverence. What we shall lose is the opportunity 
 of reverencing a person who is held to excel aU other 
 persons in a very high degree. The loss in this 
 will depend very much on what each man finds to 
 reverence in his fellow men. And it must also be 
 remembered that an object of reverence is only of 
 
 u 
 
290 SOME DOGMAS OF RELIGION 
 
 value to us in so far as we appreciate it. God is a 
 better object of reverence for me than Socrates, not in 
 proportion as he is superior to Socrates, but in pro- 
 portion as I can reahze his superiority to Socrates. 
 And, if I cannot realize his superiority to Socrates at 
 all, then, even if God exists, my reverence for Socrates 
 may be of more value than my reverence for God, In 
 the same way, a boy's reverence for the captain of the 
 eleven may be of much more value than his reverence 
 for God, however sincere the latter may be. For it is 
 possible that his idea of the captain of the eleven may 
 be one much more adequate to excite reverence than 
 his idea of God. 
 
 240. There remains love. By love of God I mean 
 something entirely distinct from reverence and admira- 
 tion and gratitude. I mean a feeling of one person for 
 another, which is not unworthy to bear the same name 
 as the feehng of friend for friend. That, of course, 
 must go, if it is believed that the person that was loved 
 never existed. 
 
 Love will not cease. There are other persons to 
 
 love. And the non-existence of God would leave it 
 
 as possible as it was before that love should be 
 
 the central fact of all reality. It might still be true 
 
 that nothing else had value. It might still be true 
 
 that nothing else had existence, except lovers and 
 
 their love,^ But some love would have been poured 
 
 out on a dream — or rather a reflexion — which could 
 
 not return it. Whether the friends whom all men 
 
 may find could compensate for the friend whom some 
 
 men thought they had found is a question for each 
 
 man to answer. It is a question which can never be 
 
 answered permanently in the negative while there is 
 
 still a future before us. 
 
 ^ I have endeavoured to develop this theory in my Studies in 
 JJegelian Cosmology^ chap, ix. 
 
CONCLUSION 
 
 241. The result of our investigations has been 
 almost entirely negative. We have, indeed, come to 
 the conclusion that dogma is important. But when 
 we inquired how it could be established, we only 
 found that some of the most common ways of establish- 
 ing it were inadmissible. We found reason to think 
 that immortality, if true, would involve pre-existence, 
 and that some of the usual objections to the truth of 
 immortality were untenable. But we found no reason 
 for a positive belief that immortality was true. Again, 
 we were able to see that there was no reason why 
 human volitions should be exceptions to any law of 
 the complete determination of events, but we did not 
 establish any such law. And our conclusion that 
 there was no reason to suppose that God existed gave 
 us no information as to what did exist. 
 
 All this was inevitable. The only way of coming to 
 any conclusions on matters of religious dogma is by 
 means of metaphysical arguments. This was the 
 result of our consideration of the question in Chapter II. 
 And since I have not put forward here any positive 
 metaphysical position, it was impossible to hope for 
 positive conclusions. Negative conclusions on matters 
 of dogma can be arrived at without the standpoint of 
 a positive metaphysical theory. We may point out 
 that the arguments for a certain dogma are untenable, 
 and that therefore there is no reason to believe it, or 
 we can go further and point out that the dogma is self- 
 contradictory, or incompatible with indubitable facts, 
 
 U 2 
 
292 SOME DOGMAS OF EELIGION 
 
 and must therefore be false. We can do this, even if 
 we are not prepared to establish any other dogma on 
 the same subject. 
 
 To refute a dogma may thus sometimes be easy, 
 while to establish a dogma is always hard, especially 
 as the subject-matter of various dogmas is so closely 
 connected, that it is difficult to establish the truth as 
 to one of them unless we can also determine the truth 
 about all the rest. And in order to establish such 
 dogmas as may form a basis for religion it seems neces- 
 sary to establish a complete system of metaphysics. 
 We need, for religion, to be able to regard the universe 
 as good on the whole, and it does not appear how we 
 could do this, except on the basis of a general theory 
 as to the ultimate nature of reality. 
 
 242. We are sometimes justified in basing action or 
 emotion on a proposition whose truth we have not our- 
 selves investigated. A man who has never examined 
 the evidence for the law of gravitation may reasonably 
 act on the assumption that it is true, because the con- 
 sensus of opinion among competent students of the 
 subject is so overwhelming that he has good grounds 
 to beheve it on their authority. But we are not justi- 
 fied in assuming in this manner the truth of dogmas. 
 For on these there is not the necessary consensus. 
 No dogma — at any rate, no dogma of rehgion — is 
 asserted which is not also denied by able students. 
 
 It follows that a man is not entitled to beheve 
 a dogma except in so far as he has investigated it for 
 himself. And since the investigation of dogma is a 
 metaphysical process, and religion must be based on 
 dogma, it follows further that no man is justified in 
 a religious attitude except as a result of metaphysical 
 study. The result is sufficiently serious. For most 
 people, as the world stands at present, have not the 
 
CONCLUSION 293 
 
 disposition, the education, and the leisure necessary 
 for the study of metaphysics. And thus we are driven 
 to the conclusion that, whether any religion is true or 
 not, most people have no right to accept any religion 
 as true. 
 
 This result is serious because, if this theory is true, 
 it will probably win increased, though not general, 
 acceptance, as inquiry in matters of religion becomes 
 more general, and the weight of authority and tradi- 
 tion becomes less. The number of people who wish 
 to hold a religion, but are xinable to do so, will become 
 larger. And this wiU increase the amount of human 
 suffering. 
 
 But it seems inevitable. What people want is a 
 religion which they can beheve to be true.^ Since 
 they are confronted on all sides with religions different 
 from their own, and with the denial of all rehgion, it is 
 inevitable that they should ask themselves why they 
 beheve their religions to be true. And when the 
 question is once asked, what can avert a widespread 
 recognition that the truth of rehgion can only rest on 
 foundations too controversial to be taken on trust, and 
 too obscure for many people to investigate ? 
 
