■ 9724 IAAY2 01948 a RRMW3 'B* A UdXl 1954 D Ml Jtlt*^T97IF" M, Cornell University Library BF412 .C5ff American men of letters : their nature a olin 3 1924 029 194 177 I Cornell University ^ Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31 9240291 941 77 AMERICAN MEN OF LETTERS Their Nature and Nurture EDWIN LEAVITT CLARKE, M. A. Aamstant Professor of Economics and Sociology, Hamilton College Sometime University Fellow, Colvmbia University SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE Faculty of Political Science Columbia University NEW YORK 1916 AMERICAN MEN OF LETTERS Their Nature and Nurture EDWIN LEAVITT CLARKE, M. A. Assistant Professor of Economics and TSoelology, Hamilton OoUege Sometime University Fellow, Golwmbia University SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE Faculty of Political Soence CoLiMBiA University NEW YORK I9I6 e.v. 5^ COPYRIGHT, I916 BY EDWIN LEAVITT CLARKE MY FATHER AND MOTHER TO WHOM I OWE THE NATURE AND NURTUKE WHICH MADE THIS STUDY POSSIBLE PREFACE The plan of this dissertation was conceived in 1911, as a result of reading the fascinating pages of Professor Lester F. Ward's Applied Sociology. Ward's work was based on an inductive study of the nature and nurture of French men of letters, Alfred Odin's Genese des Grands Hommes. Ward had been profoundly impressed by Odin's work. In the Applied Sociology he suggested the desir- ability of making other inductive studies which should be modeled after Odin's, and applied to many nations and fields of activity. When this study was undertaken, the author believed that the opinions advanced in Ward's work were in every way justified by the evidence. Results attained in his own work, however, have convinced him that nurture is not predominant over nature to the extent that Ward sup- posed. Nevertheless he still agrees that Ward's plea for the socialization of opportunity is quite warranted. If this study in any way strengthens the case for the extension of opportunity to any who are at present denied their birth- right, he will feel that the work has served its purpose. In conclusion, acknowledgments are due; first, to the many authors and relatives of authors who courteously an- swered the questionnaire sent to them; second, to teach- ers in Columbia and colleagues in Hamilton College who reviewed the statistical' aspects of the work, and finally, to Dr. Alvan A. Tenney of Columbia University, for his con- structive criticism and generous advice. E. L. C. January, 1916. 7] 7 CONTENTS rACK Preface 7 List of Tables 11 CHAPTER I Theories of Nature and Nurture The Theory of Galton 13 The Theory of Ward 15 The Theory of the Importance of Both Nature and Nurture ... 17 CHAPTER II Method of Investigation General Statement of Method 20 Odin's Definition of Men of Letters 21 Definition of American Men of Letters 22 Odin's Method of Compiling a Roll of Men of Letters 24 Method of Compiling a Roll of Men of Letters for the Present Study . 27 Classification of Men of Letters According to Prominence . . 31 Collection of Data. . . . 32 CHAPTER III Analysis and Interpretation of Data Method of Analysis 34 Thesis of the Present Study 3S Influence of the Environment 37 The Rise and Decline of American Letters 37 Social Environment 40 Geographic Environment 49 Local Environment 61 Educational Environment 66 Economic Environment 71 Environment as Indicated by Occupation of Parents of Men of Letters 73 9] 9 lO CONTENTS [lO PAGB Environment as Indicated by the Occupation of Men of Letters Themselves 76 Environment as Indicated by Early Religious Training. ... 80 Environment as Indicated by Birth-rank 82 Criticism of the Theory of Galton 84 Criticism of the Theory of Ward 90 Influence of Heredity 91 CHAPTER IV Summary and Conclusions 95 APPENDIX A Literary Families 103 APPENDIX B Biographical Tables 107 APPENDIX C American Men of Letters, Classified by Conjugal Condition, BY Period of Birth, and by Median Number of Children Born to Them 164 APPENDIX D Alphabetical List of American Men of Letters With Date of Birth 165 LIST OF TABLES I. Number of American Literati Born Prior to 1851, Clas- sified by Period of Birth • • ■ 38 II. Absolute and Relative Numbers of American Literati of White Race, Born within the Present Territorial Limits of Continental United States Prior to 1851, Classified by Period of Birth . . 39 III. American Literati Classified by Sex and by Rank, by Period of Birth ... 42 IV. American Literati Classified by Field of Chief Activity and Period of Birth . . 43 V. American Literary Women, Classified by Field of Chief Activity and Period of Birth 45 VI. American Literati Classified as of One, Two, or Three or More Fields of Activity, by Period of Birth. 46 VII. American Literati Classified by Field of Chief Activity and Region of Birth ... 47 VIII. American Literati Classified by Sex and by Rank, by State or Province of Birth, Together with the Relative Fecundity in Literati of Each State or Province. 50 IX. American Literati Classified by Rank, by Region of Birth. 53 X. American Literati Classified as of One, Two, or Three or More Fields of Activity, by Region of Birth 54 XI. American Literati Classified by State of Birth and Period of Birth . . SS XII. Relative Literary Productivity of the Several Groups of States -57 XIII. American Literati Classified by Rank and by Character of Birth-place (State Capital, Chief City of State, County Seat, and Other Places) .... 60 XIV. Absolute and Relative Numbers of Literary Persons Born in Important Cities. • . . 63 XV. Education Received by American Literati, Classified Ac- cording to Field of Chief Activity. . . 67 XVI. Education Received by American Male Literati, Clas- sified by Period of Birth 69 II] II (2 I-IST OF TABLES [l2 PAGE XVII. Education Received by American Women of Letters, Classified by Period of Birth . 69 XVIII. Education Received by American Literati, Classified by Sex and by One or More than One Field of Activity 70 XIX. Early Economic Environment of American Literati ^2. XX. Occupations of the Fathers of American Literati . • 74 XXI. Occupational Distribution of American Male Literati • 77 XXII. Occupational Distribution of American Male Literati, Classified by Period of Birth 78 XXIII. Occupational Distribution of American Literary Women. 79 XXIV. Early Religious Training of American Literati, Classified by Region of Birth - 80 XXV. American Literati Classified According to Size of Family and Birth-rank .... . ... 82 XXVI. A. Per cent Distribution of American Literati Born in the United States, Classified According to the Nation- ality Strain of Their Origin, as Indicated by Surnames. 88 B. Per cent Distribution of the White Population of the United States (1790), Classified According to the Na- tionality Strain of Their Origin, as Indicated by Sur- names . . . ... 88 XXVII. Literary Relatives of American Literati, Classified Ac- cording to Degree of Relationship 93 CHAPTER I Theories of Nature and Nurture This monograph summarizes a study of the nature and nurture of American men of letters. The task attempted was to isolate for investigation the chief factors in each of these influences, to throw some Hght on the importance of each in the development of men of letters, and to show the bearing of the facts discovered on some of the chief theories of nature and nurture. Sir Francis Galton makes a very satisfactory statement of the meaning of the terms nature and nurture when he says: The phrase " nature and nurture " is a convenient jingle of words, for it separates under two distinct heads the innumer- able elements of which personality is composed. Nature is all that a man brings with himself into the world; nurture is every influence from without that affects him after his birth. The distinction is clear: the one produces the infant such as it actually is, including its latent faculties of growth of body and mind; the other affords the environment amidst which the growth takes place, by which natural tendencies may be strengthened or thwarted, or wholly new ones im- planted. Neither of these terms implies any theory; natural gifts may or may not be hereditary ; nurture does not especially consist of food, clothing, education or tradition, but it in- cludes all these and similar influences whether known or unknown.^ ' Francis Galton, English Men of Science: Their Nature and Nur- ture (London, 1874), p. 12. 13] 13 14 AMERICAN MEN OF LETTERS [14 Throughout this study the terms nature and nurture are used in the sense of Galton's definition. There are three important theories of nature and nur- ture on which impinge the facts presented in this study. These theories are briefly summarized in the following paragraphs. Galton states clearly the position of those who hold that nature is stronger than nurture. His opinion can be pre- sented fairly by brief quotations from his classic work, Hereditary Genius.^ His first proposition is stated in the opening sentence of the volume, as follows : " I propose to show in this book that a man's natural abilities are derived by inheritance, under exactly the same limitations as are the form and physical features of the whole organic world." In the second place Galton argues for the preponderant in- fluence of nature over nurture, saying: I believe, and shall do my best to show, that, if the " eminent " men of any period had been changelings when babies, a very fair proportion of those who survived and retained their health up to fifty years of age, would, notwithstanding their altered circumstances, have equally risen to eminence.