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/cletails/cu31924027252232 
 
PE 1345.097" """""*"" '■"'"^ 
 ''**W,.?!?,5',.,te!yon.of thpught-connecfi 
 
 THE 
 
 3 1924 027 252 232 
 
 MEANING AND FUNCTION 
 
 OF 
 
 THOUGHT-CONNECTIVES. 
 
 EDWARD T. OWEN, 
 
 Professor of the French Language and, IMerature 
 in the XIniversity of Wisconsin. 
 
 Main propositions advanced: 
 
 The thought-connective reinstates a preceding thought. It also 
 expresses a relation between a preceding and a following thought. 
 
 The relation is obtained by thinking from the following to the 
 preceding thought. 
 
 The relation and reinstated preceding thought combine as adjunct 
 of following thought, being used as adverb of its verb. . 
 
 In semi-grammatical parlance the thought-connective is a rein- 
 stative adverb. 
 
 Reprinted from the Transactions of the Wisconsin Academy of Sciences, 
 Arts, and Letters, Vol. XII. 
 
 llasued March, 1898, in advance of general publication.] 
 
THE MEANING AND FUNCTION OF THOUGHT- 
 CONNECTIVES. 
 
 The present investigation would naturally cover all species of 
 that grammatical genus known afe conjunctions. For the pres- 
 ent, however, the writer is prepar^To treat those words only 
 which smooth the passage from one statement to its successor. 
 Even this limited effort, further confined to the English lan- 
 guage, seems to require a quasi-apology. 
 
 Thought-connectives are presented in some detail by grammar. 
 They form a considerable part of the arsenal of rhetoric. They 
 are in logic the back-bone of syllogistic reasoning. Psychology 
 notes them as the expression of perhaps the highest states of 
 consciousness. There will surely rise the question : "Why thrash 
 old straw?" 
 
 It may be answered in general that this straw has indeed been 
 forked over and pitched about a great deal, but has been put 
 to very little genuine thrashing. Some grain has been knocked 
 out and a great deal of chaff; but the most and the best of the 
 kernels remain. 
 
 It is claimed in particular that the thought-sciences have, so far 
 as might be, ignored language, neglecting the vast repertory 
 of thought-methods, true and false, of which language is the 
 register. On the other hand the most successful later effort of 
 linguistic science has confined itself to verbal form. The bril- 
 liant achievements in this field have blinded the workers and 
 their watchers to the existence of other fields. It is charged 
 indeed, with more or less plainness, that language-study has 
 well-nigh forgotten that language is after all but the expression 
 of thought. 
 ' To a limited extent it is rational to insist on the analogy be- 
 tween the special language of mathematics and the general 
 
2 Owen — Meaning and Fwnction of Thought-Connectives. 
 
 language of daily life. And what, it may be asked, would bo 
 said of mathematicians, if they neglected quantity and number — 
 if they confined themselves to the history of plus and minus or 
 to the comparative study of these symbols in their use by 
 different peoples? 
 
 The importance of such investigation is, doubtless, in general 
 language greatly increased. But that which general language 
 stands for is also greater than that which mathematical symbols 
 stand for. If then the special language of mathematics must 
 not eclipse the special thought expressed, so also the student 
 of general language should keep in view the general thought 
 of humanity. Thought-study and word-study are two treat- 
 ments of the same problem. It is in these days required that 
 each shall use the aid of the other. 
 
 Judged by this commandment of the new linguistics, philol- 
 ogy has left undone some things which, as linguistic leader, it 
 ought to have done. Grammar has done much that it ought not 
 to have done. Its blind have followed its blind with unfaltering 
 constancy. The pitiful confusion of this pseudo-science offers a 
 convenient introduction to the present subject in the following 
 quoted definitions : " To be is the copula ; " " To seem is a 
 copulative verb; " "And is a copulative conjunction; " " Or is a 
 disjunctive conjunction." Of this absurd crescendo 4;he last 
 term, "the disjunctive conjunction," is specially noteworthy. 
 This no doubt is that long-mourned offspring of classic imagi- 
 nation, the " lucus a non lucendo. " " Or " is ranked as a joiner 
 because it disjoins. Again the meaning of a copulative con- 
 junction depends upon the meaning of a copula. But upon the 
 latter there is, in grammar, no consensus; on the contrary there 
 is conspicuous disagreement. 
 
 A single case of such confusion on the part of those in lin- 
 guistic authority might warrant a break for liberty. Nor is 
 precedent wanting. Chemistry and astronomy broke loose from 
 alchemy and astrology. Even during the present generation 
 almost every science has changed its data, its methods and its 
 terminology. With such tendencies it would be surprising, if 
 the spirit of the time endured much longer the absurdities per- 
 petrated centuries since by men of scholastic bias, often backed 
 
Expression of Thought. 3 
 
 by no proper knowledge of more than a single language. In 
 fact a new linguistic effort has begun. Warned by the failures 
 of grammar, it strives to be intelligible. Taught by the over- 
 sight of philology, it aims at greater completeness. 
 
 PART I. DEVELOPMENT OP THEORY. 
 
 The following discussion is largely put in the first and second 
 person, for the sake of a little gain in objectivity. It takes up 
 in turn the judgment, the nature of its elements and their bear- 
 ing one upon another. It shows how language on the one hand 
 names these parts and on the other indicates their mutual bear- 
 ing. Alongside of this naming and this indicating it admits 
 the occurrence of irrelevant elements. Having thus mapped out 
 the lingustic field, it seeks to find in it the proper place for the 
 connective. 
 
 I. EXPRESSION OF THOUGHT. 
 
 LANGUAGE SYMBOLIZES JUDGMENTS. 
 
 I begin with the judgment, because the aim of language is 
 obviously to express a mental total of no less completeness. I 
 may indeed, by uttering a single word, arouse in your mind 
 the idea which it names. Also you may infer, unless I falsify, 
 that the idea is in my own mind. I may for instance rouse you to 
 think of a chimney and to the opinion that I am thinking of 
 the same. But this involves some effort on your part as well 
 as my own ; and plainly the game is not worth the candle. 
 
 Nor is it enough for you to know that two or more ideas are 
 together in my mind. Every such combination is in a way a 
 mental experiment, a sort of mental tasting of a mental mix- 
 ture. To carry out the suggested comparison, suppose you 
 taste in my presence an unfamiliar fruit. It is of no great in- 
 terest to me to know that you have it in your mouth. What 
 might be useful to me is to know whether you like it ; and this I 
 might learn from the expression of your face. So too, 
 
 when you taste a mental combination, my gain for the most 
 part lies in knowing how the combination affects you. In my 
 
4 Owen — Meaning and Function of Thought- Connectives. 
 
 knowledge of such affects lies also your own greatest gain. 
 Your ability to inform me of your personal attitude toward 
 your own ideas and combinations is the basis of my ability to 
 serve you. I must know not merely that you are thinking of 
 something; I must also know that something as liked or dis- 
 liked, feared, hoped or desired, believed, doubted or disbelieved. 
 
 In the linguistic beginning was, likely enough, the desire. 
 It is at least quite in the line of evolutionary principles to con- 
 ceive the development of language as aided by the survival of 
 the fittest. And he no doubt was linguistically fittest who 
 could tell what he wanted. To this primitive power time has 
 added the ability to express the other personal attitudes ; and 
 now it may be said that the purpose of language is in general 
 to express the speaker's personal attitude toward an idea or 
 combination. 
 
 Something of this personal attitude all agree to find in the 
 judgment. To form a generally acceptable definition of this 
 thought-form is however hardly practicable — still less so to reach 
 an agreement as to how far the sentence is confined to judg- 
 ment-expression. It is however unnecessary to do either. Some 
 forms of thought are recognized by all as judgments; and some 
 sentences are universally felt to be their expression.' 
 
 To such judgments and such sentences I will, so far as may 
 be, confine myself. The principles developed in their discussion 
 will, I think, be found to apply to other thought-forms and 
 other word-combinationsf by whatever name it may seem best 
 to call them. 
 
 ANALYSIS OF THE JUDGMENT. 
 
 The judgment may be analyzed in several ways. So, for that 
 matter, may an insect. I may regard the creature as made up 
 
 • To my own mind a judgment is always the belief in the reality (or 
 unreality) of a relation between two ideas. Even in its imperative form, 
 it vouches for the relation of wisher to what ig wished. In the dubitativ» 
 aspect it assures uncertainty. Even interrogatively formed it asserts the 
 desire for information. 
 
 On the other hand what a sentence aims to express is always a judgment 
 as above conceived, though plainly the sentence often leaves a gap to be 
 filled by inference or external suggestion and often admits irrelevant 
 elements. 
 
Expression of Thought. 5 
 
 of head and residuum — or of tail and what is left. For sev- 
 eral reasons I elect the analysis into head, thorax and abdo- 
 men. I prefer something analogous with judgments, reaching 
 results as follows. Let "6 exceeds 4" serve as illustration. 
 The mental process registered is this : I have passed in mind 
 from 6 to 4. In so doing I have become aware of something 
 foreign to 6 and foreign to 4, namely superiority. The judg- 
 ment expressed consists then of three terms : a first term (6), a 
 last term (4) and the relation between them, which may also be 
 known as mid-term.' 
 
 It should be noted that the relation may take on either of two 
 aspects according to the direction of the thought- transit. Thus, 
 if I think from 6 to 4, I obtain superiority. But, if I think 
 from 4 to 6, I obtain inferiority. Viewed from the stand-point 
 of superiority, inferiority is in this case a reverse relation. 
 Conversely", superiority may be known as proverse. The names 
 progressive and regressive are also in use. 
 
 When now the judgment consists of merely a relation and the 
 terms between which the relation obtains, I call the judgment 
 elementary, minimal or essential. 
 
 The judgment may, of course, be amplified. All admit that 
 amplification must follow the method of the original structure. 
 That is, if I add to my essential judgment, the new must be in 
 relation with the old or with one of its parts. Were the new 
 not so in relation, it would be out of relation or irrelevant. That 
 is, the new would, so to speak, have nothing to do with the old. 
 
 In practice the relation of old and new is often unexpressed. 
 When I amplify "man" into "white man," you do not require 
 me to say that the relation between " man " and " whiteness " 
 is that of object to its quality. 
 
 ' However any of these elements be emphasized or overlooked, it is hard 
 to believe that any one can ever be absent from a thought deemed worthy of 
 expression. I have elected the names applied to them, because "first term " 
 and " last term " are broader and also less confuaed with other names than 
 subject and predicate (also object) and because the name of " mid-term " 
 is a constant reminder to keep track of the often neglected relation. 
 
 It is easy to show further that this relation is in my example conceived 
 as real (not this time unreal) and that this reality I believe. But these 
 matters being unnecessary to the present investigation, I neglect them. 
 
6 Owen — Meaning and Functi(m of Thought- Connectives. 
 
 The new ideas are related sometimes with one old one and 
 sometimes with another. It is convenient to use with all alike 
 the name of adjunct, meaning thereby an added idea plus its re- 
 lation expressed or understood. 
 
 In illustration let an essential judgment be expressed by the 
 sentence: " the priest employed a lawyer." Suppose now that 
 I wish to amplify this judgment by an idea of priority. This 
 idea must appear, as last term in a new thought, whose first term 
 may be either of the original three terms. It is accordingly 
 possible to construct three amplified judgments, respectively ex- 
 pressible by three amplified sentences. (1) " The prior or former 
 priest employed a lawyer. " (2) " The priest employed a former 
 lawyer. " (3) " The pijiest formerly employed a lawyer. ' In the 
 last case that which is conceived as prior is the " employment, " 
 the relation of employer to employee. In the other cases it is 
 respectively the " priest " and the " lawyer. " 
 
 Illustrating again, I might use the phrase " a feeble old 
 priest. " My first term in this case has two adjuncts of equally 
 direct bearing upon itself. A suflBcient number of examples 
 would justify the proposition that the same term may have co- 
 ordinate adjuncts to the number of n. 
 
 Again I might speak of a " remarkably old priest. " In this 
 case primarily the " priest " is conceived as " old " and second- 
 arily this " age " is thought of as " remarkable. " Generalizing 
 again, I make the further proposition that each adjunct may 
 have a sub-adjunct and this last another and so on to the »th 
 degree of subordination. 
 
 Terms, then, and adjuncts are the elements for which the sen- 
 tence properly stands. But other elements sometimes intrude. 
 For instance, "All men (shut the door, please) are animals." 
 Of such elements it is enough to say that, although they are in 
 the thought, they are not of it. 
 
 THE SYMBOLIZING PROCESS. 
 