 The result may be evil, but that is, unfortunately, 
 no ground for denying its truth. It is no more evil 
 than cancer, famine, or madness, and these are all 
 real. But it is not as evil as may appear at first 
 sight. 
 
 243. In the first place, supposing that the universe 
 as a whole was good, so as to form a satisfactory basis 
 for a rehgion, this would not be affected by the fact 
 that most people on this planet at present were not 
 entitled to beheve it, and that a large number of them 
 
 ^ The sense in which it can be said that a religion is true is 
 explained in Section 8. 
 
294 SOME DOGMAS OF RELIGION 
 
 did not believe it. Eeligion involves believing in 
 something good, and the belief in religion is itself 
 something good. If a religion is true, but not believed, 
 the second good vanishes, but not the first. Thus if 
 it should be true that all men are immortal, and are 
 progressing to a state of great perfection — to take one 
 dogma of religion which is sometimes maintained — 
 the evil inflicted on any of them by their present lack 
 of religion would be only temporary. Like all other 
 evils it would be destined to be removed some day. 
 
 It is necessary to be clear on this point, because the 
 two goods are occasionally confounded. We are asked, 
 for example, to consider how evil the world would be 
 without the belief in immortality, and, in expounding 
 this, a confusion is occasionally made between the 
 happiness which we have before death in believing in 
 immortahty, and the happiness which we may have 
 after death if immortality is true. But the latter 
 would not be lost by losing the behef in immortality, 
 for men would not perish when their bodies perished 
 because they had previously believed that they would 
 do so. 
 
 244, In the second place, it must be remembered 
 that the man who has no religion cannot have a bad 
 one. If the mass of Englishmen ceased to beheve in 
 any religion, many of them would lose much happiness 
 by ceasing to believe in heaven, but many of them 
 would gain much happiness by ceasing to believe in 
 hell. If a loss of happiness would result from the loss 
 of belief in God, there would be a gain, both in happi- 
 ness and otherwise, in freedom from the belief that 
 the creator of such a world as this was omnipotent, 
 and was yet to be worshipped. Loss of religion might 
 be an evil to the average man, but would certainly 
 not be an unmixed evil. 
 
CONCLUSION 295 
 
 245. And, thirdly, the extent to which rehgion 
 would be lost may be exaggerated. The study of 
 metaphysics will perhaps never be very common, but 
 it may be more common in the future than it is at 
 present. The world's leisure is increasing, and much 
 of it may be devoted to study. And if study at present 
 is rarely study of metaphysics, that is largely because 
 metaphysics seems unpractical. If, however, people 
 find that they cannot have religion without it, then 
 it will become of all studies the most practical. Its 
 results, indeed, may not be raore practically useful 
 than those of some other subjects. For some results 
 of study are, in our present civilization, essential to 
 life, and life is a condition precedent of religion. But 
 elsewhere we can enjoy the results without investigat- 
 ing them ourselves. I can eat breads although I have 
 never learnt to plough or bake. I can be cured of an 
 illness, though I have never learnt medicine. But if 
 — and this is the case at present- — I have no right to 
 rely on any metaphysical result which I have not 
 myself investigated, then the study of metaphysics 
 will be for many people the most momentous of all 
 studies. And this may produce important resiilts. 
 For, after all, one great reason why so few people 
 have reached metaphysical conclusions for themselves, 
 is to be found in the fact that so few people have tried 
 to reach them. 
 
 Once more, if our interest is for the happiness of 
 mankind, we may console om-selves with the reflection 
 that the large majority of men, while human nature 
 remains what it is, are not likely to give up traditional 
 opinions merely on the ground that they have no 
 logical right to hold them. Like opium-eaters, they 
 will preserve their happiness at the expense of their 
 intelligence, though, more fortunate than opium-eaters, 
 
296 SOME DOGMAS OF KELIGION 
 
 their dreams will not unfit them for practical life^ 
 and will — sometimes — not quit them before death. 
 And when the average man has changed so much 
 that he rejects all beliefs that he is not logically 
 entitled to hold, he may well have changed so much 
 as to have some logical right to a religion. 
 
 Finally, it is possible that a time may come when 
 metaphysics may attain the same certainty in a higher 
 sphere which is now often reached by science in 
 a lower sphere. If there was the same consensus of 
 expert opinion that immortality, for example, could be 
 proved as there is now that the law of gravitation can be 
 proved, men in general might accept the truth of immor- 
 tality without investigating it, in the same way that 
 they now accept the law of gravitation without investi- 
 gating it. And in this way, perhaps, sufficient dogmas 
 might be accepted to form the basis for a religion. 
 
 There is no impossibility in this. It would be rash 
 to infer that metaphysics will never pass out of the 
 controversial stage because they have not done so yet. 
 But they have not done so yet, nor is there any sign 
 that they are about to do so. Any school of philosophy 
 which has for a few years a very marked predominance 
 over its rivals is tempted to teach its conclusions dog- 
 matically to those who have no time or capacity for 
 its arguments, and to assure them that they can take 
 the results on trust because now, at last, all philosophic 
 opinion is on the same side. But a few years are 
 sufficient to show that such agreement as had ever 
 existed was only temporary. There is not now, and 
 there is not likely to be soon, such an agreement on 
 metaphysics as would justify any man in accepting 
 any dogmas on authority. 
 
 Many dogmas, indeed, are supported by such a 
 weight of authority that, even without investigating 
 
CONCLUSION 297 
 
 them, we may reasonably conclude that they are not 
 transparently absurd, and that there is much to be 
 said for them. And this is something. But we want 
 more than this — especially when incompatible dogmas 
 are supported in this way, as is often the case. 
 
 246. In spite, then, of the alleviations which I have 
 just pointed out, we are here confronted with one 
 of the great tragedies of life. Many men desire pas- 
 sionately to know the truth as to the great problems 
 of religion. And no man may believe any solution of 
 these problems to be true unless he has tested it him- 
 self. Even if he tests for himself, and comes to some 
 conclusion, his conclusion must lack that confirmation 
 by the unanimous agreement of inquirers which plays 
 so great a part in knowledge elsewhere. 
 