^ A little later Galton says : I have endeavored to show in respect to literary and artistic eminence — r. The men who are gifted with high abilities . . . easily rise through all the obstacles caused by inferiority of social rank. 2. Countries where there are fewer hindrances than in England, to a poor man rising in life, produce a much larger proportion of persons of culture, but not of what I call eminent men. • Francis Gallon, Hereditary Genius (London, 1869). 'Ibid., p. 38. ,,, I^] THEORIES OF NATURE AND NURTURE jc 3. Men who are largely aided by social advantages, are unable to achieve eminence, unless they are endowed with high natural gifts.^ Finally, Galton seeks to show that the great differences in the achievement of nations are due almost solely to differ- ences in the innate ability of their citizens. Nowhere does he express this idea in a single sentence, but it is discussed at length in a chapter on " The Comparative Worth of Different Races." * There Galton contrasts whites with negroes and ancient Greeks with modern Englishmen, ar- guing in each case that superior achievement is due almost entirely to superior natural ability. These brief quotations and statements serve to presentV the most important part of Galton's theory, namely, that irrespective of environmental conditions, innate ability ac- counts chiefly for the appearance of leaders in nations and for the superiority of one nation over another. In diametrical opposition to this point of view stands the theory championed by Professor Lester F. Ward. He believes that a favorable environment accounts almost en- tirely for the appearance of genius.' To use his own words : ... So far as the native capacity, the potential quality, the " promise and potency " of a higher life are concerned, those swarming spawning millions, the bottom layer of society, the proletariat, the working classes, the " hewers of wood and drawers of water," nay, even the denizens of the slums ... all these are by nature the peers of the boasted " aris- tocracy of brains " that now dominates society and looks 1 Francis Galton, op. cit., pp. 42-43. /y^ ' Ibid., ch. XX. 'Lester F. Ward, Applied Sociology (Boston, 1906). 1 6 AMERICAN MEN OF LETTERS [i6 down upon them, and the equals in all but privilege of the most enlightened teachers of eugenics.^ Again Ward says : The amount of visible genius has never exceeded one-tenth of I per cent, but it is proved that at least two hundred times as much exists and might be brought out. This would raise it to 20 per cent. But when we recognize the many forms that genius takes we cannot escape the conclusion that some measure of genius exists in nearly everyone. All this genius is scattered somewhat uniformly through the whole mass of the population.^ Finally Ward remarks : It turns out, then, that after all the discussion of heredity, and the hopes hung upon the idea of utilizing it in the interest of race improvement, it is a fixed quantity which no human power can change, while the environment, which Galton af- fected to despise, is not only easily modified, but is in reality the only thing that is modified in the process of artificial selection, which is the essential principle of eugenics itself. All the improvement that can be brought about through any of the applications of that art must be the result of nurture, and cannot be due to any change in nature, since nature is incapable of change.* Ward's theory is thus, apparently, in irreconcilable oppo- sition to that of Galton. Ward seems to hold that im- proved nurture is the only means of improving the race, which is worthy of consideration, while Galton seems to hold that only improvement of the blood of the nation can permanently advance society. ' " Eugenics, Euthenics and Eudemics," The American Journal of Sociology, vol. xviii, p. 754. ' Ibid., p. 744. « Ibid., pp. 749-750. 17] THEORIES OF NATURE AND NURTURE ly In marked contrast to these two extreme views stands the opinion of more moderate sociologists, who hold a third theory that both nature and nurture are important As Professor Charles H. Cooley says : Nothing is more futile than general discussions of the relative importance of heredity and environment. It is much like the case of matter versus mind; both are indispensible to every phase of life, and neither can exist apart from the other : they are coordinate in importance and incommensurable in nature. One might as well ask whether the soil or the seed predomin- ates in the formation of a tree, as whether nature does more for us then nurture.^ ' Charles Horton Cooley, Social Organisation (New York, 1909), p. 316. Professor Edward L. Thorndike admirably clarifies the whole sub- ject when he says: It is impossible at present to estimate with security the relative shares of original nature, due to sex, race, ancestry and accidental variation, and of the environment, physical and social, in causing the differences found in men. One can only learn the facts, and interpret them with as little bias as possible, and try to secure more facts. . . . Many of the false inferences about nature versus nurture are due to neglect of the obvious facts : that if the environments are alike with respect to a trait, the differences in respect to it are due entirely to original nature; that if the original natures are alike with respect to a trait, the differences are due entirely to differences in training; and that the problem of relative shares, where both are effective, includes all the separate problems of each kind of environment acting with each kind of nature. Any one estimate for all cases would be absurd. Many disagreements spring from a confusion of what may be called absolute achievement with what may be called relative achievement. A man may move a long way from zero, and nevertheless be lower down than before in comparison with other men : absolute gain may be relative loss. One thinker may attribute differences in achievement almost wholly to nurture, while another holds nature to be nearly supreme, though both thinkers possess just the same data, if the former is thinking of absolute and the latter of relative achievement. . . . The influences of environment are differential, the product vary- ing not only in accord with the environmental force itself, but in accord with the original nature upon which it operates. Edward L. Thorn- dike, Educational Psychology, Briefer Course (New York, 1914), pp. -397-398. l8 AMERICAN MEN OF LETTERS [i8 Cooley further presents his position in the form of a simile, as follows : Suppose that one were following a river through a valley, and from time to time measuring its breadth, depth and current with a view to finding out how much water passed through its channel. Suppose he found that while in some places the river flowed with a swift and ample current, in others it dwindled to a mere brook and even disappeared altogether, only to break out in full volume further down. Would he not be led to conclude that where little or no water appeared upon the surface the bulk of it must find its way through under- ground channels, or percolate invisibly through the sand? Would not this supposition amount almost to certainty if it could be shown that the nature of the rock was such as to make the existence of underground channels extremely prob- able, and if in some cases they were positively known to exist ? I do not see that the inference is any less inevitable in the case before us. We know that a race has once produced a large amount of natural genius in a short time, just as we know that the river has a large volume in some places. We see, also, that the number of eminent men seems to dwindle and disappear; but we have good reason to think that social conditions can cause genius to remain hidden, just as we have good reason to think that a river may find its way through an underground channel. Must we not conclude, in the one case as in the other, that what is not seen does not cease to be, that genius is present though fame is not ? ^ Of the three theories of nature and nurture outlined in the foregoing pages, the last one is generally accepted by contemporary sociologists. Most of them agree with the eugenist that his theory of racial improvement contains ' Charles H. Cooley, " Genius, Fame, and the Comparison of Races," Annals of the American Academy of Political and Social Science, vol. ix, pp. 317-358. ip] THEORIES OP NATURE AND NURTURE ig a valuable element of truth. Likewise they agree with the euthenist when he says that it is possible vastly to im- prove a people by ameliorating its environment. Select- ing elements from the theory of each party, they hold that both nature and nurture are important, that neither alone can adequately explain the appearance of genius. It is this theory which the facts presented in Chapter III of this study seem to support, and which will be maintained in the discussion which is to follow. CHAPTER II Method of Investigation In beginning the present investigation, the first step necessary was to frame such a conception of men of letters as to permit somewhat detailed study of the various influ- ences of nature and nurture. By classifying men of letters in groups, formed according to the nature and importance of their work, it became possible to consider the effect of the various influences upon the members of each of these groups. By this method investigation could be made to discover whether supposedly potent influences had the same effect in all groups, or whether the contrary was the case. The supposed effect of the various influences could thus be at least partially verified or disproved. A roll of one thousand men of letters was first obtained by procedure as objective as the nature of the case per- mitted. Concerning each of these men all reasonably ac- cessible facts were collected which it was believed would throw light on the influence of nature and nurture in their lives.