 It is the peculiar office of language to establish in the hearer's 
 (or reader's) mind the thought of the speaker. This process is 
 commonly called conveyance. The figure is somewhat unfortun- 
 ate, as it is obviously impossible to convey any act or product 
 
Expresswii of Thought. 7 
 
 of the mind. Strictly speaking, language is a process by which 
 the hearer is induced to erect in his own mind a thought-struc- 
 ture like that of the speaker. This result is effected by stimuli 
 known as words.' Its first stage is accomplished when the 
 hearer's mind contains the elements of the speaker's thought. 
 
 Let it be supposed that this stage has been reached. Much 
 still remains to be done. For instance, both speaker and hearer 
 may be thinking of a missionary, a cannibal and the relation of 
 eater to food; but one may have it that the cannibal eats the 
 missionary, while the other supposes that the missionary de- 
 vours the cannibal. In the expression of more complex thought 
 such possibilities become probabilities of the greatest import- 
 ance. It appears then plainly that speaker and hearer must not 
 only have in mind the same ideas ; they must also build them 
 together by the same plan. 
 
 Of the several ways of meeting this need I will mention only 
 the strictly linguistic. These are two in number, often con- 
 fused, but radically different. Using an objective illustration, 
 suppose I give you the pieces of a dissected map — that of the 
 United States — and invite you to put them rightly together. You 
 have the elements of a structure, but not its plan. If now I 
 give you the latitude and longitude of each state-capital, they 
 constitute a sort of plan. The instructions furnished for each 
 state are moreover independent of all other states. They may 
 in that sense be called absolute. 
 
 Precisely analogous instructions are given by language, 
 though confined for the most part to word-endings. In the 
 sentence " They hear us " the ending of the word "they" is an 
 order to use the word as first term. Instruction of this kind I 
 call functional. 
 
 Suppose now that, in giving you the pieces of the map, I 
 name no latitude or longitude ; you are still amply provided with 
 instructions, but of another sort. Examine for instance the 
 Indiana section. On its southern edge you note a particular 
 curve. This detail you may take as an order to put Indiana 
 next to some other state with a similar curve, namely Kentucky. 
 
 ' Other stimuli, gesture etc., are in this investigation neglected. 
 
8 Owen — Meaning and Function of Thought-Connectives. 
 
 But the order is this time relative, not absolute. It is not 
 hinted where on the globe each state or either belongs. It 
 is however distinctly declared that they belong together. 
 
 Instruction of this relative type is' also offered by words. 
 Given the Latin " bonarum, " its termination constitutes an order 
 to put the idea of goodness with some idea conceived as plural, 
 feminine and genitive. This sort of instruction I call associa- 
 tional. An idea with the function-label may be conceived as 
 saying to any other idea: " No matter where yow go, J go here. " 
 An idea with the association-label says rather to owe other idea: 
 " No matter where you go, I go with you. " 
 
 Thes.e distinctions, guiding as they do the course of my in- 
 vestigation, seem to me worthy of emphasis, and again because 
 they are generally neglected. For instance, in the case of the 
 verb, forms which stand for meanings (e. g. tense-signs), others 
 which name function (e. g. the ending of the adjunctively em- 
 ployed participle) and still others which show association (or 
 agreement, e. g. personal endings) are made to ride together in 
 one categorical omnibus called a conjugation. As well bring to 
 a mason, in one hod, bricks and plans, mortar and specifications 
 and claim to ease his task by collective presentation. 
 
 MtTLTIPLE SYMBOLIZATIOM. 
 
 In the study of language it is helpful to bear in mind that 
 the same word is often charged with several duties. The power 
 to perform them is a mere survival or revival from a time in 
 which, it is believed, the judgment Itself was expressed by a 
 single word. Of this method even historic examples are not 
 wanting. Thus the Latin " pluit " has no less expressive power 
 than the English " it is raining. " Again, in the sentence " the 
 birds sing " it is plain that, while one term only is expressed 
 by " birds, " the other term and the relation are necessarily ex- 
 pressed by " sing. " That is, the word " sing " in some way 
 stands for two ideas, which on other occasions maybe expressed 
 by a word for each (e. g. " the birds are singing "). The action 
 of the word in such cases may be known as multiple symboliza- 
 tion. This name may be applied also to the action of the word 
 
Expression of Thought. 9 
 
 when, instead of standing for more than one idea, it stands 
 for an idea and also for that idea's function or association.' 
 
 REINSTATIVE STMBOLIZATION. 
 
 In developing my theory it is further convenient to make use 
 of what may be called the doctrine of reinstative syrnbolieation. 
 Of this let the sentence " I met Brown to-day " serve as intro- 
 duction. In this sentence the word " Brown " excites in the 
 hearer's mind an idea of which, it will be assumed, he would not 
 otherwise be conscious. The symbolizing act may in this case 
 be called introductory or initial. 
 
 The idea excited by the word " Brown " must plainly remain 
 in consciousness till all the elements of thought to be expressed 
 
 ' When a word expresses more than one idea, the naming of one 
 of the ideas may or may not be effected by the termination (prefix, in- 
 fix or other variation). When, in addition to an idea, its association or 
 function is to be indicated, the aid of a termination is usually invoked. 
 In the highly inflected languages almost every word is in some of these ways 
 a multiple symbol. 
 
 Historically the instructional elements (functional and associational) are 
 always original idea-namers. It is accordingly not surprising that a 
 given termination should exhibit on different occasions the different pow- 
 ers which it has successively acquired. Thus the Latin " cognovit" on one 
 occasion contains in itself its subject (he); on another the personal ending 
 merely shows association with a subject expressed by another word; at the 
 same time this ending, being used with no other kind of thought-element, 
 has functional value, being evidence that the idea expressed is the mid- 
 term (principal verb). 
 
 The same word-ending may then express a thought-element, its associa- 
 tion or its function — sometimes one and sometimes another — sometimes 
 one only, sometimes two, sometimes all three. The distinction made is 
 therefore useless for the purpose of ending-classification. As well endeavor 
 to classify the cattle in the field by their momentary place or doings. 
 
 But, apart from the possible interest to the zoologist, the where-abouts 
 and the what-abouts of each animal may have an immediate interest to 
 its owner. He can hardly afford to ignore differences between here and 
 there or between doing this and doing that. So too with these termina- 
 tions; though I cannot classify them by the differences in their values, 
 I must not forget that these values are different. I must also recognize 
 the particular value of each on each particular occasion. Fortunately 
 such recognition is rarely difficult. 
 
10 Owen — Meaning and Function of Tlwught- Connectives. 
 
 ^re in the hearer's mind. Otherwise it would be impossible for 
 the hearer to make a thought-structure identical with that of 
 the speaker. He would simply be unable to build completely 
 without complete materials. 
 
 On the other hand, at the end of the sentence, when the 
 thought-structure has been completed, it and its elements are al- 
 lowed to disappear from immediate attention. This disappear- 
 ance is indeed necessary to the reception of another thought. 
 That such is the fact may be shown by any somewhat complex 
 example; e. g. 348-5-6 = 58. The elements of this judgment 
 or equivalents thereof are held by the hearer till the total judg- 
 ment is received. Suppose, now, that this first statement is 
 succeeded by the following: 32X17 = 544. It is plainly im- 
 practicable to form the second thought-structure, so long as any 
 elements of the former are still in immediate attention. 
 
 Applying these principles to a verbal statement, let the for- 
 mer example serve. " I met Brown to-day. " When the hearer 
 has formed the thought-structure expressed by these words, all 
 the ideas therein contained are allowed to disappear from imme- 
 diate attention. But they do not disappear bej'-ond recall. This 
 may be shown by an extension of the above example. " I met 
 Brown to-day. He was in excellent health." In the second 
 sentence it is very evident that the word " he " recalls to the 
 hearer's mind the idea which had been initially presented by the 
 word " Brown. " This action of the word '" he " is what is meant 
 by reinstative symbolization. 
 
 Eecalling now what was said of the symbolizing process, and 
 bearing in mind not only words but also word-endings (etc.), I 
 offer, as at least intelligible, the following list of 
 
 POSSIBLE WORD-VALUES. 
 
 I. Structural. — Words stand for the ideas or materials of which 
 a thought is constructed. These are 
 
 a. First, last, and mid-term. 
 
 b. Adjuncts (direct or indirect) of any of these. 
 
 II. Non- Structural. — Word-values in this case are 
 
 a. Instructional (helps to put materials rightly together) 
 further distinguishable as 
 
Expression of Thought Connection. 11 
 
 (1) Functional; these show as what thought-element an idea 
 is to be used; they declare absolute position; 
 
 (2) Associational ; these show that a particular idea can be 
 used only with another particular idea; they declare relative- 
 position. 
 
 h. Independent. 
 
 Supposing this list to be complete, the question rises: what is 
 the value of the thought-connective ? Before seeking an answer, 
 let the connective first be isolated from its sister species in the 
 grammatical classification. All these species are ranged under 
 the genus conjunction. One species is conceived to join separ- 
 ate sentences; thus: "Roses are red but lilies are white." 
 Another is said to join single words; thus: "He and I were 
 there. " Another still is thought to unite larger portions of the 
 sentence; thus: "He stood up, that all might see him." The 
 present investigation will be extended to conjunctions of the- 
 first kind only ; that is, to words which are said to join sentences. 
 From the view-point of syntax or sentence-study these words 
 are sentence-conjunctions. From the view-point of thought- 
 study they would properly be known as judgment-conjunctions. 
 In semi-popular parlance they are called thought-connectives. 
 This phrase is not absolu1;ely definite; but, being enough so for 
 practical purposes, it will henceforth be employed. To reach a 
 perfectly distinct, though somewhat arbitrary line of demarca- 
 tion, only those sentences between which a period is allowed to- 
 stand will be ranked as separate. 
 
 II. EXPRESSION OF THOUGHT CONNECTION. 
 
 The connecting or joining of thoughts is obviously a mere 
 figure of speech and cannot be expected to present with exact- 
 ness any process of the mind. Yet a little examination of the 
 figuratively taken physical process is helpful in studying the 
 mental process which it pictures. 
 
 MEANING OF CONNBCTON. 
 
 Viewed in its simpler aspects, what is known as connection or 
 junction supposes unbroken continuity in the unit which it pro- 
 
12 Owen — Meaning and Function of Thought- Connectives. 
 
 duces. Mere contact is not suflSoient. Two balls are not con- 
 ceived as joined by the mere contact of their surfaces. Indeed in 
 one view contact is needless. The two balls of a dumb-bell are 
 conceived as joined by the handle, though they themselves are 
 each at some distance from the other. Yet the molecular con- 
 tinuity is maintained throughout the total mass. 
 
 This illustration may be viewed in another way, which will, 
 I believe, be found more helpful. The handle and one ball may 
 conveniently be taken as forming one total. The handle and the 
 ■other ball form another total. The handle is then a common 
 factor of both totals. As it is obviously possible to take the 
 same view of all joined bodies, the condition of junction may 
 be defined as the possession of a common factor. 
 
 The common factor may vary in size from that of either body 
 (or both) down to nothing. The cases which develop from such 
 variation have been carefully distinguished and are well known as 
 
 Inclusion: the common factor is the greatest possible part 
 (i. e. all) of one body; or, the common factor is the greatest 
 possible part (i. e. all) of the other body; 
 
 Coincidence: the common factor is the greatest possible part 
 (i. e. all) of both bodies; 
 
 Exclusion : the common factor is the smallest possible part of 
 one body or of the other body or of both.' 
 
 ' These cases may be conveniently developed as follows: Let two areas 
 overlap each other thus: 
 
 C being the common factor, A -f C one area and C -f- B the other. 
 
 (1) Let C be increased to the value of A -f- C; or, what is more conven- 
 ient, let A -f C be diminished to C. The right hand area now includes 
 all of the left. 
 
 (2) Conversely, let C -f- B be diminished to C. The left hand area now 
 includes all of the right. 
 
 (3) Let C be increased to the value of A -|- and of C -|- B. The 
 possibility of this presupposes the equality of the two areas. For con- 
 
Expression, of Thought Connection. 13 
 
 VARIETIES OF THOUGHT CONNECTION. 
 
 Passing now from the objective to the ideal, I shall endeavor 
 to show that, in all cases in which thoughts are conceived to be 
 joined, one of them is either wholly or partially included in the 
 other. I do not therefore need to consider the case in which 
 two thoughts are mutually exclusive. The same is true with 
 the case of thoughts conceived as coincident. Moreover this in- 
 teresting case, however occurring, would naturally be expressed 
 by a plural and not by the help of anything called a conjunction. 
 
 There remain to be considered then three cases only.' A. 
 Two thoughts have a part only of each in common. B. The 
 first contains the whole of the second. C. The second contains 
 the whole of the first. 
 