 This is sad, and would be sad even if dogma were 
 not essential to religion. For whether dogma is essen- 
 tial to religion or not, it is clear that many men do 
 desire, and will desire, to know the truth about such 
 dogmas as the existence of God and the immortality 
 of man. 
 
 It would not, perhaps, be a dispiriting conclusion 
 that truth on these great matters can only be attained 
 by long and toilsome efforts, if we were certain that it 
 would some day be attained. But this, certainty must 
 itself be dependent on dogma, and cannot, therefore, 
 be available for the comfort of those who have not 
 attained it. It is evident that various persons die in 
 a state of uncertainty about these dogmas, and others 
 are certain of incompatible conclusions, so that some 
 at least must be wrong. We should have therefore to 
 be certain, at least, of a hfe beyond death, if we were 
 to be certain that we should all eventually know the 
 truth on such subjects. 
 
 247. It is sad, but I do not know why it should be 
 
298 SOME DOGMAS OF RELIGION 
 
 thought strange. Is knowledge so easy to get that the 
 highest and deepest of all knowledge is likely to be 
 had for the asking? Or is everything good so com- 
 mon, that we should expect that religion— almost the 
 best of all earthly things — should be never absent where 
 it is desired ? 
 
 The matter would be different if we held to the old 
 opinion that it was a sin not to reach the true religion 
 in this life. For then religion must be open to all who 
 choose to take it, and cannot be dependent on meta- 
 physical abihty, which does not depend on the will. 
 And again, the matter would be different if the true 
 religion was essential to morality. For it is notorious 
 that the comprehension of metaphysics is not essential 
 to morahty. But it would seldom be said now that 
 the attainment of true religion was either a duty or 
 a condition of dutifulness, but only that it was a great 
 happiness. And, in this case, what right have we to 
 expect that it will not be rare ? 
 
 An exclamation reported ^ of Jesus has been, rather 
 unfairly, twisted into a canon of knowledge declaring 
 that the kingdom of heaven is hidden from the wise 
 and prudent and revealed unto babes. Such a prmciple 
 is sure to be popular, for it enables a man to believe 
 that he is showing his meekness and humility by the 
 confident assertion of propositions which he will not 
 investigate and cannot prove .^ Yet some other words 
 
 ^ Matt. xi. 25 ; Luke x. 21. 
 
 ^ This criticism applies, naturally, not to the teacher who made 
 the exclamation, but to the teachers who turned it into a principle. 
 It may be remarked that the principle is applied rather capriciously. 
 Many men would be prepared to use it as a ground for distrusting 
 Hegel, and for trusting the peasantry of Ireland or Wales (as the 
 case may be). But they would not admit it as a ground for 
 accepting the peasantry of Morocco as a safer guide in religion 
 than Thomas Aquinas. And yet Thomas Aquinas was a wise 
 man, and the religion of a peasant in Morocco would be eminently 
 
CONCLUSION 299 
 
 reported ^ of the same teacher might be remembered. 
 ' The kingdom of heaven is hke unto a man that is 
 a merchant seeking goodly pearls : and having found 
 one pearl of great price, he went and sold all that he 
 had, and bought it.' All that he had, — but if it is 
 too little ? The greater the price, the fewer can pay it. 
 
 Sixteen centuries after the death of Jesus, the Jewish 
 race produced another great religious teacher, in whom 
 philosophical insight and religious devotion were 
 blended as in no other man before or since. 'If the 
 way' — so he ends his account of the beatific vision 
 which frees man from all sorrow and all sin, and makes 
 death the least of all things — ' if the way which I have 
 pointed out as leading to this result seems exceedingly 
 hard, it may nevertheless be discovered. , . . But all 
 things excellent are as difficult as they are rare.' 
 
 Perhaps it will not always be so. Perhaps time as 
 it goes on will bring its imperfection nearer to the 
 perfection of eternity. But, here and now, dare we 
 deny that Spinoza is right ? ' Omnia praeclara tarn 
 difScilia quam rara sunt.' 
 
 childish. But, of course, it is only ignorant orthodoxy that is 
 childlike simplicity. Ignorant heterodoxy is childish supersti- 
 tion. 
 
 ^ Matt. xiii. 45. 
 
 THE END 
 
OXFORD : HORACE HAKT 
 TRIKTEU TO THE UNIVEKSITY 
 
Telegrams : 41 and 43 Maddox Street, 
 
 ' Scholarly, London.' Bond Street, London, W., 
 
 February, 1906. 
 
 Mr. Edward Arnold's 
 
 List of New Books. 
 
 A STAFF OFFICER'S SCRAP-BOOK. 
 
 By LIEUT.-GENERAL SIR IAN HAMILTON, K.C.B. 
 
 Demy %vo. With Illustrations, Maps, and Plans. 
 iSs. net. 
 
 General Sir Ian Hamilton's story of the operations of the 
 Japanese First Army, under Marshal Kuroki, at which he was 
 present as British Attache, is generally recognised as the best 
 and most authoritative account accessible to English readers. 
 
 The Sixth Thousand is practically exhausted — a remark- 
 able testimony to the welcome accorded to this fascinating 
 work, which is described by The Times as by far the most inter- 
 esting book on the Russo-Japanese War that has yet appeared 
 from the pen of an eye-witness. 
 
 LONDON : EDWARD ARNOLD, 41 & 43 MADDOX STREET, W. 
 
2 Mr. Edward Arnold's List of New Books 
 
 SOME DOGMAS OF RELIGION. 
 
 By JOHN ELLIS McTAGGART, Litt.D., 
 
 Lecturer in Moral Sciences, Trinity College, Cambridge. 
 
 Demy Svo, los. 6d. net. 
 