^ These facts were classified under appropriate heads and then tabulated in such a way as to show the effect of each influence upon the various classes of men. The tables are presented in Chapter III. Their significance is there, discussed, and the way in which the facts presented impinge on the three theories of nature and nurture is indicated. The final chapter presents a summary of the facts thus or- ganized. * All the data on which the study was based are given in Appendix B. 20 [20 21 ] METHOD OF INVESTIGATION 2 1 At the outset of the study the term men of letters had to be clearly defined. All definitions of the term must be more or less arbitrary. It was found, however, that Professor Alfred Odin, who had made a similar study of French lit- erary people, had framed a reasonably adequate definition of the expression.^ The fact that his definition seemed to have proved satisfactory for the purposes of his investiga- tion was also a strong argument for employing the same definition in the present study. Moreover, adoption of Odin's definition would make possible a fair and accurate comparison of his conclusions with those to be obtained in this study. For these reasons, it was decided to adopt Odin's definition. It is as follows : By men of letters we mean authors whose writings are of general interest, and all those, relatively few in number, who, without having written themselves, have none the less con- tributed directly and in an appreciable degree to the develop- ment of literature.^ Odin classified all literati under twelve heads, as follows : 1. pat. (patrons). This group includes patrons, founders and directors of schools, theatres, societies and literary salons, bibliophiles; in a word, all those who, without entering one of the three following groups, have helped in the development of literature by other methods than writing. 2. lib. (librarians). Librarians, printers, calligraphers and all those who have aided in similar fashion in the dissemination of literary works. 3. act. (actors). Dramatic artists of all kinds, including ' A. Odin, Genese des Grands Hommes, Gens de Lettres Frangais Modernes (Paris, 1895). The present study is in many ways modeled after that of Odin, and generous acknowledgment must be made of extensive use of his method of procedure. » Ihid., p. 310- 22 AMERICAN MEN OF LETTERS [22 those singers who have especially distinguished themselves by their acting. 4. or. (orators). 5. pub. (publicists). Authors of polemic or propagandic writings. 6. narr. (narrators). All those who, without marked polemic, artistic or scientific bias, relate facts or describe objects which they have seen close at hand; that is to say, most memoirists, chroniclers, authors of letters or descriptions of voyages, as well as many historians, geographers, econom- ists, etc. 7. erud. (erudite). Authors of scholarly researches based on literary documents, biographers, most historians and philo- logists, a part of the theologians, jurisconsults, etc., as well as authors of translations themselves destined especially for the erudite. 8. pop. (popularizers). All authors who serve as inter- mediaries between specialists and the general public, that is to say, in addition to popularizers in the narrow sense, authors of translations, school manuals, and, in general, of any work of instruction or popular edification. 9. spec, (speculative). Those whose writings possess pri- marily an abstract character ; philosophers in the narrow sense, many moralists, estheticians, educators, sociologists, theo- logians, jurisconsults, etc. 10. pr. (prose writers). All those who write in prose with the chief purpose of entertaining the reader, or to obtain certain artistic effects, such as novelists, feuilletonists, letter writers d la Balzac, a large part of the critics, as well as most of those who are called simply writers or literary people. 11. p. (poets). 12. dram, (dramatists).^ This classification developed a general conception of men of letteis. For the purposes of the present study it next 1 A. Odin, op. cit., pp. 356 et seq. 23] METHOD OF INVESTIGATION 23 became necessary to adopt a definition and develop a gen- eral conception applicable to American conditions. The complete definition finally adopted was : American men of letters are men of letters, within the meaning of Odin's terms, both men and women, born and brought up within the present borders of continental United States and Can- ada, in homes and schools where English was spoken, who did their work in the English language.^ This defini- tion was still somewhat arbitrary, but a more liberal one would have been subject to the criticism of admitting to the roll literati who were not born and brought up in an essen- tially American environment, a fatal defect in a study of American authors.^ There have been many thousand American men of letters as defined above. Obviously only a portion of them could be studied. The most important were naturally to be pre- ferred, for data concerning them were found to be much more abundant than in the case of minor literati. The compilation of a roll of their names presented a problem 1 To avoid monotony the terms men of letters, literati, literary per- sons, authors, and writers, are hereafter used as synonyms. ' A litterateur might of course be foreign born and yet be essen- tially American, because of having lived in an American environment from infancy. Desirable as it would have been to include such literati in the study, there were counter considerations which made the attempt seem inadvisable. It would have been necessary to decide at what age a person must come to this country in order to be brought up in an American environment. No age could have been chosen which would not be arbitrary. On the other hand, it would have been impossible to decide in the case of each foreign-born litterateur whether he was brought up in an essentially American environment. The remedy for the exclusion of the foreign-born would therefore have been worse than the evil to be cured. The number of writers thus excluded is so small as to be negli- gible. The reader will probably miss only the names of Audubon, Hamilton and Parton. More recent names were automatically ex- cluded by the fact that no authors born after 1850 have been included in this study. 24 AMERICAN MEN OF LETTERS [24 of considerable difficulty. In the interests of equity all writers of the same degree of importance had to be included in the list, all others had to be rejected. No ready-made roll satisfied this condition. Authors of biographical diction- aries do not agree in their lists. They appear to include and reject names of minor importance in the most arbitrary manner. Many mediocre writers are included by some compilers and excluded by others, while various men of obvious merit are by some compilers omitted entirely. This failing in ready-made lists was found to be particu- larly serious, since minor authors who are treated in such an arbitrary manner constitute the vast majority of all men of letters. To avoid this source of weakness, inherent in any ready-made roll, there was but one method of proce- dure ; the investigator had to make a list for himself, using a method as scientific and as little arbitrary as was pos- sible.^ In the preparation of a scientific list certain general con- ditions had to be satisfied. 1. The data had to be collected according to an indis- putably objective method, quite independent of the person- ality of the investigator. 2. The relative number of facts collected had to be large enough to be representative. 3. The absolute number had to be large enough to per- mit significant statistical work. 4. The subject under investigation had to be fairly familiar to the investigator. 