 A. Two thoughts contain each a part of the other. 
 
 This case may be illustrated most clearly by a form of sentence 
 which, though little used, has appeared in widely different 
 languages. Thus: (from old German) "Up from his seat arose 
 Herr Hagen spoke as follows. " There are here two complete 
 sentences, presenting two complete judgments. One is: "Hagen 
 rose from his seat. " The other is : " Hagen spoke as follows. '* 
 The two have a common factor, namely the first term "Hagen." 
 This term is only once named and only once thought.' 
 
 venience, instead of increasing, let A + C and C + B be each diminished 
 to C. C then stands for two areas; and these are coterminal or coincident. 
 
 (4) Let now C be diminished as much as possible, namely to zero. Ac- 
 cording to the view-point one may say with equal truth that C, as part of 
 A -j- C, is the smallest possible part of C + B; the left area is excluded 
 from the right. Again C as part of C -|- B is the smallest possible part 
 of A + C; the right area is excluded from the left. And still again C as 
 part of both A + C and C -|- B is the smallest possible part of both; the 
 two areas are mutually exclusive. 
 
 ' I necessarily omit thoughts occurring in different minds; for obyiously 
 the thoughts of one mind have no existence for another until they become 
 also the thoughts of that other. Simultaneous thoughts in the same mind 
 I also omit, because in linguistic presentation they are perforce expressed 
 in succession or as a plural. 
 
 ^ In such cases the common factor is usually symbolized a second time 
 by " who," the second symbol allowing a second inflection expressing as- 
 
14 Owen — Meaning and Function of Thought- Oonnectives. 
 
 The same principle is very extensively employed with other 
 terms and with adjuncts; but I believe grammarians have not 
 ■called the process conjunctive except in some uses of the adverb; 
 e. g. "He departed as I arrived." In this example the word 
 ""as " is plainly an indefinite time-namer. That the idea of time 
 belongs with " arrived " has never been questioned. That this 
 idea also belongs with "departed" cannot be questioned; for 
 the whole purpose of time -naming is to show when "he de- 
 parted. " This time-idea is then first and foremost an element of 
 the first thought. As such it is indefinite. This indefiniteness 
 is removed by its association with "my arrival." The very aim 
 and purpose then of the time-idea fails, unless it be taken as 
 a factor of both thoughts. 
 
 From the above and other examples it would be easy to show 
 that when two thoughts have a part only of each in common, 
 they are not expressed by what are usually ranked as separate 
 ■sentences. This case does not then lie within the limited 
 field of this investigation. 
 
 B. The first thought contains the whole of the second. 
 
 The simplest cases show the second thought as last term of the 
 first thought, e. g. " I believe his being honest. " " I believe him 
 to be honest. " I believe that he is honest. " These three forms, 
 so far as function of the second thought in the first is concerned, 
 are identical. The notion of conjunction does not, however, oc- 
 cur to the grammarian till he reaches the form: " I believe that he 
 is honest. " And even with this form the notion seems to be losing 
 ground. It is now quite generally believed that the word " that " 
 is a substantive of indefinite meaning, that this meaning is 
 •symbolized a second time and definitely in what follows. 
 
 From the above and other examples it would be easy to show 
 that when the first of two thoughts contains the whole of the 
 second, the two are ordinarily not expressed by separate sen- 
 tences. Such expression does, however, sometimes occur ; e. g. 
 "What I believe is this. A is honest." Clearly enough the 
 
 sociation and function of the common factor in the second thought. But 
 the two symbols stand for a single idea and this idea makes but one ap- 
 pearance, a continued appearance, in the mind. 
 
Expression of Thought Connection. 15 
 
 word " this " presents indefinitely that which is believed ; and 
 that which is believed is again symbolized by the second sen- 
 tence. The indefinite prefigures the definite and may be said 
 to be the definite in a germinal stage of development. But as 
 such cases are comparatively rare and obviously different from 
 those to follow, they may be excluded from this investigation. 
 
 C. The second thought contains the whole of the first. 
 
 The two may be expressed by a single sentence, e. g. " That A is 
 honest I do not doubt. " But more commonly two independent 
 sentences are employed, e. g. "A is honest. That I do not doubt. " 
 Cases in which the first sentence is so obviously repeated as a 
 main term of the second may be neglected on account of their 
 simplicity. 
 
 In the cases to be considered neither sentence is usually con- 
 ceived to be repeated in the other, but a word of the second 
 is believed to join the two. This word is classed as a con- 
 junction and is known by the special name of a thought-con- 
 nective. 
 
 It is my object in this investigation to demonstrate that, in 
 these cases also, what the thought-connective names is part and 
 parcel of the second thought. In so doing, I will first endeavor 
 to show that such a theory must be true by establishing the 
 untruth of all other possible theories. 
 
 In choosing examples it would be strictly proper to use only 
 sentences of some length, because the accepted sign of separate- 
 ness, the period, is usual only between such sentences. But, 
 as a matter of convenience, short examples are chosen, it being 
 always easy to conceive them as sufficiently amplified. 
 
 WHAT VALUES THE THOUGHT-CONNECTIVE CANNOT HAVE. 
 
 Using my list of possible word- values (see p. 10) and follow- 
 ing the order of convenience, it is plain that 
 
 First, the thought-connective might be functional. That is, 
 it might show as what thought-element some idea is to be used. 
 But this it does not; e. g. "And George the dragon slew". 
 The poetic arrangement is chosen for the purpose of leaving to 
 order no power to settle the choice of subject. It is accordingly 
 
16 Owen — Meaning and Fitnction of Thought- Connectives. 
 
 uncertain whether George slew the dragon or the dragon slew 
 George. That is, the function of either noun, its choice of 
 position as first or last term, is uncertain. This uncertainty 
 is in no degree relieved by the presence of the word " and. " 
 Also " and " has no influence in fixing " slew " as mid-term. 
 That is, the conjunction is not to be classified as functional.' 
 Second, the thought-connective might be associational ; e. g. 
 
 " And large exceedingly fierce men dragons slew. " The func- 
 tion of the idea named by " exceedingly" is clear. It must be 
 taken as an adjunct of one of the first term's adjuncts, either 
 with " large " or with " fierce. " But its choice between these 
 two, that is to say its association, is uncertain. This uncer- 
 tainty is in no degree relieved by the presence of the word 
 
 " and. " That is, this conjunction is not to be classified as as- 
 sociational.'' 
 
 Third, the thought-connective might be independent of the 
 rest of the sentence — a foreign substance in the verbal organ- 
 ism ; e. g. " And George the dragon slew. " The word "'and, " 
 if irrelevant, is required to join what follows to what has pre- 
 ceded, while itself completely disconnected from what follows. 
 It seems unnecessary to argue that this is impossible. The 
 moment that union ceases between " and " and what follows, the 
 union between what follows and what precedes must also cease, 
 unless maintained by some other means than any afforded by 
 "and." That is, the conjunction cannot be classified as inde- 
 pendent.' 
 
 ' The conjunction obviously gives no instruction as to its own function. 
 
 ^ The conjunction obviously gives no instruction as to its own associa- 
 tion. 
 
 'It might be claimed that, even though the word " and " be not itself a 
 joiner, it shows that thoughts are to be joined by the hearer or have been 
 joined by the speaker; that is, " and " might be either an order to join or 
 a statement of junction. While this view contains much truth and the 
 appearance of more, it shrinks in value when we find no indication whether 
 a statement or an order be designed. The absence of such indication sug- 
 gests at least the sub-importance to the speaker of what he omits to name. 
 In illustration let the thought expressed by our example be fully devel- 
 oped. Using first the the mandatory value of " and," the sentence " And 
 George the dragon slew " does not mean " Join George the dragon slew to 
 
Expression of Thought Connection. 17 
 
 The above examination of the thought-connective in so small 
 a number of examples cannot be regarded as proving that it 
 never has the values successively tested. It shows, however, 
 that the possession of such values, if it ever occurs, cannot be 
 regarded as essential — cannot, therefore, be properly made the 
 basis of classification. It seems, accordingly, evident that the 
 thought-connective must be classed as the only other element 
 possible, namely a structural element. The doctrine of the 
 growth of instructional sentence-elements from so-called empty 
 words favors the expectation that the thought-connective, at 
 least in primary usage, will be found to be fully structural. 
 
 Fourth, then, and last, the thought-connective may be struct- 
 ural. If such, it might be a term of the essential judgment. 
 But this, to the most superficial observation, it is not. Recall- 
 ing the example: "And George the dragon slew," it is plain 
 that " and " is not the first term, otherwise known as subject. 
 It is not the last term, otherwise known as sometimes object and 
 sometimes predicate. It is not the mid-term or relation, other- 
 wise known as verb or copula. 
 
 WHAT VALUE THE THOUGHT-CONNECTIVE MUST HAVE. 
 
 Turning from the discussion of what the thought-connective 
 is not, it remains to show if possible what the thought-connec- 
 tive is. There remains one class only of ideas from which, by 
 the preceding argument, it has not been excluded, namely the 
 class of adjuncts. It is therefore provisorily assumed that the 
 thought-connective enters thought-structure as an adjunct. An 
 
 a preceding statement." It means rather " Gfeorge slew the dragon — a 
 statement to be joined by you to its predecessor." That is, the order to 
 join the statement is of less importance than the statement itself. Or, 
 using the explanatory value of " and," the meaning is not " I have joined 
 George the dragon slew to a preceding statement," but rather " George slew 
 the dragon — a statement joined by me to its predecessor." That is, the 
 explanation is much less important than the statement. Indeed, for my 
 own part, I can not, in the given example, say either that I am telling 
 you to do the joining or stating that I have done it myself. All that I feel 
 sure of is that I do conceive a junction. Its particular form and method 
 
 will be developed in another place. 
 
 2 
 
18 Owevr— Meaning and Function of Thought- Connectives. 
 
 attempt will now be made by the study of a single thought-con- 
 nective to develop a more detailed theory. The whole body of 
 thought-connectives will next be classified. The theory will 
 then be tested by successive applications to all classes. 
 
 For preliminary examination of thought-connectives let the 
 word " therefore " serve as a type ; e. g. " He invited me. 
 Therefore I came. " The meaning of the word " therefore " is es- 
 sentially "on account of that " or, more briefly, " from that. " Not 
 only is this evident to ordinary observation, but etymology, his- 
 torical and comparative, declares it. It is accordingly clear at 
 the outset that the present case is one of multiple symboliza- 
 tion. That is, the single word "therefore" stands for ideas 
 which are often represented separately. 
 
 The exact value of each of these ideas merits careful observa- 
 tion. And plainly the usual demonstrative value of " that " is 
 not sufficient. Its present value is reinstative. That is, it brings 
 back to attention something which has preceded. ' In the present 
 case, moreover, it brings back not a mere thought-element but 
 a whole thought. The proper test of this opinion is to put the 
 whole preceding thought in place of the word " that. " Accord- 
 ingly "He invited me. On account of his inviting me I came. " 
 Plainly this substitution neither adds to nor subtracts from the 
 speaker's meaning. " His inviting me " may be accepted, there- 
 fore, as the true value of " that. " ' This element of the connec- 
 tive is, then, reinstative or anaphoric, not initial, not deiktic. 
 
 • The propriety of thus defining " that " may be further developed as 
 follows. It is evident that the same mind-stuff can be regarded as one or 
 as composed of parts, according to the speaker's choice. To illustrate ob- 
 jectively, a horse, buggy and connecting harness may be regarded as either 
 one or three at convenience. Indeed the language of the people supplies 
 in the word "rig" a name for the unanalyzed combination. But as 
 rule such names do not exist for thoughts. Names for the most part are 
 symbols merely of thought-elements or ideas. Even when conceived as a 
 unit, the thought as a rule can be distinctly presented only by combining 
 the names of its parts, one part appearing as nucleus and the others being 
 clustered about it. The question rises then: which part shall be put as 
 nucleus? In answering, we must avoid a common but erroneous valua- 
 tion of thought-elements. 
 
 The phraseology of grammar presents relations as of subordinate Value. 
 
Expression of Thought Gonnecikm. 19 
 
 The remaining element, the phrase " on account of " or its 
 equivalent " from, " can best be appreciated after some exam- 
 ination of the preposition. This thought-element, like the verb, 
 is a relation-namer. But it differs from the verb in being em- 
 ployed in the special case in which a relation and a last term 
 are combined as an adjunct of what stands thereto as a first 
 term ;■ e. g. " The book on the table. " In this phrase " book " and 
 " table " are plainly a first and a last term and the relation be- 
 tween them is expressed by " on. " This relation and the last 
 term are taken together as an attribute or accident of the first. 
 That is, the book is characterized by " on-the-tableness, " In 
 
 Subject and predicate (firat and last terms) are conceived as of primary 
 importance; other elements are noted as " mere relations." In opposition 
 to such valuation it may be observed that the habitual effort of the mind 
 is to find relations. In an advanced stage of mental development we may 
 indeed start with a relation and institute a search for terms to fit it. This 
 process is, however, unusual. As a rule we are not finders of terms for 
 relations. We are finders of relations for terms. Man might indeed be 
 distinguished as the relation-finding animal. So far then as results are 
 more important than data, so far in the thinker's mind the relation out- 
 values the other terms. 
 