 This book attempts to prove, in the first place, that beliefs as to 
 certain matters of fact beyond our empirical experience are essen- 
 tial for religion, and of fundamental importance for human life. 
 Secondly, it is maintained that such beliefs cannot legitimately rest 
 on faith, but only on argument. Reasons are then given for thinking 
 that the ordinary objections to the belief in human immortality are 
 fallacious. It is suggested that the most reasonable form for the 
 doctrine of immortality to take is one which makes each person to 
 have existed for many years before the existence of his present body' 
 and perhaps for all past time. A chapter on Free- Will endeavours 
 to show that Determinism must be accepted, and that its acceptance 
 would have no bad effects on morality. 
 
 THE CHURCH OF CHRIST. 
 
 Disttation CMxqcb of tbe IRl^bt IRcv. (Bcorgc 1R(&&ing, S).2)*, 
 3Fic0t JBfsbop of SoutbwelU 
 
 Collected and Edited by his Wife, Lady LAURA RIDDING. 
 
 Demy Svo. lo. 6d, net. 
 
 For some time before his death the late Bishop of Southwell had 
 intended to republish in a collected form his Visitation Charges to 
 his Diocese. These Charges are five in number, the last of them, 
 which was to have been delivered at the Synod summoned to meet 
 at Southwell on June 30, 1904, being in the unfinished state in which 
 the Bishop's final illness found it. In preparing them for the press, 
 Lady Laura Ridding has omitted those passages which are of purely 
 local, or of temporary, though public, interest, but in other respects 
 the Charges appear in their original form, and constitute a valuable 
 body of teaching on many of the great Church questions of the day. 
 There is a full synopsis of the subject-matter, which enables the 
 reader to see at a glance the points dealt with under such main 
 headings as The Holy Communion, The Law of the Church, 
 Education, etc. 
 
Mr. Edward Arnold's List of New Books 3 
 
 THROUGH INDIA WITH 
 THE PRINCE. 
 
 By G. F. ABBOTT, 
 
 Knight Commander of the Hellenic Order of the Saviour; 
 Author of 'Songs of Modern Greece,' 'The Tale of a Tour in Macedonia,' etc. 
 
 Demy 8vo, With Illustrations. los. 6d. net. 
 
 Mr. Abbott, whose graphic and amusing description of his last 
 visit to Macedonia will be well remembered, is at present accom- 
 panying Their Royal Highnesses the Prince and Princess of Wales 
 on their Indian tour as special correspondent of the Calcutta 
 Statesman. His tale of this tour possesses a characteristic which 
 will probably distinguish it from other versions — he attaches 
 far more importance to the real condition of the various peoples 
 visited than to the external display made by their rulers. This does 
 not mean that Mr. Abbott fails to describe adequately durbars and 
 similar ceremonies. On the contrary, his descriptions of these 
 incidents are as vivid and convincing as ever. But he does not 
 allow his attention to be distracted by them from the more serious 
 problems of British rule in India, and his record of the Royal 
 progress will possess a more permanent value on that account. 
 
 CONCERNING PAUL AND 
 FIAMMETTA. 
 
 By L. ALLEN HARKER, 
 
 Author of 'The Intervention of the Duke,' 'Wee Folk, Good Folk,' 'A Romance 
 OF the Nursery,' etc. 
 
 Crown Svo. 5s. 
 
 This is a charming book of the same kind as Mr. Kenneth 
 Grabame's 'Golden Age' — a book about children for grown-ups. 
 There are five children, including Fiammetta, full of quaint fancies 
 and amusing pranks — an entirely lovable little party, with delightful 
 dogs and a very characteristic Aunt Eunice. 
 
4 Mr, Edward Arnold's List of New Books 
 
 SURGICAL NURSING 
 
 BnD tbe iprinc(p[cs of Surflcr^ for 'Hnxscs. 
 By RUSSELL HOWARD, M.B., M.S., F.R.C.S., 
 
 Lecturer on Surgical Nursing to the Probationers of the London Hospital ; Surgeon 
 
 TO Out-patients, Royal Waterloo Hospital for Children and Women ; 
 
 Surgical Registrar, London Hospital. 
 
 Crown Svo, With Illustrations, 6s, 
 
 This is an exceedingly lucid and comprehensive handbook on the 
 subject, and contains all the most approved methods very clearly 
 arranged. 
 
 THE LAWS OF HEALTH. 
 
 By D. NABARRO, M.D., B.Sc, D.P.H., 
 
 Assistant Professor of Pathology and Bacteriology, University College, London. 
 
 Crown 8vo. With Illustrations, is. 6d. 
 This volume takes the form of a very simply-written primer of 
 health, in which the direct application of the science of physiology 
 to every-day life is shown, while care is taken to avoid technical 
 terms whenever possible. The author's chief aim is to give such 
 explicit directions as will, if acted upon, help the reader to develop 
 a sound mind in a sound body, and, at the same time, to demonstrate 
 in a simple manner why each rule or warning is given. He also 
 shows in almost every chapter the effect on the tissues and nervous 
 system of a misuse of alcoholic drink and tobacco. 
 
 NEW AND REVISED EDITION. 
 
 FOOD AND THE PRINCIPLES OF 
 DIETETICS. 
 
 By ROBERT HUTCHISON, M.D. Edin., F.R.C.P., 
 
 Assistant Physician to the London Hospital and to the Hospital for Sick Children, 
 Great Ormond Street. 
 
 Demy Svo. With 3 Plates in colour and numerous Illustrations in 
 the text. i6s. net, cloth ; 15s. net, paper. 
 
 This important work, the first edition of which was described by 
 the Guardian as ' one of the most enthralling books ever published 
 on the subject,' has been thoroughly revised by the author in the 
 light of the experience of recent years, and is now absolutely up to 
 date. 
 
Mr. Edward Arnold's List of New Books 5 
 
 THE AENEID OF VIRGIL. 
 
 With a Translation by CHARLES J. BILLSON, M.A., 
 
 Corpus Christi College, Oxford. 
 