5. As far as possible, the investigation had to be based on well-known men, so that the material used could he veri- fied by any one who might desire to do so.' 1 A. Odin, op. cit, pp. 358 et seq. ^ Ibid., pp. 291 et seq. 25] METHOD OF INVESTIGATION 25 Only the first of these conditions presented a real prob- lem. In its solution the method used by Odin was followed exactly. A statement of this method is therefore necessary at this point. Says Odin : The importance of a book necessarily corresponds ... to the success of the work. Consequently we must include in our list all men of letters whose success with the public is beyond question, and who are assured of not falling into oblivion by this very success. The only question is to know what is the most authentic criterion of success. We possess a criterion for men of letters which is relatively easy to ascertain, and whose value cannot be contested. It is simply the diffusion of their works. ^ This diffusion has two aspects, that of time and that of degree. Some authors enjoy very great popularity for a short period, but soon sink into comparative obscurity. Others enjoy a more modest but lasting popularity. Public esteem may be greater or less in degree. In time it may be more or less enduring. Whatever its extent in either re- spect, the facts are always readily determinable. Hence the appreciation of the relative importance of men of letters usually presents little difficulty. As a result of the method employed the slightest good faith on the part of a student suffices for the attainment of a high degree of objectivity. An investigator almost never has to estimate the success of a work. If his sources are at all abundant, as they have to be in a study of this kind, they answer the question them- selves.^ Odin states in detail the exact standards by which he measured the importance of men of letters as follows : 1 A. Odin, op. cit., p. 362. ' Ibid., p. 363. 26 AMERICAN MEN OF LETTERS [26 The most significant and at the same time the most explicit are; the number of editions and reprints; the number and success of translations, allowance being made as far as pos- sible for the personality of the translator ; finally, the number of imitations, adaptations, plagiarisms, etc. These are the most certain tests of the success which a work may have had.^ Odin also mentions other less important supplementary- tests. For example: Works which perhaps do not appear important in themselves, but which have caused keen argument, or which have become known abroad, cannot be entirely insignificant. Sometimes the very fecundity of certain authors is a proof of their success. This is the case, for instance, when a poor author writes to gain his liveUhood.^ In addition there are many very subsidiary criteria which Odin used only rarely, such as the frequency of mention or quotation, or the eulogies of well-known critics, supported by reasons. Even when such objective criteria are used, it is obvious that there must often be resort to personal judgment. An example, taken from Odin, of circumstances necessitating personal judgment will sufficiently illustrate this point. Reprints, for instance, are far from always signifying the same thing. They are quite frequently due to fortuitous causes, absolutely independent of the merit of the work and of the interest which it arouses in the public. Now it is a descendant of the author who re-edits his works through filial reverence, • • . now we see some work, disdained by con- temporaries and unknown to posterity, suddenly acquire im- portance in the eyes of certain specialists, for a reason abso- lutely foreign to its literary value. The same is true of the 1 A. Odin, op. cit., p. 364. ^ Ibid., p. 365. 27] METHOD OF INVESTIGATION 27 number of editions, which has only very relative significance. A single new edition of a large and costly work may sometimes signify as much as many editions in other cases. Fortunately it is almost always easy to determine the true significance of the different tests, for the simple reason that they serve as checks to each other. ^ Odin's foregoing criteria apply only in part to the first four categories of men of letters. With respect to patrons, librarians, actors and orators he used other tests. The speeches of orators are indeed frequently printed; but it is well known that the reception which the public reserves for the printed address does not always correspond to the popularity of the orator. . . . For the other three categories even this test is lacking. I have therefore been obliged in all these cases to restrict myself to the tests which were only sub- sidiary for the other groups. As for orators and actors, I have considered first of all the impression which they have produced on contemporaries, as it has been reported by wit- nesses worthy of trust. Here I hardly risk deceiving myself, for it is at least as easy to judge impartially of the success attained by an orator or actor as to appreciate exactly the popularity which a written work has enjoyed. I have had more difficulty in deciding which of the " patrons " and " librarians " had a right to appear on the list. For these two groups information was not always as abundant and explicit as could be desired. Thus it may be that I have erroneously omitted more than one person who was really important. Nevertheless, I have reason to suppose that the number of these omissions cannot be considerable." Following Odin's method as closely as possible, the in- vestigator began his compilation of a list of American men of letters. Five encyclopedias of biography and literature were carefully studied. These were : 1 A. Odin, op. cit., p. 365. " Ibid., p. 366. 28 AMERICAN MEN OF LETTERS [28 A Critical Dictionary of English Literature and British and American Authors, S. Austin Allibone (Phila- delphia, 1882). Appleton's Cyclopaedia of American Biography (New York, 1887). A Supplement to Allibone's Critical Dictionary of English Literature and British and American Authors, John Foster Kirk (Philadelphia, 1891). Lamb's Biographical Dictionary of the United States (Boston, 1900). The National Cyclopaedia of American Biography (New York, 1898). These five works were selected as constituting the most recent and exhaustive compilations concerning American letters and biography. The volumes of Allibone and Kirk contained practically no biographical notices, but simply the names of authors, titles of books written, and the num- ber of editions and translations of each. These works were especially useful in determining the diffusion of a work in time and space. The other three encyclopedias were typical biographical dictionaries. Their use was essential in determining the importance of literati who did not write, as well as in estimating the popularity of authors who wrote after 1891, when Kirk's volumes appeared. It soon became apparent that the different sources were not equally reliable. The works of Appleton and Lamb seemed satisfactory in every way. Their articles were dig- nified and moderate in tone, and their statements were ap- parently always justified by the sources on which they were based. On the other hand, the National Encyclopedia often seemed extravagant in its praise of an author, devoting more space to writers of doubtful merit than to other men of established reputation. Consequently it was frequently 29] METHOD OF INVESTIGATION 29 deemed necessary to discount its assertions to some extent. The volumes by Allibone and Kirk seemed quite reliable as regards statements of the nature and amount of work done by an author. Librarians consider them to be standard works. In all cases of doubt as to whether an author was suffi- ciently important to be admitted to the list, additional works were consulted. The most important of these were : A Dictionary of American Authors, Oscar Fay Adams (Boston and New York, 1905). An American Anthology, Edmund Clarence Stedman (Boston and New York, 1900). Chamber's Cyclopaedia of English Literature (Phila- delphia, 1904). The Cyclopaedia of American Literature, Evert A. and George L. Duyckinck (Philadelphia, 1881). Who's Who in America (Chicago, various dates). In spite of the deficiencies of these sources, it is believed that collectively they furnished an adequate criterion of the importance of American men of letters. The name of each litterateur mentioned in the foregoing volumes, who seemed to have any claim to a place on the roll, was put on a numbered card. On the card was also written the name of the class or classes of literary activity in which the author appeared to have made a significant record. When a man of letters had distinguished himself in several fields he was noted as belonging in all of them, though later in the statistical summaries he was counted only in that one in which he had achieved the greatest dis- tinction. On the card was also recorded a list of important books written, including a statement of reprints, new edi- tions, and translations, as well as any other facts which seemed to warrant the inclusion of the author in the ranks of American literati. 30 AMERICAN MEN OF LETTERS [30 In the consideration of the names of candidates for the roll, the alphabetical order was followed. This method not only expedited the work, but served to prevent the intru- sion of any possible bias in favor of a particular time or place. Dates and places of birth were not noted till the final list had been completed. It seemed best to admit to the roll only persons bom prior to the year 185 1. There were two reasons for this restric- tion. The biographies of writers born after 1850 were found to be few and incomplete. Moreover, it seemed un- fair to pass judgment on an author before it was certain that he had achieved his maximum literary reputation. For most of the younger writers such a decision could not be made. On the other hand, it appeared that few writers who had attained the age of sixty-four ^ would be likely to alter their status in the literary world to any important degree. Hence it seemed quite safe to consider as candi- dates for the list all authors born before 1851. None of the chief sources used were published after 1900. Inasmuch, however, as no eligible author who was little known before 1900 subsequently sprang into promi- nence, it seems probable that these sources included the names of all persons who were sufficiently important at the time of the compilation of the roll (1914) to deserve a place on the roll of the thousand foremost American men of letters born prior to 185 1. The first preliminary survey gave a total of nearly thir- teen hundred names. This entire roll was carefully scruti- nized a second time, and the sources again consulted. It was then apparent that some authors had been included • The list was compiled in 1914. Inasmuch as the latest reference works were consulted, including Who's Who in America for 1914-1915, it seems probable that the true rank of each living author was de- termined with adequate accuracy. 