 Suppose now that a thought conceiYed as a unit is to be presented by 
 the names for its parts. It is plainly possible to use as nucleus either first 
 or last term, distinguishing it by its relation to the other term. But such 
 a method violates the thinker's estimate of values. The proper method is 
 rather to use as nucleus the relation and to distinguish this relation by 
 the terms between which it holas. Having said for instance that 6 ex- 
 ceeds 4, if I wish to distinctly present this thought as a unit, I use the 
 phrase "the superiority of 6 to 4," not "the 4-exceeding 6," not "the 
 by-6-exceeded 4." 
 
 So also in the present case, having stated analytically that " He invited 
 me;" having obviously expressed by " invited " a relation (that of inviter to 
 invited) between him and me, which can be learned from no other source; if 
 now I wish to distinctly present my thought again as a unit, but by means 
 of its part-namers, my descriptive phrase will be of some such form as 
 " his inviting me," " his invitation of me," or, " my invitation by him." 
 
 ' I believe that sufficient testing will show an opportunity to apply here 
 Morgan's excellent distinction between focal and marginal consciousness; 
 that, when several relations enter into one thought-structure, that which is 
 dominant (or centrally, focally conscious) will be found to be always ex- 
 pressed by the principal verb; that the preposition will be found to name 
 in all cases a relation of a subordinate, eccentric or marginal character. 
 
20 Owen — Meaning and Function of Thought- Connectives. 
 
 other words, the prepositional phrase is an adjunct. This ad- 
 junct is in the present case used as an adjective. Such an ad- 
 junct in other cases may be used like an adverb; e. g. " He 
 sank with great rapidity. " Accompaniment (figuratively put 
 for characterization) by rapidity is treated as an adverbial ad- 
 junct of '■ sank." 
 
 It should further be noted that, in the above examples, the 
 relation named by the preposition is obtained by thinking from 
 the first-mentioned term to the last. Thus, in the phrase " the 
 book on the table " the speaker plainly obtained the relation by 
 what may be called a thought-transit from the " book " to the 
 " table. " For, if he had thought from the table bo the book, 
 he would have developed the relation named by " under. " For the 
 relation expressed by " on " I have chosen; as above noted, the 
 name of proverse or forward, giving to that expressed by " un- 
 der " the name of reverse or backward. And it should also be 
 observed that between any two terms either relation may be de- 
 veloped; for it is obviously possible to make the thought- tran- 
 sit from either one to the other. 
 
 It should further be noted that the order of ideas in thinking 
 is ilot necessarily that of the words employed to express them. 
 For instance, using an order allowed in poetry, " The earth 
 above the sky is blue. " Now there is no earth above the sky. 
 Accordingly it must bo that, in spite of the order of words, 
 the order of ideas is that expressible by " the sky above the 
 earth." So far as observed, such violations of order in the 
 prepositional phrase are quite rare. It may then, within the 
 bounds of safety, be said that the preposition usually names 
 a relation obtained by a thought-transit from its first-named 
 term to its last-named term. 
 
 With thought-connectives the opposite is the case. Given 
 again the example : ' ' He invited me. Therefore I came. " " His 
 invitation " is the first-mentioned term. " My coming " is the 
 last-mentioned term. " Therefore " contains a relation between 
 the two. This relation is not developed by a thought-transit 
 from the invitation to the coming. Were such the case, the 
 developed relation would be named by " toward " or an expres- 
 sion of causation, e. g. " His invitation was toward my com- 
 
Expression of Thought Connection. 21 
 
 ing, favored my coming, caused my coming. " The actual rela- 
 tion, expressed by " from, " is plainly developed by passing in 
 mind from " my coming " back to " his invitation. " This rela- 
 tion would be asserted by saying: "My coming was from the 
 invitation, on account of it, caused by it. " In short, " I came 
 from or for that. " 
 
 It appears then that the word " therefore " has two simultan- 
 eous meanings. It names a reverse relation between two 
 thoughts. Of these two thoughts it reinstates the former. 
 
 In remains now to be seen in what way these two meanings 
 are structurally employed. To determine this it should be noted 
 that the structure of the second thought is the outcome of a 
 compromise between two conflicting desires. The speaker 
 wishes, on one hand, to assert " his coming. " On the other 
 hand, he wishes to assert that " this coming was the result of 
 an invitation. " Both these wishes he might easily realize by 
 using two separate sentences as follows. (1) " I came. " (2) 
 " That was from his invitation. " But linguistic economy urges 
 the reduction of these two structures to one. This result can be 
 obtained only by making one of them some element of the other. ' 
 On the one hand, the assertion that " I came " might be sac- 
 rificed, (1) becoming an element of (2). This would lead to such 
 a sentence as " My coming was from his invitation " or " My com- 
 ing was from that, (therefore). " On the other hand the assertion 
 that " the coming was from the invitation " may be sacrificed, 
 (2) becoming an element of (1). The election of this alterna- 
 tive leads to such a sentence as " I came from his invitation — 
 from that — therefore. " 
 
 It needs no demonstration to show that in the elected form- 
 ula the two words, " I " and " came, " express by some means 
 three terms. The word " I " names an actor. The word 
 " came " names an action. These two have something to do 
 with each other, are in some relation; otherwise they could 
 not constitute a thought. That the relation is that of actor 
 to his own action and that it is, by multiple symbolization, 
 expressed in the word " came, " along with the speaker's belief 
 
 ' Unification by common factor has been eliminated from the present 
 field. 
 
22 Owevr— Meaning and Fimction of Thought- Connectives. 
 
 in its reality, does not need to be proven. For, without such 
 proof, it is plain that in some way all essential elements of 
 thought are presented by the sentence " I came. " That is, the 
 places for terms are all preempted. If therefore a new element 
 is to be added, it must take its place as an adjunct.' 
 
 The connective is then adjunctive, but, as the sequel will 
 show, of a somewhat peculiar type. The adjunct as a rule is 
 coupled in the mind with a single term, as shown by previous 
 examples. But in the case before us the adjunct is coupled 
 with a whole thought. It does not really belong with " com- 
 ing, " except so far as the coming is my own, as distinguished 
 from the coming of any other person. 
 
 The adjunct of this sort may be known as a thought-adjunct 
 in distinction from the ordinary or term-adjunct. With special 
 forms for term-adjuncts languages are generally well-provided, 
 even differentiating sometimes the adjunct of one term (e. g. 
 adjective as adjunct of first or last term) from that of another 
 (e. g. adverb as mid-term-adjunct). But in special forms for 
 thought-adjuncts language seems to be quite deficient. When 
 therefore it is necessary to express the adjunct of a whole 
 thought, that variety of term-adjunct is used as a make-shift, 
 which seems most nearly to meet requirements. The variety 
 elected is the adjunct of the mid-term, the term, that is, which 
 expresses relation. In other words the chosen expression of 
 thought-adjunct is the adverb, in its stricter sense of verb-lim- 
 iter.' The reason for this choice is in part the fact that the 
 relation, as indicated above, is by far the most important 
 thought-el ement. 
 
 ' The use of make-shifts is oiie of the most common linguistic phenom- 
 ena, is indeed one of the most important agents of linguistic progress. As 
 the boy makes use of his jack-knife now as screw-driver, now as gimlet 
 and again as hammer, so a word designed for a particular purpose is forced 
 to serve, in the lack of a better word, the different purpose of the moment. 
 
 To illustrate for the special case in hand, suppose I wish to chetracterize 
 a thought as not my own, but that of another — a zoologist, the thought 
 being that mastodons are pachyderms. I do not say that the mastodons 
 of his mind, are pachyderms; for this is telling you my own opinion of 
 what he ranks as mastodons. Nor do I eay that mastodons are pachyderms 
 of his mind; for this is telling you my opinion of mastodons in terms of 
 
Expression of Thought Connection. 23 
 
 From the preceding reasoning, if correct, it follows that in 
 some way the ideas expressed by " therefore " must form a 
 thought-adjunct and that this thought-adjunct will choose the 
 same linguistic expression as that employed for the adjunct of 
 the mid-term or verb. It remains only to note the details of 
 this arrangement. And in so doing it should be observed that 
 the use of adjuncts is based upon this principle: that the itse 
 of an idea as a term of one thought does not disqualify it from 
 serving at the same time as term in some other thought.^ Ac- 
 cordingly in the thought expressed by the sentence " I came, " 
 let the structural value of the idea expressed by " came " be 
 whatever it will; it is still true that this idea may serve again 
 as first term to a new relation and a new last term. This now 
 is exactly what happens. The coming serves again as the first' 
 term of a new thought. The last term of this new thought is 
 the reinstated invitation. The relational element of the new 
 thought is most conveniently expressed by " from. " 
 
 At this point occurs a slight readjustment of structure. The 
 causal relation and the last term taken together are regarded as 
 in the attributive relation to the first term. That such read- 
 justments do often occur may be shown by a few examples. In 
 the sentence " dogs eat meat " it is certainly possible, and 
 indeed highly convenient, to regard the word " eat " as naming 
 the relation of eater to food. But this relation and the last 
 term, "meat," may be combined in the word "carnivorous;" 
 and the ideal total presented by " carnivorous " may be con- 
 ceived anew in the attributive or objeet-to-quality relation with 
 the first term. This readjusted structure is expressed by the 
 sentence, " dogs are carnivorous. " Again, " a cave under the 
 
 what he ranks as pachyderms. I say rather that mastodons are, in his 
 mind, pachyderms. That is, the relation between the two groups Is de- 
 clared as obtaining in his mind, not my own. To obtain in his mind, this 
 relation must be accompanied in his mind by its terms. The idea of pres- 
 ence in his mind accordingly diffuses from the relation over the other 
 thought-elements. That is: the adjunct which, in form, belongs to the 
 mid-term only, is conceived as, in value, the adjunct of the whole 
 thought. 
 
 ' So a northeast corner-stone serves at the same time as part of a north 
 wall and part of an east wall. 
 
24 Owen — Meaning and Function of Thought- Connectives. 
 
 sea " is paraphrased by the words, " a submarine cave. " In 
 both examples the very word- and sentence-structure shows that 
 relation and last term are combined, and that the combination 
 is conceived in the new attributive relation to the first term. 
 Analogously, the relation named in the word " therefore " (by 
 the element " for " or " from " ) and the reinstated thought 
 (expressed in substantive value by the element " there " ) com- 
 bine into a total, which is conceived anew as in the attributive 
 relation with "came." This attributive relation, being easily 
 inferred, is not expressed. In other words, the elements com- 
 bined under the word " therefore " constitute an adjunct in the 
 last thought. Or, in grammatical parlance, "therefore," with 
 the value of " from that, " is construed like a prepositional 
 phrase; it is adverbially used; it limits "came."' 
 
 The argument for the adverb-theory, thus far general and 
 largely of an a-priori character will be followed by a process of 
 testing. Meantime observe that the word " therefore ' and many 
 other thought-connectives contain an etymological element of 
 so-called demonstrative character, which has been accounted for 
 by giving it reinstative value. But in still other connectives, 
 e. g. " consequently, " no such element appears. Yet the mental 
 total is the same, whichever connective be employed. By the 
 use of the word "consequently" precisely the same result is 
 reached as by the use of the phrase " in consequence of that. " 
 That is, in either case a preceding thought is reinstated. Two 
 explanations of the word "consequently" are, therefore, possi- 
 ble. It may be claimed that by multiple symbolization it actu- 
 ally expresses both relation and reinstated thought. Or it may 
 be claimed that " consequently " names only relation and that, a 
 first term being thereto necessary, the hearer's mind must sup- 
 ply such first term for itself. For plainly the relation between 
 two thoughts cannot be felt with one thought in the mind and 
 
 ' It may add clearneas to this opinion to cite one of the distincter formu- 
 lations of antagonistic opinion. And none seems more distinct than that 
 set forth by Girault-Duvivier, on page 895 of his " Grammaire des Gram- 
 maires." " The conjunction differs from the adverb. It does not modify 
 a verb (nor an adjective, nor an adverb)." " It differs from the preposi- 
 tion. It does not express the relation of one thing with another." 
 
Single JRelations Between Thoughts. 25 
 
 the other out. In both cases, then, the thought-result is the 
 same. The terminology of the first view being simpler, that 
 view will be used in this investigation, as a working hypothesis. 
 That is, every thought-connective will be treated as not merely 
 necessitating, but as accomplishing of itself the reinstatement 
 of a former thought. 
 
 PART II. APPLICATION OP THEORY. 
 