 2 vols. Crown A^to. 30s. net. 
 
 These handsome volumes, which are printed in old-style English 
 type on a deckle-edged antique paper, and are bound in drab holiand 
 with paper label on the back, contain on the left-hand page a text 
 based on Conington's, and on the right a line-for-line translation in 
 blank verse. 
 
 THE ROMANCE OF EMPIRE. 
 
 By PHILIP GIBBS, 
 
 Author of ' Facts and Ideas,' ' Knowledge is Power,' etc. 
 
 Crown divo. With Illustrations. 6s. 
 
 In this book Mr. Gibbs tells, in his characteristically interesting 
 style, the story of the expansion of Britain, beginning shortly before 
 the time of Elizabeth, and bringing the account down almost to the 
 present day. Each great division of our Empire beyond the seas is 
 dealt with in turn, and without any sacrifice of historical accuracy 
 or proportion the author gives to his narrative the attractiveness of a 
 well-told romance. 
 
 SHAKESPEARE AND HIS DAY. 
 
 B StuOg ot tbc XToplcal JSIcment in Sbakc0peate auD in tbe 
 JBlisabetban Drama. 
 
 By J. A. DE ROTHSCHILD, 
 
 Trinity College, Cambridge. 
 
 Crown Svo. 5s. net. 
 
 This work was originally written as the Harness Prize Essay of 
 igoi. The author, considering that no field could offer a wealthier 
 fund of Elizabethan remains than the contemporary dramas, has 
 devoted himself almost entirely to the drama as a source of infor- 
 mation, although contributions have occasionally been levied on 
 pamphlets and other writings. The Elizabethan background which 
 is evolved as contemporary allusions are massed together is an 
 achievement equally useful and interesting to the lover of the 
 literature of the period. 
 
6 Mr. Edward Arnold's List of New Books 
 
 VALVES AND VALVE GEAR 
 MECHANISMS. 
 
 By W. E. DALBY, M.A., B.Sc, M.Inst.C.E., M.I.M.E., 
 
 Professor of Engineering, Citv and Guilds of London Central Technical College. 
 
 Royal Svo, With mtmevous Illustrations, 21s. net, cloth ; 
 20s. net, paper. 
 
 Valve gears are considered in this book from two points of view, 
 namely, the analysis of what a given gear can do, and the design of 
 a gear to effect a stated distribution of steam. The gears analyzed 
 are for the most part those belonging to existing and well-known 
 types of engines, and include, amongst others, a link motion of the 
 Great Eastern Railway, the straight link motion of the London and 
 North-Western Railway, the Walschaert gear of the Northern of 
 France Railway, the Joy gear of the Lancashire and Yorkshire 
 Railway, the Sulzer gear, the Meyer gear, etc. A chapter is added 
 on the inertia stresses in the links of a valve gear, and an actual 
 example of the inertia loading of a Joy gear is fully discussed. 
 
 A MANUAL OF PHARMACOLOGY. 
 
 By WALTER E. DIXON, M.A., M.D., B.Sc. Lond,, 
 D.P.H. Camb., 
 
 Assistant to the Downing Professor of Medicine in the University of Cambridge, 
 Examiner in Pharmacology in the Universities of Cambridge and Glasgow. 
 
 Demy ^vo. 15s. net, cloth; 14s. net, paper. 
 
 This text-book, which is prepared especially for the use of students, 
 gives a concise account of the physiological action of Pharmacopceial 
 drugs. The subject is treated from the experimental standpoint, 
 and the drugs are classified into pharmacological groups. The text 
 is fully illustrated by original tracings of actual experiments and by 
 diagrams. 
 
 The author's aim throughout has been to cultivate the reasoning 
 faculties of the student and to subject all statements to experiment, 
 in the hope that pharmacology may thus be learnt like any other 
 science, and consist in something more than the mere committal to 
 memory of many disjointed and often unassociated facts, as it has 
 been too often in the past. 
 
Mr. Edward Arnold's List of New Books 7 
 
 RACES OF DOMESTIC POULTRY. 
 
 By EDWARD BROWN, F.L.S., 
 
 Secretary of the National Poultry Organization Society; 
 
 Author of 'Poultry Kekping : An Industry for Farmers and Cottagers,' 'Industrial 
 
 Poultry Keeping,' ' Pleasurable Poultry Keeping,' etc. 
 
 Crown ^to. With Illustrations, 6s, net. 
 
 This important and comprehensive work, by an admitted master 
 of his subject, will be welcomed by all who are interested in poultry- 
 keeping. Chapters I. and II. deal with the origin, history, and 
 distribution of domestic poultry, and with the evolution and classi- 
 fication of breeds ; the next ten chapters are devoted to the various 
 races of fowls ; Chapters XIII. to XV. treat of ducks, geese, and 
 turkeys. The remaining chapters are on external characters and 
 breeding. There are also Appendices on Nomenclature, Judging, etc. 
 
 A FISHING CATECHISM 
 
 AND 
 
 A SHOOTING CATECHISM. 
 
 By COLONEL R. F. MEYSEY-THOMPSON, 
 
 Author of 'Reminiscences of the Course, the Camp, and the Chase.' 
 
 Two volumes. Foolscap Svo. 3s. 6d. net each. 
 
 Lovers of rod and gun will welcome these valuable handbooks 
 from the pen of an admitted expert. The information given is abso- 
 lutely practical, and is conveyed, for the most part, in the form of 
 Question and Answer. As the result of some fifty years' experience, 
 the author seems to have anticipated every possible emergency, and 
 the arrangement is especially calculated to facilitate easy reference. 
 There are special chapters on fishing and shooting etiquette, and at 
 the end of each book is a chapter dealing with the legal side of the 
 subject. 
 
 'The questions are direct, and the answers equally direct ; it is difificult to 
 think of other questions which might have been put, so wide is the range covered 
 by query and reply ; and, last and best recommendation of all for a book of this 
 kind. Colonel Meysey-Thompson recognises that no question must be ruled out 
 as too easy, or as being one of the things that every duffer knows.' — County 
 Gentleman. 
 