31 ] METHOD OF INVESTIGATION 31 previously who did not fully measure up to the standard required for admission to the list. The names of authors whose importance seemed doubtful were carefully indi- cated. A third revision, not less thorough than the others, determined with a considerable degree of certainty what names were to be included in the final roll. Attention naturally centered on names of doubtful importance. To avoid all possibility of bias, however, every name, whether doubtful or not, was carefully considered a fourth time. The final list was found to contain one thousand and six names. The investigator had made no conscious attempt to ob- tain exactly one thousand names. He had no idea whether he would have nine hundred or eleven hundred names in the final list. Inasmuch, however, as the number obtained was so near one thousand, it seemed desirable to reduce the list to that number to facilitate calculations. The names of five authors of children's stories were finally selected for elimination, because judged to be the least important on the roll. They were found only in Kirk's work, the least im- portant of the sources used. It would have been useless to retain them, for no biographical facts about the authors were available. The other name eliminated was that of a man who never put pen to paper as an author, but who dic- tated an account of King Philip's war, a narrative valued solely for its historical significance. Since this man was the only person on the list who did that kind of literary work, it seemed reasonable that he should be the sixth per- son to be dropped from the roll, particularly as there were apparently no other authors who could be considered less important. The facts noted during the compilation of the list of names facilitated division of the literati, during these sur- veys, into two classes, those of major and those of minor 32 AMERICAN MEN OF LETTERS [32 importance. This division made it possible to determine whether the more prominent writers were born in circum- stances different in kind or in degree from those in which mediocre authors appeared.^ This division was made in accordance with the degree of success which men of letters had achieved, measured, as before, by the diffusion of their works in time and space.^ The more prominent authors formed the smaller of the two groups. For convenience its members are hereafter designated as men of talent. The minor literati who con- stituted the other group are called men of merit.' Men of talent were classified as follows : ( i ) authors whose works had been translated into foreign languages ; * (2) those writers whose works were very widely read in other English-speaking countries during their lifetime; (3) writers whose works were read extensively after their 'death ;^ and (4) those literary patrons, Hbrarians, actors, and orators whose reputation endured after their decease. After the final roll had been determined, the desired facts 1 Cf. Odin, op. cit., pp. 374 et seq. 2 Cj. supra, p. 25. * In making this classification foreign works were consulted, espec- ially Meyers Grosses Conversations-Lexicon (Leipzig and Vienna, 1906), and La Grand Encyclopedic (Paris, no date). This was done in order to discover to what extent the works of the more important American authors were read and esteemed abroad. * Exception was made of authors of works of missionary or tem- perance propaganda, writers of text-books on non-literary subjects, explorers who owed their success as authors chiefly to their sub- jects, authors of works of special interest to a foreign people because dealing with some phase of their national life or history, and authors whose residence abroad apparently caused the translation of their works. "^ From this category were omitted those authors whose works were valued chiefly as historical sources, and authors of posthumous works which enjoyed only ordinary success. 33] METHOD OF INVESTIGATION 33 bearing upon the nature and nurture of each person on the list were collected. A questionnaire sent to living authors and to the immediate relatives of others met with an unex- pectedly cordial reception. One hundred and seventy-five schedules were returned, more than seventy per cent of those sent out. Biographies furnished abundant informa- tion in regard to perhaps fifteen per cent of the thousand literati. Facts about the others were gathered from ency- clopedias, magazine articles, and various scattered sources. Many facts could not be discovered, but those collected were sufficiently numerous to be representative, and to serve as the basis of significant statistical calculations.^ When all available facts concerning each author had been collected and recorded on the individual cards, the process of analyzing the data was begun. It was then a simple matter to isolate for consideration any recorded fact, by means of sorting the cards. The results of this analysis and interpretation constitute the subject-matter of the next chapter. • In the absence of reason for believing that the facts collected are biased, there is no statistical error in proceeding to draw inferences from samples chosen by any unprejudiced method. Cf. A. L. Bowley, An Elementary Manual of Statistics (London, 1910). CHAPTER III Analysis and Interpretation of Data This study, as has been stated in the preface, was origi- nally undertaken with the intention of making an investiga- tion exactly parallel to that of Odin. The plan was to dis- cover, with respect to American men of letters, whether Odin's contention that nature is much more important than nurture was sustained. It was the belief and hope of the author that the data collected would lend themselves to such interpretation, and thereby be in harmony with Professor Ward's argument for the preponderant influence of en- vironment over heredity, as presented in his interesting work, Applied Sociology} As the work progressed, how- ever, and as the tables on heredity were prepared, it became evident that, in order to reveal the whole truth, methods of manipulating the data which were not used by Odin would have to be employed. Hence it became necessary to scruti- nize from as many angles as possible the data which "had been collected, instead of simply following the method of analysis which Odin had used.^ Tables were therefore pre- pared to present the data from many points of view. Some- times a table was made simply to present facts in a con- venient summary. More frequently, however, one was pre- 1 Lester F. Ward, Applied Sociology (Boston, 1906). ' As a result of this modification of the plan of study, a few tables are introduced in the following pages which are quite unlike any pre- sented by Odin. In the main, however, his method of analysis was closely followed. 34 [34 35] ANALYSIS AND INTERPRETATION OF DATA 35 pared in the hope that it would throw some light on the particular phase of the subject under investigation. The first table prepared, showing the absolute number of literati bom in each decade, is an instance of the first type. This summary was naturally followed by a study of the relative number of men of letters born in each decade. As it was found that the number fluctuated, a search was made for an explanation of the variation. Again, it appeared that certain families and environments had produced unusual numbers of authors. This discovery led to the preparation of many additional tables, which, it was thought, might possibly make clear the reason for the facts observed. Some of these tables were later discarded because they appeared to have no particular significance. The others are presented and discussed in the following pages. Those presented were not originally prepared in the order in which they now stand. They are given in this sequence simply because this seems to be the manner of presentation best adapted to bring out the conclusions which resulted from studying the data in as unbiased a manner as possible. The conclusion to which a consideration of the tables seemed to lead is introduced here in the form of a thesis. It is not an arbitrary dogma to be defended at any cost, nor is it the theory which the investigator expected to find the data sustaining when he began his work. Rather is it a gradually developed conclusion which he felt obliged to accept as the result of his study. The arbitrary form of presentation is therefore used simply for the sake of defi- niteness and brevity. This thesis is as follows. In all ranks of American society there have been found men and women of literary