 The remainder of this investigation endeavors (I) to general- 
 ize, for the single thought-relations expressed by the English 
 language, the opinion derived from the examination of the word 
 " therefore; " (II) to exhibit the use of the connective in the expres- 
 sion of multiple relations; (III) to apply the present view to 
 the serial relations which occur in what is known as continued 
 reasoning. It takes into consideration the connectives listed on 
 page 178 of Prof. E. H. Lewis' "History of the English Para- 
 graph." From this list, however, there have been omitted the 
 following as of doubtful connective value : yea, nay, first, add to, 
 so far, hitherto, thus far, once more, it is true, at least, finally, at 
 last, at length, well. On the other hand, a few have been added 
 which seem not to have occurred in the texts examined by Prof. 
 Lewis. 
 
 I. SINGLE RELATIONS BETWEEN THOUGHTS. 
 
 It is preliminarily assumed that in all cases the function of 
 the thought-connective is to exhibit relation between the thought 
 of the moment and a preceding thought. This is but saying 
 that the connective aims to show what, in popular parlance, two 
 thoughts have to do with each other. This being admitted, it 
 is obvious that the study of connectives is first of all a study of 
 relations. To simplify this study it is desirable to classify such 
 relations as are presented by thought-connectives. The diffi- 
 culty of such classification may be inferred from Lloyd Morgan's 
 belief that the ability to fasten the attention directly upon rela- 
 tions is the one power which distinguishes the minds of men 
 from the minds of other animals. With man himself the power 
 
26 Owen— Meaning and Function of Thought- Connectives. 
 
 to focus attention upon relations comes late and the words which 
 distinctly express relations are late in linguistic development. 
 In fact the perfect differentiation of one relation from another 
 would be a maximum of intellectual attainment. It is, therefore, 
 readily conceded that the suggested lines of demarcation be- 
 tween classes of relation are more or less arbitrary. Classifica- 
 tion is accordingly made in the interest, mainly, of convenience. 
 It should further be noted that, on account of the general lack of 
 power to clearly discriminate between relations, there is much 
 inconsistency in the use of relation-naming words. The same 
 word stands for different relations and different words "stand for 
 the same relation. There will then be no attempt to exhaust 
 the possibilities of relation-naming, the present aim being- not 
 so much to establish meaning as to exhibit structure. 
 
 The relation most frequently expressed by thought-connectives 
 may be known as common group-membership. In treating this 
 relation it is important to distinguish at once the different 
 aspects in which a group may be considered. 
 
 In the first place the same group may be conceived as either 
 homogeneous or heterogeneous. For instance, the difference 
 between two objects being emphasized, they may be called 
 ■■ an apple and a pear. " The group in this case is conceived as 
 heterogeneous. On the other hand, their differences being over- 
 looked, they may be called " two fruits. " The same group is 
 now conceived as homogeneous. Omitting the interesting pos- 
 sibilities of groups formable either in time or space from ob- 
 jects, qualities, actions and relations, I merely draw from my 
 illustration the general principle that honiogeneous idea- groups 
 may be expressed by plurals. 
 
 Homogeneous thought-groups are somewhat rare. As would 
 be expected from the expression of thought-adjuncts by mid- term- 
 adjuncts (see p. 22), it appears that the indication of thought- 
 plurality may be accomplished by indicating plurality of mid- 
 term or relation-word. To illustrate, suppose a homogeneous 
 series or group of visits on my part to my uncle; a single one 
 of these may be expressed by the sentence : " I visited my uncle. " 
 If I wish to pluralize this in English, I do so by a pluralizing 
 
Single Belations Between Thoughts. 27 
 
 adverb, obtaining the sentence: "I repeatedly (several times) 
 visited my uncle." In French I may use the form "visitais," 
 the verb itself having in this case a genuinely plural (frequenta- 
 tive) value, not at all to be confounded with that of the spurious 
 plural " visitames. " 
 
 I do not know any case in which thoughts consecutively ex- 
 pressed (the only ones considered in this investigation) are char 
 acterized by the connective as homogeneous. For although 
 thoughts thus characterizable are often connected by "and," it 
 will be seen that they are actually characterized somewhat differ- 
 ently. I therefore neglect homogeneous thoughts, as not handled 
 by connectives, and pass to the examination of heterogeneous- 
 groups. These will be found to be of three sorts: concordant, 
 discordant and neutral. 
 
 Of the members of a concordant group it is conceived that, 
 though they are different, they exist together naturally, with- 
 out occasioning surprise. Thus, age and feebleness are obvi- 
 ously heterogeneous qualities; yet they are grouped together in 
 the description of an individual, without occasioning surprise ;^ 
 indeed, given the first, the second rather fulfills an expectation.. 
 To illustrate the second group- variety, let the same individual 
 be characterized not only by age, but also by activity. The facts 
 of the case permit activity to group itself with age, the two 
 being qualities of the same person. But tl^e presence of activity 
 with age is, to say the least, unexpected. This group may, then, 
 be characterized as discordant. 
 
 To illustrate the last kind of group let the same individual be 
 distinguished firstly, by honesty; secondly, by beauty. The 
 facts of the case permit honesty to group itself with beauty, the 
 two being qualities of the same person. But it should be noted 
 that personal beauty and honesty are not conceived as in the na- 
 ture of things belonging together. That is, there is nothing 
 worthy of the name of concordance. Nor, on the other hand, 
 is there the slightest suggestion of discordance. This group, 
 then, being neither concordant nor discordant, may be called 
 neutral. 
 
 This distinction established, it is practicable to open the study 
 of the relations expressed by thought-connectives with 
 
28 Owen — Meaning and Function of Thought- Gormectives. 
 
 CLASS I. RELATIONS OF GROTJP-MEMBBR TO QEOUP-MEMBBR. 
 
 The group may be 
 
 (a) Concordant, employing the connectives and, and not, nor, 
 
 NEITHER. 
 
 Of these the last three merely combine " and " with a negative. 
 Attention may, therefore, be exclusively directed to " and. " The 
 historical value of this word is now given as " in addition " ; e. g. 
 "And he succeeded." That is, " additionally he succeeded." 
 It is plainly his success that was additional. Accordingly 
 " and," though valued properly as a thought-adjunct, appears 
 in the form of a verb-adjunct or adverb. But such a sentence is 
 as incomplete as the algebraic form " + B. " B is here repre- 
 sented, not as an isolated B, but as a B additional to some 
 •other quantity. So, too, in the example given, his success is 
 not an isolated success, but a success conceived as added to 
 something else. Without this something else the thought and 
 sentence are both incomplete. There is present, so far, a 
 thought expressed by " he succeeded" ; also a relation ex- 
 pressed by " and. " There is wanted another thought between 
 which and the thought expressed this relation may hold. Let 
 the example, then, be expanded into the following form: "He 
 strove. And he succeeded." It is now evident that the sec- 
 ond thought reinstates the first thought as, with the relation, 
 an adjunct of itself. Accordingly, in full, " he succeeded in 
 addition to his striving. " 
 
 Reviewing and stating fully, " He strove " expresses a first 
 thought. " He succeeded " expresses a second thought. The 
 relation between them is that of member to member of a hetero- 
 geneous group. The first thought and this relation combine as 
 adverbial adjunct in the second thought, and are expressed by 
 "and." Again, the group is conceived as concordant. Other- 
 wise the word " but " would have been chosen. Also, the relation 
 is obtained by thinking from his success to his striving, and is 
 accordingly reverse. That is, his success is additional to his 
 striving. The whole meaning may then be presented by the 
 sentences: "He strove. In concordant addition thereto he 
 succeeded. " 
 
Single Relations Between Thoughts. 29 
 
 It is conceded that absolute proof of values can be reached 
 only by an exhaustive examination of examples. The one chosen 
 in this and following cases is to be taken as offered after the 
 Euclidean method, as if I had said: let the above be any 
 example of the use of " and ". It should also be noted that 
 such cases as " I will find him and he be there " are explain- 
 ble by a totally different method and, indeed, have nothing to 
 do with the field under examination. 
 
 (b) Discordant, employing the connectives but, though, yet, 
 
 STILL. 
 
 For example : " He is old. Butjie is active. " In this exam- 
 ple the value of " but " is plainly " in discordant addition to 
 his being old." The two thoughts are conceived as a hetero- 
 geneous group in the sense above established. They are also' 
 further conceived as discordant. The relation is, then, that of 
 member to member of a heterogeneous and discordant group. 
 This relation, obtained by thought-transit from his activity 
 back to his age, is reverse. " But, " then, reinstates in substan- 
 tive function the preceding thought. With this reinstated value 
 it combines the relation. The total thus obtained is used as ad- 
 junct in the last sentence. This adjunct is put as adverb 
 with " is. " 
 
 The use of the word " though " is more peculiar. It should 
 be noted in the first place that this word is mainly used in 
 a strictly prepositional function; e. g. "Though he slay me, 
 yet will I trust him. " The meaning here is plainly that the 
 trusting will hold in spite of the slaying. By its use in such 
 cases " though " acquires the habit of regarding the immed- 
 iately following thought as the last term of the relation which 
 it expresses. Such being the use of the prepositional " though, " 
 the use of " though " as thought-connective offers some embar- 
 rassment. For instance, let it first be stated that " X did not 
 strive. " Let it be added that " he succeeded. " It is desired now 
 to express a discordant relation between his success and his lack 
 of effort. If " though " is put at the head of the second state- 
 ment, the hearer may take it prepositionally, regard the immed. 
 lately following thought as last term and conclude that " in spite 
 of his success X did not strive. " But it is desired that the hearer 
 
30 Owen — Meaning and Function of Tliought- Connectives. 
 
 should understand that " X succeeded in spite of not striving. " 
 To assure this understanding, the connective " though, " is sent 
 to the end of the sentence, giving in full the unequivocal form 
 " He did not strive. He succeeded though. " In this shape the 
 meaning of the last sentence is plainly that " he succeeded in 
 spite of his not striving. " 
 
 The position of " though " is a marked exception to the gen- 
 eral habit of connectives. That it has been occasioned by the co- 
 existent prepositional function of the same word, is further 
 suggested by the German usage of its etymological equivalent 
 " doch. " This word is not embarrassed by co-existent prep- 
 ositional function. It, accordingly, may take the usual place 
 at the sentence-head; e. g. " Er hat nicht gearbeitet. Doch ist 
 es ihm gelungen. " 
 
 It may be added that, so far as the fully expressed thoughts are 
 concerned, it is also in the case of " though " from the last to 
 the first that the thought-transit is taken. In this aspect of 
 the ease, the relation may be called, as usual, reverse. It com. 
 bines as usual with the reinstated first thought to form an ad- 
 junct of the second, this adjunct being put as adverb of the 
 verb. 
 
 "Yet" and "still " do not seem to require comment. 
 
 (c) Neutral, unqualified, employing the connectives also, bk- 
 
 6IDES, TOO, MOREOVER, TDRTHER, FURTHERMORE, AGAIN. 
 
 For example, " He is young. Also he is rich. " As total 
 thought-value of these sentences we obtain " He is young. In 
 addition (neither concordant nor discordant) thereto he is rich. " 
 In the expression of this value there falls to '" also " the duty 
 of reinstating the preceding thought and that of naming the re- 
 lation between it and the following thought. The relation is, as 
 usual, reverse. The relation and the reinstated thought form 
 together- an adjunct of the last thought, being put as adverb 
 with its verb. 
 
 " Also " is further sometimes used as an idea-connective, 
 falling as such without the field in discussion. 
 
 For example : " It is remarkable how long he contrives to re- 
 main young. He is also rich, " i. e. " rich in addition to young. " 
 Eemaining thought-elements are not reinstated. 
 
Single Relations Between Thoughts. 31 
 
 The connective " too " is, for the most part, also confined 
 to partial reinstatement of this sort. As full reinslater, it 
 retires somewhat from the initial position; e. g. "He is 
 poor. His wife, too, is extravagant." 
 
 The other connectives of this class seem to require no com- 
 ment. 
 
 CLASS II. RELATIONS OF SIMILARITY AND DISSIMILARITY. 
 
 (a) Similarity, employing the connectives likewise, in like 
 
 MANNER, SIMILARLY. 
 
 For example, "Fire melts lead. Similarly, the sun melts 
 snow. " That is, the second phenomenon stands to the first in 
 the relation of like to like. Here occur again the reinstatement 
 of previous thought, and the naming of the reverse relation be- 
 tween two thoughts. Reinstated thought and relation combine 
 to form an adjunct of the last thought, but are put as ad- 
 junct to verb of last sentence, i. e., as adverb. 
 
 "Likewise" has for the most part almost, if not altogether, 
 lost this original power of naming the relation of similarity, 
 being almost synonymous with "and" or even "also." 
 