 ' The whole handy, well-printed book is as full of information of the right 
 sort as an egg is of meat. It will delight alike the tyro and the expert, which no 
 book can do that is not thoroughly good.' — Sportsman. 
 
8 Mr, Edward Arnold's List of New Books 
 
 RECENT ADVANCES IN PHYSIOLOGY 
 AND BIO-CHEMISTRY. 
 
 CONTRIBUTORS ; 
 
 BENJAMIN MOORE, M.A., D.Sc, 
 
 Johnston Professor of Bio-Chemistrv in the University of Liverpool. 
 
 LEONARD HILL, M.B., F.R.S., 
 
 Lecturer on Physiology, the London Hospital. 
 
 ;j. J. R. MACLEOD, M.B., 
 
 Professor of Physiology, Western Reserve University, Cleveland, U.S.A. 
 late Demonstrator of Physiology, the London Hospital. 
 
 M. S. PEMBREY, M.A., M.D., 
 
 Lecturer on Physiology, Guy's Hospital. 
 
 A. P. BEDDARD, M.A., M.D., 
 
 Assistant Physician, late Demonstrator of Physiology, Guy's Hospital. 
 
 752^a^^s. Demy Svo, i8s. net, cloth; 17s. net, paper. 
 
 This book, which is edited by Mr. Leonard Hill, consists of Lec- 
 tures on Physiological subjects selected for their direct clinical 
 interest, and designed to meet the requirements of advanced students 
 of Physiology. Professor Moore deals with Vital Energy, Ferments, 
 and Glandular Mechanisms ; Mr. Hill himself with the Atmosphere 
 in its Relation to Life, the Metabolism of Water and Inorganic Salts, 
 and the Metabolism of Fat; Professor Macleod with the Metabolism 
 of the Carbohydrates, and of Uric Acid and the other Purin Bodies, 
 and with Haemolysis ; Dr. Pembrey with the Respiratory Ex- 
 change and Internal Secretion ; and Dr. Beddard with Lymph, 
 Absorption, and the Secretion of Urine. 
 
 NEW EDITION. 
 
 PRACTICAL PHYSIOLOGY. 
 
 By A. P. BEDDARD, M.A., M.D., J. S. EDKINS, M.A., M.B., 
 
 L. HILL, M.B., F.R.S., J. J. R. MACLEOD, M.B., and M. S. 
 
 PEMBREY, M.A., M.D. 
 
 Demy Svo. Copiously illustrated. 12s. 6d, net, cloth ; 
 IIS, 6d. net, paper. 
 
Mr, Edward Arnold's List of New Books 
 
 NEW FICTION. 
 
 Crown Svo. 6s. ea<;k. 
 
 THE LADY OF THE WELL. 
 
 By ELEANOR ALEXANDER, 
 
 AuTHOK OF 'Lady Anne's Walk,' 'The Rambling Rector.' 
 
 HYACINTH. 
 
 By GEORGE A. BIRMINGHAM, 
 
 Author of 'The Seething Pot' 
 
 FOLLY. 
 
 By EDITH RICKERT, 
 
 Author of 'The Reaper.' 
 
 THE HOUSE OF SHADOWS. 
 
 By REGINALD J. FARRER, 
 
 Author of 'The Garden of Asia.' 
 
 THE PROFESSOR^S LEGACY. 
 
 By MRS. ALFRED SIDGWICK, 
 
 Author of 'Cynthia's Way,' 'The Beryl Stones/ etc. 
 ' It is always a pleasure to find Mrs. Sidgwick's name on the title-page of a book, and '' The 
 Professor's Legacy " does not disappoint anticipation. . . . This is an excellent novel.' — Giearefian. 
 
 A TROMBONE AND A STAR. 
 
 By C. T. PODMORE, 
 
 Author of *A Cynic's Conscience.' 
 
 ' Extremely clever and unconventional. ... A book which should not be missed by anyone in 
 search of a real novel.' — East Anglian Daily Tz?nes. 
 
 A FLOOD TIDE. 
 
 By MARY A. DEBENHAM. 
 
 ' As a novel of adventure rather than of character, the book is successful, and a very good 
 specimen of its class.' — The Times. 
 
 THE BROWN HOUSE and CORDELIA. 
 
 By MARGARET BOOTH. 
 
 ' Two cleverly-written and attractive stories. . There is a suggestion of Jane Austen's power 
 
 in several of the situations.' — Dundee Advertiser. 
 
10 Mr, Edward Arnold's List of New Books 
 
 THREE LITTLE COOKS. 
 
 By LUCY CRUMP. 
 
 Square crown Svo. With Illustrations hy Gertrude M. Bradley, 2S. 6d. 
 
 ' No child who owns one of those precious possessions — a miniature cooking 
 stove — should be without this book. It contains many good recipes, adapted to 
 the conditions of a toy stove, and also much good advice, which may be followed 
 with advantage by those boys and girls who play at being cooks. ' — Church Times. 
 
 POLITICAL CARICATURES, 1905. 
 
 By F. CARRUTHERS GOULD. 
 Super royal ^to. 6s, net. 
 
 NEW AND CHEAPER EDITION. 
 
 THE REMINISCENCES OF SIR 
 HENRY HAWKINS 
 
 CBaron :©rampton). 
 Edited by RICHARD HARRIS, K.C., 
 
 Author of 'Illustrations of Advocacy,' ' Auld Acquaintance,' etc. 
 
 Crown Svo. With Portrait, 6s. 
 
 In this edition a few of the more technically legal passages have 
 been omitted, but all the dramatic episodes and characteristic anec- 
 dotes remain untouched. 
 
 RED POTTAGE. 
 
 By MARY CHOLMONDELEY. 
 Crown Svo, 3s. 6d. 
 
Mr. Edward Arnold's List of New Books ii 
 
 ESSAYS AND ADDRESSES ON 
 ECONOMIC QUESTIONS (1865-1893). 
 