ability. Much of this ability has been found in members of the same families, but it has been the monopoly of neither a select group of families nor of a particular nationality strain. This latent ability has 36 AMERICAN MEN OP LETTERS [36 been brought to light by favorable environmental influences, of which there are two distinct kinds. One kind may be called education, or training, and includes those influences of home and school which are particularly potent during childhood and youth. The other kind includes all the re- maining elements of environment, especially the ideals and customs of the group in the midst of which one lives. Pos- session of even the best advantages at home and in school has made possible the development of great authors only when supplemented by this second factor of environment. In short, men of letters/nave appeared chiefly when the society of their time has appreciated and demanded litera- ture. Without such incentive to write, persons with natural literary ability and adequate training have tended to turn their efforts in other directions. This thesis can be put in the form of a simile, nature being likened to seed and nurture to ground. A combina- tion of either good ground and poor seed or poor ground and good seed will produce a better crop than when poor seed is sown on poor ground. No good crop is ever pro- duced, however, without the use of both good seed and good ground. In like manner gifted children who lack oppor- tunity, and dull children who possess every opportunity, achieve far more than dull children who lack favorable con- ditions of environment. Genius, however, is usually pro- duced only by a favorable combination of innate ability and the two factors of environment mentioned in the preceding paragraph. This thesis is of course only one form of the statement , that both nature and nurture are of importance in the de- velopment of genius. It is in harmony with the opinions of those sociologists of whom Professor Cooley was quoted as representative in Chapter I. Now that the thesis has been stated and the method of 37] ANALYSIS AND INTERPRETATION OF DATA x] investigation has been made plain, the data on which the study is based can be presented and discussed. The appar- ent influence of various environmental conditions which seem to have affected the authors studied will first be con- sidered. Reason will next be given for a belief that cer- tain phenomena, explained in this study in terms of nurture, cannot possibly be interpreted chiefly in terms of natural ability, as Galton supposed. To this extent it will be shown that the results of the study harmonize with the opinions of Ward. Finally, however, evidence will be submitted for a belief that nurture alone cannot explain the development of American literati, as Ward believed, and an argument will be made so far in support of Galton's contention as to hold that original nature is at least an important factor in the development of genius. Thus it will be shown that appar- ently the theories of both Galton and Ward are partially right and partially wrong, and that a combination of the two theories, as in the thesis stated above, seems best to accord with the facts as observed. There are nine important environmental conditions which will be considered. They are as follows: (i) social en- vironment, by which is meant the ideals and customs of a group at any given time and place, (2) geographic environ- ment, (3) local environment, (4) education, (5) economic condition of parents, (6) occupation of father, (7) occu- pation of the literati themselves, (8) early religious train- ing, and (9) birth-rank in the family of brothers and sisters. Before these forces of nurture can be considered, how- ever, it is necessary to present a few facts about the history of American letters to serve as a background for the discus- sion which is to follow. These facts are presented in Tables I and II. It appears in Table I that the number of literati born in- creased very rapidly from the time of the American Revo- 38 AMERICAN MEN OF LETTERS [38 TABLE I Distribution of iooo American Literati Born Prior to 185 i, by Period of Birth' 1 1 1 1 1 1 g Before 1 1 0000 CO »-t 1 N ro, ■* v^ VO 1 t^ 00 ON 11,1 T •4 i-i M CO ^ m Period , , , -li. 1 1 i^ 1 1 1 1 1 1 1 701 HH M fO •* to \0 t^ 00 C^ h., C4 ro ■+ 1 1 t>. 1-1 M 1 M !>. 00 00 00 CO Ii4 00 Number. .. 1 6 4 3 7 8 8 18.14 34 49 103 122 178 140 169 137 lution till about 1820. After that time the absolute num- ber declined, though not at a uniform rate. The full sig- nificance of the change is brought out in Table II, which shows the relative number of men of letters born in each decade. From Table II it appears that before 1771 there were born on the average in each decade ten literary people per million of white population. This number gradually increased until, during the years 1 791-1800, there were produced twenty-three authors per million. This birth- rate remained practically constant during the two succeed- ing decades. Then there was an abrupt change. In the period 1821-30 the relative number of men of letters bom was less than sixty per cent of what it had been in the previous decade. This decline continued steadily, till in the last decade recorded the relative number of literati born was less than thirty per cent of the number born in the period of maximum fecundity.' ' There are two colored literati on the roll. They are included in all studies except those summarized in Tables 11, VIII, and XIV, where some ratio of white literati to white population is considered. The fact of this exclusion is in each case plainly indicated in discussion of the table. ' It must be born in mind that, in all probability, the period of maximum literary productivity of an author is normally between forty and sixty years after his birth. Thus the literary birth-rate indicates roughly the amount of literary activity a half-century later. 39] ANALYSIS AND INTERPRETATION OF DATA TABLE II 39 Absolute and Relative Numbers of American Literati of White Race, Born within the Present Territorial Limits of Conti- nental United States prior to 185 i. Classified by Period of Birth' Period of birth Before 1771 1771-80 1781 90 . .. 1791-1800.. l8oi-JO. ... 1811-20 1821-30.... 1831-40 ... 1841-50 . . . Total. Absolute number White population of Number of literati the period in per million of white thousands' population 68 6,735' 10 34 2,249 IS 49 3.170 '5 lOI 4.30s 23 118 S,86i 20 176 7,866 22 138 10,522 13 163 14,191 II 133 19.37s 7 980 74,274 13 1 The eighteen Canadian writers can not be considered in this table, as there was no adequate census of Canada before 1850 on which to base comparisons. The two men of African descent are also omitted. ' For the period since 1790 the figures are taken from the decennial census of the United States. Statistical Abstract of the United States, 1909. Table 20. For the period prior to 1790 the figures are based on estimates in or derived from A Century of Population Growth in the United States. Bureau of Census (Washington, 1909). The very small population figures for mountain and pacific states are omitted to make this table comparable with Table XII. In estimating the white population for the colonial period it was assumed that, during the entire period prior to the first census, the colored population bore the same relation to the white population in each colony that it did in 1790. This assumption did not entirely accord with the facts, but data on which to base more accurate estimates were not available. As a result of this assumption the period before 1771 was credited with a slightly larger number of literati per million of population than it really deserved. It is extremely unlikely, however, that the error was large enough to change the index figure given for the period even as much as from ten to nine. ' It will be noted that, with the exception of the period before 1771, the number of literati born in each decade was compared with the 40 AMERICAN MEN OF LETTERS [40 The statistics in Tables I and II thus show that the liter- ary activity of the American people has been far from uni- form in amount. This fact requires explanation. It is apparently best explained, in accordance with the thesis maintained in this study, in terms of the social environ- ment, the first of the nine environmental conditions which are to be considered. Authors of the first rank disappeared after the 'Civil War because their work seems to have been no longer appreciated.^ In other words, the social environ- ment had become relatively unfavorable to literary activity. The reason for this change seems to be as follows. During the Reconstruction period the temper of the American peo- ple was profoundly altered. It is true that before the War the spirit of commercialism was strong, but many people still had leisure which they devoted to serious reading. Later, when ail values seemed to be expressed in terms of money, the nation had less time to devote to a seemingly impractical subject like literature. Professor Cooley ad- mirably summarizes the argument for this theory when he says: The real cause of literary and artistic weakness (in so far as it white population of the United States at the end of that decade. Each white person in the United States was counted in the population of each decade at the end of which he was alive. To obtain cotnparable results the same procedure was adopted for the period prior to 1771. The population for each decade was estimated, and the number of literati born in the period was recorded. Since these figures were too small to be significant, it seemed best to combine the estimates for the entire colonial period. The number of literati born in the period before 1771 was of course found by adding the numbers born in each decade of the period. Likewise the estimated population for each decade was summed, to give a comparable population figure for the period. Only thus could significant figures be obtained. • The marked fall in the birth rate of literati took place perhaps fifty years before the corresponding decline in literature. It began after 1820, and continued till the close of the period studied. 