 Of this class examples also occur in which reinstatement is 
 only partial, the reinstative word in such cases not being re- 
 garded as a thought-connective; e. g. " Hes dined in full uni- 
 form. In like manner (i. e. in full uniform) he appeared at ^- 
 the ball. " 
 
 (b) Dissimilarity, employing the connectives contrariwise, 
 ON THE CONTRARY, ON THE OTHER HAND. Thesc are mere oppo- 
 sites of the preceding, and require no discussion. 
 
 It may be noted that the relations of class II are much con- 
 fused with those of class I, and for a very natural reason. Things 
 that are like are easily grouped. And things that are grouped 
 are apt to be conceived as more or less like. The difference be- 
 tween the relations is then merely one of view-point; and pas 
 sage from one to the other is extremely easy. The relations of 
 the first class emphasize group-existence, secondarily conceived 
 as concordant, discordant or neutral. The relations of the sec- 
 ond class emphasize similarity or dissimilarity, which may be 
 further used in determining group-formation 
 
32 Owen — Meaning amd Function of Thought- Gonnectives. 
 
 CLASS III. BELATION OF EQUIVALENT TO EQUIVALENT. 
 
 This class employs the connectives in short, in a word, in other 
 
 ■WORDS, OR, RATHER. 
 
 The first three of these forms have, in the now-considered 
 usage, lost their original special meanings, and coalesced in 
 meaning and function with " or, " which will, accordingly, be 
 chosen as type; e. g. "A is the father of B's father. Or A 
 is the grandfather of B. " In exploiting this example it should 
 be noted first that, historically, " or " is the same as " other, " 
 symbolizing then the relation of difference. But, in the now- 
 considered use of "or," this difference has been restricted in 
 scope to difference in expression or form. At the same time it 
 has joined to itself the idea of substantial equivalence in mean- 
 ing. In short, it is merely a relation of the type which I have 
 elsewhere described as composite. 
 
 Such being the relation-naming value of ''or," it remains to 
 note its reinstative value. This may be readily made to appear 
 by putting the relation only as attribute to the verb, and not- 
 ing what further term it requires. Neglecting then the ele- 
 ment of form-difference, as obvious and unimportant, let " sub- 
 stantial equivalence " name the relation. This relation and some 
 last term are to be used as attributive or adjunct with the last 
 thought. As this adjunct in practice associates itself with the 
 verb, it should have the adverbial form, " equivalently. " This, 
 however, is not all. In the present case and many others (e. 
 g. concordance, similarity above), as soon as a relation with its 
 last term is treated as an attribute of its first term, it is need- 
 lessly further conceived as itself in some relation with its last 
 term. Thus, the form "A equals B" plainly expresses two terms 
 and their relation. But, in the form "A is equal to B, " the 
 speaker, having made equality an attribute of A, goes on to feign 
 a new relation, expressed by "to, " between, " equality " and "B. " 
 That is, the equality which characterizes A belongs to B. Simi- 
 larly, our typical example is readjusted as follows: "A is the 
 grandfather of B equivalently to something. " Emphasizing now 
 the " something, " it is plain that a last term is required and is 
 to be supplied by the reinstatement of the preceding thought 
 
Single Relations Between Thoughts. 33 
 
 In full, then, "A is the grandfather of B equivalently to A's 
 being the father of B's father. " 
 
 Accordingly " or " reinstates the preceding thought and also 
 names a reverse relation between it and the succeeding thought. 
 The relation and reinstatement combine as adjunct of the last 
 thought, being expressed by an adverb to the last verb. 
 
 Of the word " rather " it should be noted that it names a 
 relation of essential equivalence in meaning, but characterized 
 by a preference for the latter of two forms of expression. Thus : 
 "A is' the father of B's father. Rather, A is B's grandfather." 
 The latter, then, is to the former in the relation of preferred 
 equivalent to rejected equivalent. 
 
 CLASS IV. RELATIONS OP ALTERNATIVE TO ALTERNATIVE. 
 
 This class employs the connectives ok, or else, otherwise, 
 
 RATHER. 
 
 For example: "A is the father of B. Or A is the uncle of B. " 
 In the preceding case attention was concentrated upon form of 
 thought-expression. In the present case attention is with- 
 drawn from form of expression and centered on the thoughts 
 themselves. The hearer is presented with two (or more) possi- 
 bilities, of which it is believed that one only is true; but 
 which is true is not known. When strictly employed, as in the 
 logical dilemma, these possibilities are exhaustive, and hence 
 mutually exclusive. Accordingly, the truth of one involves the 
 untruth of the other, and vice versa. But the linguistic method 
 is commonly the following: the first thought is put as true; 
 the second thought is then put as true in case the first is, after 
 all, not true. 
 
 Recalling our illustration, " A is the father of B. Or A is the 
 uncle of B, " let the relation of terms above described be known 
 as alternative. Plainly, this relation does hold between the 
 two propositions: first, that A is B's father; second, that he is 
 his uncle. It is, further, from the word " or " that this rela- 
 tion is learned. Again and obviously, this relation cannot be 
 perceived unless both its terms are in mind. The full meaning 
 of the second sentence is then: "alternatively to A's being B's 
 father, A is B's uncle." The first thought having, as usual, 
 3 
 
84 Owen — Meaning and Function of Thought- Connectives. 
 
 lapsed from attention at the end of the first sentence, is rein- 
 stated in the second sentence. And, plainly, no word in the sec- 
 ond sentence has any approximation to such reinstative power, 
 except " or. " It, therefore, appears thkt " or " first reinstates 
 preceding thought; that "or" further names the alternative 
 relation between preceding and succeeding thought; that the 
 reinstatement and the relation combine, as usual, to form an 
 adverbial adjunct in the last sentence. Also the relation is 
 thought backward — is reverse.' 
 
 " Rather " adds to the value expressed by " or ' that of prefer- 
 ence on the speaker's part for the latter alternative. Thus 
 
 ' The above case of dual indecision naturally suggests the allied case of 
 plural or multiple indecision, which may be illustrated in the sentence, 
 " A is a kinsman of B." This statement plainly contemplates a multitude 
 of possible relations between the two. This multitude is capable of var- 
 ious degrees of reduction, for which language does not usually take the 
 trouble to develop special forms of expression, until the number of possi- 
 bilities becomes very small. For instance, it is hardly worth while to state 
 that A is related to B in some one of six specified varieties of kinship. 
 Indeed, the colloquial phrase " one of the six " was, no doubt, invented 
 to ridicule such useless exactness. Even indecision between three possi- 
 bilities is usually undistinguished from more general indecision. Lan- 
 guages commonly have special expressions for indecision between two. 
 possibilities, but none for indecision between three or more. It is accord- 
 ing not surprising to learn that "or," by derivation, means " the remaining 
 one of too." This meaning offers indeed a hint of the original thought- 
 structure, which was no doubt entirely different from that above suggested, 
 being appparently as follows: " One (of two possibilities is:) A to be the 
 father of B." " The other (of two possibilities is:) A to be the uncle of B." 
 That such was the original thought-structure is further suggested by the fre- 
 quent introductory use, in its expression, of " either " (=one of two); thus, 
 " Either A is the father of B or A is the uncle of B." But in this expres- 
 sion each possibility is asserted, under the restriction, however, of an ad- 
 junct. If then linguistic form is to express actual thought, the idea of 
 possibleness must be found in this adjunct. Accordingly, "either" must 
 suggest the idea of possibility. This possibility is of special kind, and, as. 
 noted above, may be called alternativeness. Also, the first thought is re- 
 garded as the first alternative and the second thought as the remaining 
 alternative. Putting all this into an adverbial phrase, the first part expands 
 into "As the first of two alternatives, A is the father of B, " or "Firstly 
 of two alternatives, A is the father of B." Correspondingly, the second 
 part becomes " As the remaining alternative, A is the uncle of B," or 
 " Secondly of two alternatives, A is the uncle of B." 
 
Single Belations Between Thoughts. 35 
 
 "Rather A is the uncle of B" means in full "As preferred al- 
 ternative to A's being the father of B, A is the uncle of B. " 
 The structure is otherwise obviously the same as in the case of 
 
 or. 
 
 CLASS V. CONDITION RELATIONS. 
 
 (a) Relation of condition to conclusion, expressed by " if " 
 or an equivalent word or phrase. 
 
 This relation, so far as observed, is named in the first part 
 of a complex sentence; e. g. "If you had helped, I should 
 have succeeded." This value of "if" is plainly the same 
 as that of a preposition, as appears in the equivalent " mth 
 your helping. " This value, in combination with the substantively 
 taken following thought, forms an adverbial adjunct of the last 
 thought. But the case falls without the present field. 
 
 (b) Relation of conclusion to condition, expressed by "then," 
 " in that case, " " so. " 
 
 The condition is often independently formulated. It is rein- 
 stated as part of the conclusion by the thought-connective- 
 6. g. " Suppose you had been there. Then you would have 
 helped. " The word " then " means " in that case " or " as con- 
 clusion to that condition. " Accordingly " then " (1) reinstates 
 your supposed presence ; (2) names the relation of conclusion to 
 condition between your helping and your presence. Stated in 
 full, but with modified perspective, the value is accordingly: 
 " Suppose you had been there. Your helping would have been 
 in relation of conclusion to condition with your being there. " 
 This relation, as usual, is reverse. It and reinstatement com- 
 bine as adverb in the la&t sentence. 
 
 CLASS VI. CAUSAL RELATIONS. 
 
 (a) Relation of effect to cause, employing the connectives 
 
 HENCE, CONSEQUENTLY, ACCORDINGLT, SO, SO THAT,' THEREFORE, 
 THUS. 
 
 ' Of these the form " so that " is strictly incorrect. Its proper function 
 is to introduce a result employed as a measure of the intensity of some 
 preceding element; e. g. "The iron is so hot that I cannot hold it." That 
 is, the iron is hot to the degree indicated by my inability to hold it. The 
 different use of "so " in the present class of cases appears in the example, 
 
36 Owen — Meaning and Function of Thmight- Connectives. 
 
 The original (?) value of "so " is " in that degree;'' e. g. "The 
 boy is so tall" (accompanied by an explanatory gesture). From 
 this it has made the somewhat difficult transition to the value 
 of " in that manner, " e. g. " He stood so" (accompanied by a de- 
 scriptive attitude). Both these values, primarily deiktio or 
 demonstrative, easily become anaphoric or reinstative. Thus 
 " The sun melts snow by its heat. So fire melts lead. " That 
 is, " Fire melts lead in a previously mentioned, reinstated 
 manner." Another easy transition develops the meani;ig "by 
 the same agency " or " from the same cause. " The final change 
 to the value of " therefore " is much more violent and invites a 
 passing observation. 
 
 Let X stand for a phenomenon which is neither distinctly nor 
 centrally in mind. Let A and B stand for two other phenomena 
 both distinct and central. Let both these last be somewhat 
 vaguely conceived as caused by X. If now A is described as occa- 
 sioned by X, it is natural to say " So B is occasioned. " But the 
 case offers two possibilities. X may occasion both A and B. 
 Or X may occasion A, which in turn occasions B. That is, X 
 may occasion A directly and B indirectly. Now " so " (= from the 
 same cause) can properly be used of B only when B is, like A, 
 directly caused by X. But it is easy to make the slip of using 
 " so " with a B which is only indirectly caused by X — a B 
 which is directly the result of A instead of being, concomitantly 
 with A, a result of X. That is, what I have elsewhere de- 
 scribed as the " minor relation " of co-effect to co-effect, is con- 
 fused with the major relation of effect to cause and the same 
 word, namely "so," is used to name both relations; e. g. "He 
 invited me. So I came;" i. e. " I came (not as co-effect, with the 
 inviting, of the invitation's cause, but) as effect of the invita- 
 tion (itself)." 
 
 In the preliminary study of connectives the present class was 
 
 " The iron is too hot. So that I cannot hold it." Either the word " that " 
 must be rejected as superfluous or a considerable ellipsis must be filled, 
 " So " alone, unencumbered by " that," will be covered by the explanation 
 which is to follow. If " that " is retained, it is necessary to fill the ellipsis 
 with some such result as "So it is true that I cannot hold it," " So you 
 see, etc.", " So it is plain, etc." All these sentences may be treated like 
 those with "so " alone; but presumably no one of them is really intended. 
 
Single Relations Between Thcyughts. 37 
 
 sufficiently treated in the typical " therefore. " It is enough, then, 
 to note again that in the present cases a preceding thought is 
 reinstated in substantive function. This reinstated thought, 
 combined with relation, is used as adverbial adjunct in the 
 second thought. The relation is reverse. 
 
 (b) Relation of effect to counter-cause, employing the connect- 
 ives HOWEVER, NEVERTHELESS. 
 