 WITH INTRODUCTORY NOTES (1905). 
 
 By the RIGHT HON. VISCOUNT GOSCHEN. 
 
 Demy Svo. 15s. net. 
 
 ' One of those rare and desirable works — an economic treatise based on 
 practical and personal experience, and at the same time interesting and 
 readable. ' — Manchester Guardian. 
 
 ' It is written in graphic and incisive language. Its qualities will, we are 
 convinced, appeal to many readers who would be deterred from studying more 
 formal and elaborate treatises, for they will find here complicated facts set forth 
 with great lucidity and directness. . . They will feel that they are throughout 
 in close contact with the real circumstances of the actual situation.' — Economic 
 Journal. 
 
 FINAL RECOLLECTIONS OF A 
 DIPLOMATIST. 
 
 By the RIGHT HON. SIR HORACE RUMBOLD, Bart., 
 
 G.C.B., G.C.M.G., 
 
 Author of ' Recollections of a Diplomatist,' and ' Further Recollections 
 OF A Diplomatist.' 
 
 Demy Svo. 15s. net, cloth ; 14s. net, paper. 
 
 ' He appears to have met and known every remarkable man and woman of 
 his time who was to be met with in Europe. This last volume is, indeed, like its 
 predecessors, a thoroughly fascinating study.' — Daily Chronicle. 
 
 LORD HOBHOUSE: 
 
 A MEMOIR. 
 By L. T. HOBHOUSE, and J. L. HAMMOND, 
 
 Author of 'Mind in Evolution,' Author of ' C. J. Fox: A Study.' 
 
 Demy Svo. With Portraits, 12s. 6d. net. 
 
 ' No more conscientious public servant than the late Lord Hobhouse ever 
 existed, and it is only right that the community on whose behalf he spent 
 laborious days should be able to appreciate his full worth. That end will be 
 agreeably accomplished by the readers of this compact and eloquent memoir.' — 
 Athenauvi. 
 
12 Mr. Edward Arnold's List of New Books 
 
 THE LIFE OF JOHANNES BRAHMS. 
 
 By FLORENCE MAY. 
 
 Two volumes. Demy Svo. With Illustrations, 21s. net, cloth; 
 20s. net, paper, 
 
 ' There have been many valuable contributions to Brahms literature, but none 
 that has yet appeared is of equal importance with Miss May's volumes.'— TA^ 
 Times. 
 
 ' Quite the most complete and comprehensive life of the master which has so 
 far been produced in this country.' — Westminster Gazette. 
 
 ' Bids fair to remain for many years to come the standard biography in the 
 English language.' — Yorkshire Post. 
 
 A FORGOTTEN JOHN RUSSELL. 
 
 JQcing Xetters to a ^an of :SBnsi\\CB6, 1724=1751. 
 
 Arranged by MARY EYRE MATCHAM. 
 Demy Svo. With Portrait. 12s. 6d. net. 
 
 • A vivacious picture of society, mainly naval, in the reign of the second 
 George. John Russell appears to have been a distant connection of the Bedford 
 family. . Miss Matcham is to be congratulated on her judicious editing of 
 
 this fresh and pleasant volume. Her John Russell has been most tactfully rescued 
 from oblivion.' — AthencBum. 
 
 THEODORE OF STUDIUM : 
 
 HIS LIFE AND TIMES. 
 By ALICE GARDNER, 
 
 Associate and Lecturer of Newnham College, Cambridge ; 
 Author of 'Julian the Philosopher,' 'Studies in John the Scot,' 'Rome the Middle 
 
 OF the World,' etc. 
 
 Demy Svo, With Illustrations. los, 6d. net. 
 
 ' Miss Gardner shows, as she has done before, a fondness for the byways of 
 ecclesiastical history. She has now taken as the subject of a careful study 
 Theodore, who became Abbot of the Monastery of Studium, in Constantinople, 
 towards the end of the eighth century. . . . The author writes without any 
 ecclesiastical bias in one direction or the other, doing full justice to Theodore's 
 nobler qualities, but not concealing or minimizing his faults. The volume is 
 illustrated by some excellent photographs.' — Glasgow Herald. 
 
Mr, Edward Arnold's List of New Books 13 
 
 THE GREAT PLATEAU. 
 
 3Bcir\Q an account of Bjploratlon in Central Zibet, 1903, anC) of tbc 
 <5artoF; JBspeMtfon, 1904=1905. 
 
 By CAPTAIN C. G. RAWLING, 
 
 Somersetshire Light Infantry. 
 
 Demy Svo, With Ilhtstrations and Maps, 15s- net, cloth ; 
 14s. net, paper. 
 
 ' Of exceptional value as a record of travel, and its interest is enhanced by 
 an admirable map and many exceedingly fine illustrations.' — Standard. 
 
 IN THE DESERT. 
 
 By L. MARCH PHILLIPPS, 
 
 Author of 'With Rimington.' 
 
 Demy Svo, With Illustrations. i3S. 6d, net, cloth ; 
 IIS. 6d. net, paper, 
 
 ' A very fine book, of great interest and fascination, that is difficult to lay aside 
 until read at a sitting' — World. 
 
 'There are many that go to the desert, but few are chosen. Mr. March 
 Phillipps is one of the few. He sees, and can tell us what he has seen, and, 
 reading him, we look through his eyes and his sympathies are ours.' — The Times. 
 
 TWO YEARS IN THE ANTARCTIC. 
 
 :S&zim a Bartativc of tbc :fl3titlBb National Antarctic BjpcMtion. 
 By LIEUTENANT ALBERT B. ARMITAGE, R.N.R., 
 
 Second in Command of the 'Discovery,' 1901-1904 ; and of the Jackson-Harmsworth 
 Polar Expedition, 1894-1897, 
 
 Demy Svo. With Illustrations and Map, 15s. net, cloth ; 
 14s. net, paper. 
 