41 ] ANALYSIS AND INTERPRETATION OF DATA 41 exists) I take to be chiefly the spiritual disorganization incident to a time of rather sudden transition. . . . No matter how gifted an individual may be, he is in no way apart from his time, but has to take that and make the best of it he can ; the man of genius is in one point of view only a twig upon which a mature tendency bears its perfect fruit. . . . Any ripe de- velopment of productive power in literary or other art implies not merely capable individuals but the perfection of a social group, whose traditions and spirit the individual absorbs, and which floats him up to a point whence he can reach unique achievement. The unity of this group or type is spiritual, not necessarily local or temporal, and so may be difficult to trace, but its reality is as sure as the principle that man is a social being and cannot think sanely and steadfastly except in some sort of sympathy with his fellows. There must be others whom we can conceive as sharing, corroborating and enhancing our ideals, and to no one is such association more necessary than to the man of genius. ... no doubt such questions afford ground for infinite debate, but the under- lying principle that the thought of every man is one with that of a group, visible or invisible, is sure, I think, to prove sound; and if so it is indispensable that a great capacity should find access to a group whose ideals and standards are of a sort to make the most of it.^ Among other significant facts in the history of American letters which seem to be explained in the light of this proposition better than by either the theory of Galton or of Ward there are seven which are particularly worthy of notice. The first appears in Table III. From this table it is evident that from colonial times to the period at which this study ended, there was a fairly steady decline in the pro- portion of literati of superior achievement, called men of talent. It seems probable that the same influences which 1 Cooley, Social Organisation, pp. 162 et seq. 42 AMERICAN MEN OF LETTERS [42 caused fewer potential men of letters to devote themselves to authorship had an especially strong effect on men of ex- ceptional ability. It is apparently reasonable to assume that men of genius are more dependent upon their environment than are others, for, as Cooley remarks : " being thinner- skinned, they are more suggestible, more perturbable, and TABLE III American Literati Classified by Sex and by Rank, by Period of Birth Rank ' Before 1771 & 1 1 7 00 78 32 II I no 12 to 89 33 8 1 00 i-i 127 28 12 II '55 23 "3 139 39 1 00 91 26 20 3 117 23 16 III 29 °r <» 97 34 31 7 131 38 22 128 4« ■f 86 22 26 3 io8 29 21 112 25 e2 Men of merit ......•••• 43 22 3 24 9 I 30 14 3 2 6s 31 1 96 7 7 70 33 641 .-•I 8 Men of talent Women of merit Women of talent • ■■■■■■ 112 29 Total Men 65 3 33 I 44 S 10 16 859 141 Total Women Women, per cent Total merit 4 46 22 3 25 9 14.1 753 247 Total talent Talent, per cent 32 26 S3 32 27 22 21 24 18 24.7 peculiarly in need of the right sort of surroundings to keep their delicate machinery in fruitful action ".' Presumably the best potential American authors, those endowed with the finest sensibilities, were the persons whose sensitive minds were most ready to give up the pursuit of letters when conditions became unfavorable. Thus the fact that the relative number of literati of talent began to decline 1 For method of assignment of an individual to the rank, talent or merit, see p. 31. 'Cooley, op. cit., p. 165. 43] ANALYSIS AND INTERPRETATION OF DATA 43 while the absolute number of authors was still increasing, is probably to be explained on the ground that the men of superior ability were the first to sense the baneful influence of approaching philistinism. The second of the seven noteworthy facts which seem best explained as due to changes in the social environment is presented in Table IV. This table shows that the decline TABLE IV American Literati Classified by Field of Chief Activity and Period OF Birth.' Chief Field of Activity ' Patrons Librarians . ■ ■ Actors Orators Publicists Narrators .... Erudite Popularizers ■ Speculative . . Prose writers Poets Dramatists . ■ ■ Total . S Before 1771 3 10 6 13 17 8 2 7 I 68 34 49 S I 2 3 2 9 13 38 8 10 II I 103 2 4 S 8 10 IS 40 6 12 17 3 122 5 IS 16 30 38 10 21 30 3 2 7 2 I II 12 22 32 6 27 18 3 2 13 I 10 S 32 35 I 38 25 4 178 !l40 J169 Total 2 6 2 5 , 7 1 IS 26 7 48 16 3 137 23 33 24 71 70 IS7 249 50 166 132 15 1000 of American letters was not manifest in all kinds of litera- ture. The number of authors diminished in nine of the twelve fields of literary activity under consideration, but the number of actors, dramatists, and prose writers did not I In the compilatioti of this table each author was counted only in that field of activity in which he seemed to have attained the greatest distinction. 2 The exact character of these classes is defined on pages 21-22. 44 AMERICAN MEN OF LETTERS [44 decline during the last few decades studied/ The theory- advanced above readily explains this apparent exception to the general tendency. Activity increased in the three kinds of work which were in harmony with the spirit of the time. This activity furnished what the people demanded. The environment being favorable, the number of literati in these three fields naturally tended to increase. The third noteworthy fact, discovered from data not here presented, is that in these three fields in which activity was increasing, apparently because of greater popular in- terest, there was not a growing proportion of literati of talent compared with those of merit. It might seem that, according to the theory that when literature is in popular favor conditions stimulate the production of literary genius, an increase in the number of men of talent in these fields should have been expected. In reality, however, such an increase would not harmonize with that theory, while the decline in the ranks of men of talent observed is quite in accord with it. This paradox is explained as follows. In the first place, it must be remembered that, at the time when the authors born in the latter decades studied were writing, popular taste in fiction and the drama was not at all what it had been several decades previously. Even the attitude of the public toward the players had changed. People did not then have, as formerly, enduring interest in an actor. The desire of the public was for the recent. " Popular " books were lauded, and it was not fashionable to read books ' These facts were further verified by the results of another analysis in which the method of procedure served as a check to the one used in compiling Table IV. In this case each litterateur was counted onca for every line of activity in which he had achieved distinction. Re- sults differed so little from those noted in Table IV that it seemed needless duplication to print even the summaries. It is quite evident that literary activity declined at approximately the same rate as did the number of literati. 45] ANALYSIS AND INTERPRETATION OF DATA ^t which had been tested by time. Popular taste was also provincial, rather than catholic. It did not care for litera- ture which was universal in its appeal, but preferred that which dealt with matters of local and transient interest. This provincialism may well account for the small number of writers of talent in the later decades of the study. Authors of merit merely produced what the people would read. It was a poor literature which could not command a foreign audience, or even hold the attention of Americans for any length of time. In other words, popular taste had by its very nature made it increasingly difficult for a litterateur to win recognition as a man of talent, though comparatively easy for a man to attain the rank of a man of merit. The fourth significant fact to be noted in connection with the theory under consideration is given in Table V, which contains an analysis of the fields of activity of liter- ary women. The table shows that women did considerable TABLE V American Literary Women Classified by Field of Chief Activity and Period of Birth Chief Field of Activity Before 1771 r 1 8 M M 1 00 1 00 M r 00 T HI oT 1^ Total Futroiis • • • • • • • ■ Libr&risLns <■■•■• •■•< Actofs .•••.. ■■>■<■ •■ I I 2 3 1 3 20 2 7 f Orators ............. Publicists .•....• I 2 2 11 5 5 4 26 N&rratnrs ■■■>....... I I 4 I 8 7 I 2 6 17 9 Erudite I 2 PoDiiI^rizers •■•■■ ■■■■ I 2 2 3 J Prose writers - 2 I 3 4 10 6 61 Poets 30 Oramfltists .