 By this is meant the relation which exists between a second 
 phenomenon and a first, when the first does not cause the second, 
 but rather tends]Jto cause the contrary; e. g., "The cable was 
 severely strained. Nevertheless, it held firm. " The strain is 
 plainly conceived as causative. Indeed, there is a vivid sugges- 
 tion of its causing a rupture. Such is its tendency and such 
 would be the result, were it not for the intervention of oppositely 
 causative force or resistance. Such a force is found in the mole- 
 cular attraction of the hempen fibres, and in the friction of fibre 
 on fibre, obtained by torsion. It is of these forces that the hold- 
 ing firm must be regarded as the result. In the present case then 
 the related terms are, first, the cause of a result opposite to the 
 one mentioned ; second, the result of a cause opposite to the one 
 mentioned. The relation is broadly one of antagonism. But this, 
 like every other, has its proverse and reverse aspects. To form 
 suitable names for these, it seems best to start with the names 
 of other causal relations. Beginning, as usual, with the last 
 phenomenon of the example, it is plain that it at least occupies 
 the place ordinarily occupied by a result or effect. Let the 
 holding firm of the cable be, then, conceived as an effect. The 
 relation of the strain to this effect is obviously not that of a 
 cause, but of the opposite of a cause. To name this relation, 
 such words as hindrance, prevention, deterrent, etc. are hardly 
 broad enough. There is need of a word expressing what tends 
 to produce an opposite effect, what tends to hinder actual effect 
 and what merely fails to occasion actual effect. It is to express 
 these meanings that the word " counter-cause " is employed. ' 
 
 The word " nevertheless, " like " therefore, " contains an 
 etymological element, namely " the, " of reinstative and relation- 
 
 ' Plainly the relation might also be known as that of counter-effect to 
 cause; but this terminology has been found less convenient. 
 
38 Owen — Meaning and Function of Thought- Connectives. 
 
 naming power. In the case of " therefore " this element was 
 seen to mean " on account of that " ; but in the present case it 
 means " in spite of that. " " Never " (= not) and " less " con- 
 bine in the meaning "to the same degree," which may be neg- 
 lected as being a mere emphasizer.' The relation expressed by 
 "in spite of " is reverse and the value of "that " reinstative. 
 Their combination, strictly adjunct of the last thought, is 
 taken as adverb to its verb. 
 
 ' This interpretation may be itemized as follows. Imprimis the facts re- 
 quire us to introduce the second statement by some such formula as " an- 
 tagonistically thereto." In developing this value from "nevertheless" 
 we must be careful both with meanings and with their combination. The 
 meanings are obviously " from" (which usually means on account of), 
 " that " (reinstative), "never "(with the essential value of "not") and 
 " less," which defines itself. In the combination of these meanings the 
 important question is what to do with "not." If put with "held," the 
 meaning is that the cable did not hold, which meaning is untenable. If 
 " not " be put with "firm," the meaning is that the cable held not firm or 
 perhaps even weakly, which is untenable. If " not " be put with "it," 
 (the cable), the meaning is again useless. If "not" be put with "on ac- 
 count of," a particular relation is excluded; but as the number of possible 
 relations is immense, the exclusion of a wrong one gives no practical 
 help in finding the right one. The result then of such exclusion would be 
 merely to leave the hearer in total ignorance of the relation between the 
 two thoughts contemplated. This also is presumably not intended. If 
 "not" be put with "that," the preceding thought is excluded, which 
 exclusion again gives no aid in finding a substitute. It remains only to 
 put " not " with " less, " obtaining in full: " The cable was strained. From 
 that not less it held firm." 
 
 In testing this connective formula, let it be borne in mind that the coun- 
 ter-cause is in the present case merely a cause that doesn't work, an ineffec- 
 tive cause. Now a moment's reflection will convince one that, before the 
 mind begins to deal with ineffective causes, it must have become familiar 
 with causes through their effectivity. This is but another way of saying 
 that, in the order of acquisition or development, the effective cause precedes 
 the ineffective. Such being the case, with the well-known habits of lan- 
 guage in mind, it is eminently probable that the connective mechanism of 
 the ineffective is a mere modification of that already in use with the effec- 
 tive. Turning for a moment to the latter, suppose that as the result of strain 
 the cable breaks. I might prefer to say with some'reserve, " From that less 
 (desto weniger) it held firm ." If now I wish to adjust this mechanism to 
 usage with an inefficient cause, I am likely enough, as countless examples 
 might show, to imagine that everything will be well, if I merely thrust in 
 
Single Belations Between Thoughts. 39 
 
 As previously indicated, the boundary lines between relations 
 are frequently crossed, the speaker electing one class, even 
 when another is more conspicuous. Thus the speaker may choose 
 to regard cause and effect as concordant group-members; e. g. 
 "The cable was severely strained. And it broke." Conversely, 
 counter-cause and effect may be ranked as discordant group- 
 members; e. g. "The cable was severely strained. But it held 
 firm. " 
 
 Of the word " however " the first part, namely "how, " has the 
 value of " in any case. " This is merely emphasized in " any- 
 how. " The second part, namely " ever, " is a mere broadener of 
 indefiniteness, as appears in the use of " who " and " whoever, " 
 
 a negative, to match the change from effectiveness to non-effectiveness- 
 Thus I reach, when dealing with an ineffective cause, the form " The cable 
 was strained. From that not less it held." 
 
 Now this " not less " might be taken as suggesting either " as much " or 
 " more." But this latter plainly violates the possibilities of the case. Even 
 "as much" is seen to be a mere emphasizer of what is, in its opposi- 
 tion to what might have been. It is therefore comparatively neglected. 
 On the other hand the attention which has been directed to the possibili- 
 ties of the case develops with ease the fact that the relation conceived can 
 be only that of effect to counter-cause; and the phrase "from that" is 
 henceforth in this formula associated with that relation. 
 
 This then is but one of the many cases in which a word's particular re- 
 lation-naming power is determined by its association with phenomena whose 
 relation is obvious; one also of those in which the same formula has on differ- 
 ent occasions very different meanings. Thus " while," naming strictly time- 
 inclusion, acquires in English the meaning " though " (in spite of); but 
 in German it reaches that of '• because." " Pour " in French means 
 sometimes " for the purpose of," sometimes " because of " and sometimes 
 again " in spite of." So too the English " f or " introduces purpose (e. g. 
 "a push for liberty"), reflects upon cause (e. g. "imprisonment for 
 theft "), or names antagonism (e. g. "a man's a man for a' that "). Further 
 cases offer later. 
 
 The word " nevertheless" is employed in an interesting range of nega- 
 tive sentences; e. g. " The cable was severely strained. Nevertheless it did 
 not break." And here it is worth while to note the absurdity of the often- 
 uttered doctrine that two negatives necessarily destroy each other, leaving 
 an afllrmative. According to this doctrine " not " and " never " should 
 cancel each other, leaving: " On that account the less did it break." 
 The whole value then becomes: " The cable was severely strained. It held 
 the more firmly on that account." 
 
40 Owen — Meaning and Function of Thought- Connectives. 
 
 etc. Using "however" as connective in the former example, I ob- 
 tain " The cable was severely strained. However, it held firm. " 
 Strictly interpreting, " The cable held firm in any case what- 
 ever. " ' " In any case " means here "in case of the occurrence 
 of any other phenomenon. " This exceedingly broad reference to 
 other phenomena is restricted in usage to prior phenomena. Now 
 a prior or a previously mentioned phenomenon may be regarded as 
 (1) favorable to the occurrence of the last phenomenon, (2) unfav- 
 orable to the occurrence of the last phenomenon. Usage has 
 elected^ the unfavorable aspect to the exclusion of the other, de- 
 ciding that the relation between last phenomenon and first, 
 named by "however," shall be that of effect to counter-cause. 
 In other words " however " means " in spite of what precedes. " 
 Accordingly, as usual, the thought-connective makes a reinstate- 
 ment and names relation between last thought and reinstated 
 thought. This reverse relation and reinstatement combine as 
 adverbial adjunct in last thought. > 
 
 (c) Relation of cause to effect, employing the connectives for, 
 
 BECAUSE. 
 
 Both these words have suffered change of value of a new sort. 
 Instead of naming as formerly and in other cases an effect-to- 
 cause relation, they have been forced to name a cause-to- 
 effect relation; that is, their relation-value has been reversed.' 
 
 The primary usage of "for " is seen in the following quotation 
 from Richard II, 1. 4. " And, for our coffers . . are grown some- 
 what light, we are enforced to farm our royal realm. " Simpli- 
 fying this for convenience, " We are enforced to farm, for our 
 coffers are light. " That is, " for, " as preposition, governs the 
 
 ^So, " Two gentlemen of Verona" I. 1. "However, but a folly," etc, 
 " However " is here defined by Knight as " in whatsoever way;" better, 
 " in any case "or " in either case." 
 
 '^ This election has presumably been influenced by the use of " however" 
 to introduce concessions; e. g. " However much the cable was strained it 
 did not break." 
 
 ' Such reversal may seem at first thought diSicult. It is however ac- 
 complished by every passive verb; e. g. in the active voice the verb " em- 
 ploy " names the relation of employer to employee; thus, "A employs 
 B." In the passive, the same verb names the reversed relation of em- 
 ployee to employer; thus, " B is employed by A." 
 
Single Relations Between Thoughts. 41 
 
 following substantive clause as its direct object. The preposi- 
 tion •■ for " names a relation of effect to cause. That is, the en- 
 forcement to farm is " for, " or " from, " or • on account of " the 
 lightness of coffers. The relation-naming preposition and its 
 object form an adverbial limiter. Thus, the causal sentence is 
 an adjunct of the effect sentence. Such is the case, also, in the 
 example: "I like him for he is kind." That is, I like him on 
 account of his being kind. 
 
 But frequently this structure is replaced by another very dif- 
 ferent. The effect, e. g. my liking him, is first stated alone; 
 thus, " I like him. " Then follows the cause stated alone, but re- 
 garded as explanation of the preceding effect-sentence; thus, 
 "For he is kind." It is now no longer proper to regard the 
 last sentence as meaning • on account of his kindness. " ' " His- 
 being kind " is, rather, now the dominant thought. And any 
 relation between his* kindness and my liking must be conceived 
 in such an aspect that it may stand naturally as an adjunct 
 to " his being kind. " That is, this relation must, according to 
 the chosen habit of language, be reached by a mental transit 
 which starts with his kindness and ends with my liking. The- 
 relation thus established is that of cause to effect. In the pres- 
 ent example, therefore, it is this relation which must be expressed 
 by " for. " Testing this statement by substitution, the follow- 
 ing meaning results : " I like him. Causatively or explanatorily 
 thereof he is kind." That is, his kindness is in the cause-to- 
 effect relation with my liking. 
 
 When " because " heads the second sentence, it has the same 
 value. 
 
 In connectives of this class it appears then that the relation 
 is, as usual,' reverse. The connective makes, also, the usual re- 
 instatement. This and relation, combined as adjunct, are treated 
 in the usual way. 
 
 (d) Relation of counter-cause to effect, employing the connec- 
 tives NOTWITHSTANDING, NEVEKTHELESS, HOWEVER. 
 
 For example: "I like him. Notwithstanding, he is unkind." 
 
 ' Obviously, " For he is kind " might be taken as a mere supplement, 
 total value being the single thought: "I like him for his kindness." But 
 this meaning is not elected in the present case. 
 
42 Owen — Meaning audi, Function of Thought- Connectives. 
 
 The same reversal of values has occurred in this case as in the pre- 
 ceding. Had the above been put as a single sentence, the value 
 would have been : " My liking is in spite of his unkindness. " 
 That is, my liking would be in the relation of effect to counter- 
 cause with his unkindness. But in the actual form of the ex- 
 ample there stands first an independent statement of my liking. 
 Then follows a statement of his unkindness. His unkindness 
 does not tend to produce my liking. Indeed, it tends either to 
 hinder this liking or to produce the opposite. The unkindness 
 therefore ranks properly, not as a cause, but as a counter-cause. 
 The value of the example then, taken in this sense, is as fol- 
 lows: "I like him. Counter.causatively thereof he is unkind. " 
 Or " I like him. Unfavorably thereto, he is unkind. " - That 
 is, unkindness is the starting-point in relation-formation; and 
 the relation is the opposite of that of cause to effect, i. e. coun- 
 ter-cause to effect. 
 
 Relation then is of the usual type. It is combined with a rein- 
 stated first thought as adjunct of a second thought. This adjunct, 
 as usual, appears as adverb to the verb of the second thought. 
 
 The word " nevertheless" exhibits the present relation some- 
 what more distinctly; e. g. "The wood burns badly. Neverthe- 
 less it is dry." Of this example I elect the following meaning: 
 " The wood burns badly. Counter-causatively thereof it is dry. " 
 The relation indicated is that of counter-cause to effect, the re- 
 verse of the relation expressed by "' nevertheless " on p. 37. 
 The present relation is as usual obtained by a transit from the 
 latter to the former thought and is combined with a reinstate- 
 ment of the former thought. The combination, though adjunct 
 to the latter thought, is treated as an adverb of its verb. 
 