 •A most entertaining work, written in a plain, straightforward style which 
 at once appeals to the reader. It is very nicely illustrated and furnished with an 
 excellent map.' — Field. 
 
14 Mr. Edward Arnold's List of New Books 
 
 FLOOD, FELL, AND FOREST. 
 
 By SIR HENRY POTTINGER, Bart. 
 Two volumes. Demy Svo, With Illustrations, 25s. net. 
 
 ' Sir Henry Pottinger was one of the pioneers amongst Englishmen who have 
 found in Norway a fascinating field of sport, and to these in particular his volumes 
 will appeal. He is at once picturesque and graphic, and to the sportsman in 
 general, and to the frequenter of Scandinavian homes of sport in particular, we 
 heartily commend the book.* — Badminton Magazine. 
 
 THE QUEENS POOR. 
 
 3Lffe as tbeg finO it In Zown anD Countri?. 
 
 By M. LOANE. 
 
 Crown Svo. 6s, 
 
 ' It is a book which is not only a mine of humorous stories, quaint sayings, and 
 all that web of anecdote and quick repartee which sweetens a life at the best 
 limited and austere. It is also a study in which common-sense mingles with 
 sympathy in a record of intimate relationship with the problems of poverty,' — 
 Daily News. 
 
 Sir Arthur Clay, Bart., says of this book :— ' I have had a good deal of ex- 
 perience of " relief " work, and I have never yet come across a book upon the 
 subject of the " poor " which shows such true insight and such a grasp of reality 
 in describing the life, habits, and mental attitude of our poorer fellow-citizens. . . . 
 The whole book is not only admirable from a common-sense point of view, but it is 
 extremely pleasant and interesting to read, and has the great charm of humour.' 
 
 SHORT LIVES OF GREAT MEN. 
 
 By W. F. BURNSIDE and A. S. OWEN, 
 
 Assistant Masters at Cheltenham College. 
 
 Crown Svo. With Illustrations, ^s, 6d. 
 
 Special Cheltonian Edition, including plan of Reredos and an Introduction 
 ■ by the Rev. R. Waterfield, M,A. 4s. 
 
 The Cheltenham College memorial of Old Cheltonians who fell 
 in the South African War takes the form of a reredos in the school 
 chapel, filled with forty-four figures illustrating certain aspects of 
 English history and representative men in different callings of life. 
 It has been felt that an account of these great men would be service- 
 able, not only to those who see these carved figures every day, but 
 to a larger number of readers, who would be glad to have in a com- 
 pendious form biographies of many of the leading men in English 
 history and literature. The list extends from St. Alban to Gordon, 
 and for the sake of convenience chronological order has been 
 adopted. Illustrations are given of eight typical personages. 
 
Mr, Edward Arnold's List of New Books 15 
 
 THE WALLET SERIES OF HANDBOOKS. 
 
 The following five volumes are the new additions to this useful 
 series of handbooks, which range, as will be seen, over a wide field, 
 and are intended to be practical guides to beginners in the subjects 
 with which they deal. 
 
 Foolscap Svo., is. net per volume, paper ; 25. netj cloth. 
 
 THE MANAGEMENT OF BABIES. By Mrs. Leonard 
 Hill. 
 
 ON COLLECTING MINIATURES, ENAMELS, AND 
 
 JEWELLERY. By Robert Elward, Author of ' On Collect- 
 ing Engravings, Pottery, Porcelain, Glass, and Silver.' 
 
 MOTORING FOR MODERATE INCOMES. By 
 
 Henry Revell Reynolds. 
 
 ON TAKING A HOUSE. By W. Beach Thomas. 
 
 COMMON AILMENTS AND ACCIDENTS AND 
 THEIR TREATMENT. By M. H. Naylor, M.B., B.S. 
 
 The following volumes have been already published : 
 
 ON COLLECTING ENGRAVINGS, POTTERY, PORCE- 
 LAIN, GLASS, AND SILVER. By Robert Elward. 
 
 ELECTRIC LIGHTING FOR THE INEXPERIENCED. 
 By Hubert Walter. 
 
 HOCKEY AS A GAME FOR WOMEN. With the New Rules. 
 
 By Edith Thompson. 
 
 WATER-COLOUR PAINTING. By Mary L. Breakell 
 (* Penumbra '). 
 
 DRESS OUTFITS FOR ABROAD. By Ardern Holt. 
 
i6 Mr. Edward Arnold's List of New Books 
 
 NEW EDITION, REVISED AND ENLARGED. 
 
 COMMON-SENSE COOKERY. 
 
 3f or Bngllsb 1bousebol&6, witb ITwcnt^ /Iftenus worfie& out in Detail. 
 By COLONEL A KENNEY-HERBERT, 
 
 Author of ' Fifty Breakfasts,' * Fifty Lunches,' ' Fifty Dinners,' etc. 
 
 Large crown Svo. With Illustrations, 6s. net. 
 
 The author has so largely rewritten this edition that it is prac- 
 tically a new book. Besides being brought up to date with the 
 very latest ideas on the subject, it is much enlarge^, and now 
 contains a number of attractive full-page illustrations. 
 
 CHEAPER EDITION. 
 
 PEN AND PENCIL SKETCHES OF SHIPPING AND 
 CRAFT ALL ROUND THE WORLD. 
 
 By R. T. PRITCHETT. 
 
 Demy Svo, With 50 full-page IlUtstrations, 3s, 6d. 
 
 ILLUSTRATED EDITION. 
 
 HISTORICAL TALES FROM 
 SHAKESPEARE. 
 
 By A. T. QUILLER-COUCH CQ.*), 
 
 Author of 'The Ship of Stars,' etc. 
 
 Crown Svo, With Illustrations from he Boydell Gallery, 6s. 
 
 The value of this much-appreciated work will, it is believed, be 
 enhanced by the addition of sixteen selected illustrations from the 
 well-known Boydell collection. 
 
 LONDON : EDWARD ARNOLD, 41 & 43 MADDOX STREET, W.