 "However" agrees in meaning and function with' "neverthe- 
 less. " 
 
 It is evident that "nevertheless " in the last example. may be 
 taken with an entirely different meaning, the total value in such 
 
 ' In another possible meaning of this example his unkindness is an effect 
 not overcome by my liking conceived as cause. This meaning may be ex- 
 pressed by saying that " I like him. In spite of my liking him he is un- 
 kind to me." This meaning which merely illustrates the relation of effect 
 to counter-cause is not elected in the present case. 
 
Single Relations Between Thoughts. 43 
 
 case being " The wood burns badly. In spite of that, it is dry. " 
 This interpretation brings us out of the field of external causa- 
 tion into that of internal or mental causation. In this field the 
 ill-burning of the wood is a natural cause of the opinion that it 
 is wet. The opinion that it is dry subsists in spite of the rec- 
 ognition that it burns badly. The example then illustrates the 
 counter-causative in the intellectual field of action. 
 
 I have not thought it necessary to discriminate between the 
 treatment, in causal relations, of the external and the internal; 
 for I see no cases in which treatment differs in principle. What 
 is known as the relation of " Grund und Folge " or " datum to 
 inference, " I have then left undiflerentiated from other causal re- 
 lations. To one peculiar form of the mentally causal I however 
 invite passing attention, giving it the name, for lack of a bet- 
 ter, of 
 
 r 
 CLASS VII. RELATION OF DECISION TO MUTUALLY CONFLICTING DATA. 
 
 This class employs the connective on the whole. 
 
 For example, " X is polite, amusing, vain and fickle. On the 
 whole I like him. " The connective is defined by Webster as mean- 
 ing "all things considered," "in 7iew of all the circumstances." 
 Or, more strictly, the final statement is in a relation of mental 
 attendance on all details of the preceding. That is, all are co- 
 present in consciousness. This connective might be used when 
 all data favor the same decision. But as a matter of fact, and 
 indeed naturally enough, it appears to be employed only when 
 the data partly favor and partly oppose the decision. In such 
 case the speaker's assurance that his decision is accompanied by 
 a consciousness of both favorable and unfavorable data is quite 
 in order. It is indeed natural to regard the following thought 
 as in relation, partly of effect to cause, and partly of effect to 
 counter-cause, with the preceding data. But it is «nore natural 
 to believe that the speaker's opinion is in the single relation of 
 effect to cause with the resultant of the conflicting influences. 
 Or, letting the causes constitute a minuend and the counter-causes 
 a subtrahend, the remainder is cause, of which the decision is 
 effect. The thought-structure, in either view, is that employed 
 in the other cases considered. A full reinstatement, or a rein- 
 
44 Owen— Meaning and Function of Thought- Connectives. 
 
 stated remainder, and a reverse relation combine as adverbial 
 adjunct in the last sentence. 
 
 I have not yet found any case in which the reverse of this re- 
 lation is expressed by a thought-connective. 
 
 CLASS VIII. TIME KELATIONS. 
 
 (a) Precedence, employing the connectives previously, etc. 
 For example, " He called on me. Previously he called on you. '* 
 
 The idea of previousness has no practical value, unless accom- 
 panied by that from which it is reckoned. The necessary sup- 
 plement is obtained by reinstating the preceding thought. Ac- 
 cordingly, in full: "Previously to" or, more simply, "before 
 his calling on me, he called on you. " As usual, a reinstatement 
 and a reverse relation combine as the adverbial adjunct in the 
 last sentence. 
 
 (b) Contemporaneousness, employing the connectives at the 
 
 SAME TIME, SIMULTANEOUSLT, etc. 
 
 With these the linguistic method is the same, and does not 
 seem to require comment. 
 
 (c) Subsequence, employing the connectives subsequently, 
 
 LATER, THEREUPON, etc. 
 
 This case is the mere opposite of (a) and, as such, does not 
 appear to need examination. 
 
 CLASS IX. INDEFINITE RELATIONS. 
 
 This class employs the connectives now, then, etc. 
 
 These seem to be used, for the most part, to introduce a 
 thought conceived in a general way as in relation with the pre- 
 ceding, yet without specification of a particular relation; e. g. 
 " Now Barabbas was a robber." So far as the unaided power of 
 " now ■■ is concerned, it may mean " explanatorily of the preced- 
 ing, or antagonistically to the preceding. " Indeed, it is hardly 
 safe to say that the meaning is anything more than " relevantly 
 to the preceding. " That is, the last thought is, with the pre- 
 ceding thought, in some relation. This relation is not partic- 
 ularized and must be discovered by the hearer from the context. 
 The cause of such indefiniteness may often be found, no doubt, 
 in the mental status of the speaker. He has not himself estab- 
 lished a particular, definite relation. He only perceives 
 
Single Relations Between Tlumghts. 45 
 
 that his exposition is iuoomplete without the introduction of 
 his last statement. That is, he perceives, merely, that there 
 is some sort of relation.' Thought-structure is of the usual 
 type. What has preceded is reinstated. The reinstatement 
 and the general relation unite as adverbial adjunct in the last 
 sentence. The relation, as usual, is reverse. 
 
 CLASS X. FIGURATIVE RELATIONS. 
 
 This topic is suggested by the use of " now " in the preceding 
 example. " Now " strictly means " at the present time. " But 
 in common usage it has come to mean " at the same time as some- 
 thing else " or " immediately after something else." That is, it 
 names a time relation. In the last example it evidently passes 
 from time relation to general relation. 
 
 Time relations are also notoriously prone to be misunderstood 
 as even causal, this fact giving rise to the dictum "Post hoc; 
 ergo propter hoc. " In other words, the passage from time re- 
 lations to other relations is so easy, that it is often made unin- 
 tentionally. The fjjequent figurative use is, accordingly, not 
 surprising. 
 
 Space relations also have their figurative employment; e. g. 
 "thereupon" passes from space to time. Also "further" 
 passes from expression of increase in distance to that of augmen- 
 tation in thought-total. Such change in value concerns rather 
 lexicology than thought-structure, and does not seem to require 
 investigation. For, though the meanings of connectives change, 
 their function appears to remain the same. That is, they are 
 found to combine, in both literal and figurative meanings, a 
 reinstatement and a reverse relation, as adverbial adjunct in the 
 last sentence. 
 
 ' There is a somewhat analogous tendency to use the forms of exact and 
 important relation even when the question may fairly rise whether any re- 
 lation is conceived beyond that of mere co-presence in the speaker's mind. 
 The abuse of " thus " and " therefore," of " hence " and " consequently " 
 offers a notorious illustration. Similarly " once upon a time " goes through 
 the motions of naming a time. '* A certain man " has the air of selecting 
 a particular person. Usage of this sort, the employment of definite names 
 for indefinite ideas or no ideas, I have elsewhere classed as counterfeit 
 symbolization. 
 
46 Owen — Meaning and Function of Thought- Connectives. 
 
 II. MULTIPLE RELATIONS BETWEEN THOUGHTS. 
 
 Several relations obtain between a latter thought and one for- 
 mer thought. 
 
 It is evident that more than one mental transit may be made 
 between the same two thoughts; that each transit may develop 
 a relation more or less different from that obtained by the others ; 
 that the speaker may elect to express more than one of these 
 relations ; e. g. " A = B X C. Now therefore also C = A — B. " 
 " Now " indicates relation generally and serves as a note of warn- 
 ing to watch for relation of a more definite character. " There- 
 fore " names the effect-to-cause relation, which is obviously of 
 prime importance. " Also " ranks the following statement as 
 one of a group of truths — puts it on a footing of parity with 
 its predecessor.' 
 
 To such heaping-up of connectives English is not particularly 
 prone. It is much more affected by Greek and German. Its 
 investigation promises much psychological interest, but is hardly 
 necessary to the present discussion; for the method pursued is 
 apparently the same, whether one or many connectives be em- 
 ployed. 
 
 Several relations obtain, each between a latter thought and one 
 of two or more former thoughts. 
 
 It has already been noted that the connective may reinstate 
 part only of a preceding thought. Such use of the connective 
 lies without the present field. But it is interesting to observe 
 that this usage may coexist with that which has been specially 
 considered ; e. g. " My brother being invited, declined. My 
 sister was most strongly urged. Nevertheless she also de- 
 clined." "Nevertheless" puts her declining as opposed to her 
 
 ' It is pertinent to ask, at this point, what, to the current understanding 
 ■ of connectives, may be the motive for their simultaneous use. Is it doubt- 
 ful whether the joining process will work? Is it sought to insure results 
 by a multiplicity of conjunctive efforts? Is language at the stage of develop- 
 ment characterized in medicine by the use of the shot-gun prescription? 
 Or is a single bond of union too feeble? Must it be supplemented by other 
 bonds of possibly equal weakness? Has conjunctive prudence learned of 
 the spider to strengthen its bond by multiplying cooperative filaments? 
 
berial Relations Between Thoughts. 47 
 
 being urged, reinstates the whole preceding thought. " Also "■ 
 can be taken only as reinstating a part of the first statement. 
 Accordingly one connective reinstates the preceding thought, 
 while the other reinstates a part of the thought before the pre- 
 ceding. 
 
 Eeference to different thoughts is also at least possible with 
 full connectives ; e. g. , " My brother accepted. My mother greatly 
 needed help. But my sister also came. " That is, her coming 
 is put by "but ' as in spite of her mother's need. "But " then 
 reinstates the preceding thought in one relation. On the other 
 hand " also " groups her coming with my brother's acceptance. 
 " Also " accordingly reinstates the thought before the preceding 
 in another relation. At least the sentence may be so inter- 
 preted, which is all that the example is asked to show. 
 
 One relation obtains between a latter thought and each of two 
 or more preceding thoughts. 
 
 That is, the same connective may simultaneously reinstate two 
 different thoughts; e. g., "He is old. He is also feeble. 
 Nevertheless he is active." The last statement may be taken 
 as meaning that he is active in spite of both his being old and 
 his being feeble. 
 
 One relation obtains between a latter thought and the total of 
 two or more JorTner thoughts. 
 
 Thus, " A is the father of X. X is the father of B. Or A is B's 
 grandfather." The last statement means that " A's being B's 
 grandfather is equivalent to the two preceding statements com- 
 bined " — but not to either alone. 
 
 So again "John is a man. All men are animals. Therefore, 
 John is an animal. " The syllogistic cogency lies in the rein- 
 statement of both premises in the statement of the conclusion. 
 
 III. SERIAL RELATIONS BETWEEN THOUGHTS. 
 
 If now the preceding reasoning be correct, the connection 
 of thoughts appears in emphasized effectiveness, especially 
 in the expression of what is called connected reasoning. To 
 illustrate, let A, B, C etc. stand for a series of thoughts, say 
 n in number. The process of their expression is as follows. 
 
48 Owen — Meaning and Function of Thought- Connectives. 
 
 First comes a sentence declaring A. Then follows a sentence 
 declaring B and, further, the relation between B and reinstated 
 A. Next appears a sentence declaring C and, further, the rela- 
 tion between C and reinstated B. But the B which is on this 
 ■occasion reinstated, is a B which contained A by a previous 
 reinstatement. That is, C, in a way, contains both A and B. 
 And so also N, when reaiched, will be found to contain all the 
 thoughts from A to M inclusive. That is, by a series of rein- 
 statements, the last statement of the series is made to contain the 
 last thought plus all preceding thoughts and their relations to 
 each other. And such, no doubt, is the thought actually pres- 
 ent, at the end of an extended reasoning process, in the mind 
 of an intelligent reasoner. 
 
 The succession of thoughts is often called a chain. It were 
 better to compare it to the sphere-in-sphere of the Chinese carver. 
 The smallest sphere is contained in a larger, and that again in 
 one still larger, the outer .sphere containing all the others by 
 successive inclusion. 
 
 Like this final sphere is the final thought in continued reason- 
 ing. 1 do not mean that, as the complexity of the total 
 thought increases, each particular member preserves its original 
 distinctness of parts. Rather each thought, presented first 
 as a combination of elements, enters its successor as a unit, used 
 with other units to form a second combination. And this new 
 -combination fading in turn upon the thinker's consciousness, 
 its elements blend together till they also lose their discreteness, 
 appear as one, and as one form part of still another combina- 
 tion. 
 
 But the last thought of the series contains some representa- 
 tion of every preceding thought. In other words, successive 
 thovight-connection is successive thought-incorporation. 
 
 Madison, Wis., February, 1898. 
 
 Though attention has been confined to English, it may be added that the 
 connectives of other languages, so far as examined, seem without excep- 
 tion to confirm the opinions advanced. Those of German in particular 
 (e. g. dagegen, dessenungeachtet) are conspicuously rich in elements of 
 nnmistakably reinstative and relation-naming value.