**.■■ •■>■ Total 3 I S 7 12 23 23 38 29 141 46 AMERICAN MEN OF LETTERS [46 work in popularizing, in poetry and in prose writing. These were the fields of literary activity which the public seemed to consider most appropriate for women. It is noteworthy that, during the last few decades considered, the number of literary women increased only in the fields of acting and prose writing, two of the three fields in which the number of men also increased. The fifth of the series of facts best understood in the light of the theory of the influence of the social environment is given in Table VI. The authors are here classified, by period TABLE VI American Literati Classified as of One, Two, or Three or More Fields of Activity, by Period of Birth.' Of one field Of two fields Of three or more fields Period of birth a Per cent of all literati born in the period a B 9 3 12 20 24 36 21 43 17 1S5 =3 E-o 00 M- ^ "C HI 13 9 24 19 20 20 15 25 12 i8.s i ^ JO Per cent of all literati born in the period Total abso- lute num- bers Before 1771 .... 1771-80 1781-90 1791-1800 l8oi-io 1811-20 1821-30 1831-40 1841-50...- li 34 79 90 "3' 107 119 117 758 78 82 69 77 74 74 76 70 85 75-8 6 3 3 4 8 II 12 7 3 9 9 6 4 I 9 4 2 68 34 49 103 122 178 140 169 '37 Total 57 5-7 1000 of birth, as of one, two, or three or more fields of activity. It appears that in the period 1841-50 there were relatively 1 When the percentage figures in this table are added on horizontal lines the totals will not in every case equal 100 because of cumulative error. The inaccuracy, however, is slight. 47] ANALYSIS AND INTERPRETATION OF DATA 47 far fewer literati of two or more fields than at any previous time in American history. This decline may have been either a temporary fluctuation or a real tendency due to the same influence which caused the decline in the number of men of letters in general. If it was the manifestation of a real tendency, it can perhaps be explained by the supposition that the more versatile potential literati found it easy to adapt themselves to unfavorable conditions, and were there- fore the first to give up the pursuit of letters. The sixth fact of this series is brought out in Table VII, which shows the field of chief activity of authors born in the various groups of states.^ These figures speak for TABLE VII American Literati Classified by Field of Chief Activity and Region OF Birth Field of chief activity ' Patrons Librarians . . Actors Orators Publicists Narrators — Erudite Popularizers . Speculative . . Prose writers Poets Dramatists . . Total . i8 cm 6 14 10 '3 37 25 8S 136 30 69 61 I 487 54 40 10 316 u u (u C a S a - « •«« 6- 'S< y 5866 3548 1908 1778 y 1037 ' <'>5 2923 2614 610 278 5236 5 2170 b 1539 3 6.S2 8 336 9 1 104 5 * 73796 ^13 1 No figures can be given for Canada because population estimates are lacking. ^ Includes one man of African descent. ' The population base is derived by summing the figures indicating the white population for each decade from the founding of a state or 51 J ANALYSIS AND INTERPRETATION OF DATA 51 In the first of these tables (Table VIII), the authors are classified by state or province of birth, by sex and by rank. The most significant figures of this table appear under the column headed Index. This column contains for each state a figure indicating the average number of white literati, per million of white population, born in that state or colony prior to 185 1. The figure was derived by summing the figures indicating the white population for each decade from the founding of that state or colony up to 185 1. By this total, called population base, was divided the number of white literati born in the state or colony during the same period.^ For instance, the sum of the estimated and enum- erated decennial white population figures for Massachusetts, colony up to 1851. In the table the figures of population base are given in thousands (i. e., three naughts [000] are omitted in each case). In estimating the white population for the colonial period it was assumed that, during the entire period prior to the first census, the colored population bore the same relation to the white population that it did in 1790. While the assumption did not exactly accord with the facts, it seemed inadvisable to attempt to obtain greater ac- curacy. Better results would have been attained only at a labor cost out of all proportion to their value. As a result of the method used the states with a large colored population seem to have produced relatively more literati than should really be credited to them. The error, however, cannot be significant, for the total population of the colonial period was relatively small. * This total differs slightly from that of Table II because it excludes the population of several southern states which produced no literati before 1851. ^ This index (13) is based on a population total of 74,274 which in- cludes the populations of several southern and western states the popu- lations of which were enumerated in 1850 or earlier, but which had produced no literati and which, therefore, were not included in the separate categories of this table. The number of literati included in the calculations for this figure (13) was 980. Eighteen Canadian literati were omitted because of the lack of Canadian population estimates. The two men of African descent were also omitted. The three men of unknown region of birth were, however, included. ' The population figures are taken from census returns and estimates in or derived from A Century of Population Growth in the United States. 52 AMERICAN MEN OF LETTERS [52 from 1620 to 1850 inclusive, calculated in each decade to the nearest thousand, was 5,637. When by this sum 'was divided 246, the number of white literati born in that state during the same period, an index number of 44 was ob- tained. This procedure had the following justification. Plainly a figure derived by dividing the population of a state at the end of any decade, by the number of authors born in the state during that decade, would be an index of the relative productivity of literati by that state during the decade. Such figures could have been obtained, but because of the small numbers concerned their significance would have been slight. However, when the numbers of white literati bom during each decade are summed, and the figures for the white population living at the end of each decade are also summed and expressed in millions, and when this former sum is divided by the latter, there results a figure which indicates the average productivity of literati by a state in each decade, per million of white population. This index number is chiefly significant as a measure of the relative literary fecun- dity of the different states. The chief points brought out by the index numbers of this table are as follows : ( 1 ) With the exception of Vermont and Rhode Island, all the New England states ranked higher than their nearest competitor. New York. (2) Massachusetts and Connecticut stood far above the other New England states, and Massachusetts had a large lead over Connecticut. (3) The District of Columbia ranked next to Massachu- setts. (4) The relative importance of the southern states was slight. Of all the states south of Mason and Dixon's line, 53] ANALYSIS AND INTERPRETATION OF DATA 53 TABLE IX American Literati Classified by Rank by Region of Birth Region of birth 1 Canada New England Middle Atlantic... South Atlantic East South Central West South Central East North Central West North Central Unknown Total Rank Total Merit Talent 18 15 3 352 "35 487 246 70 316 80 «9 99 14 I 15 3 I 4 38 IS S3 3 2 5 2 I 3 753 247 1000 Talent per cent ' 28 22 19 28 24.7 Maryland alone had as high rank as the lowest of the New England and Middle Atlantic states. ( 5 ) The states formed from the Northwest Territory on the whole ranked with the southern states, distinctly lower than those of the east." ' New England — Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut. Middle Atlantic — New York, New Jersey, Pennsylvania. South Atlantic— Delaware, Maryland, District of Columbia, Virginia, West Virginia, North Carolina, South Carolina, Georgia, Florida. East South Central — Kentucky, Tennessee, Alabama, Mississippi. West South Central — Arkansas, Louisiana, Texas, Oklahoma. East North Central— Ohio, Indiana, Illinois, Michigan, Wisconsin. West North Central— Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, Kansas. 2 Not given when very few persons are concerned, for the ratio would be spuriously accurate and therefore misleading. ' Care must be used in drawing comparisons among states in cases where the recorded instances are too few to permit accurate statistical deductions. Other investigators have discovered striking differences in the pro- 54 AMERICAN MEN OF LETTERS [54 The second of the five tables considered under the subject of geographic environment (Table IX), shows the region of birth of literati, classified according to rank. It appears that the northern states have produced more persons of talent than have those of the south. The figures also show that both New England and the East North Central states produced unusually large proportions of literati of talent. Standing by themselves, the foregoing figures might seem inconclusive. They are borne out, however, by results pre- TABLE X American Literati Classified as of One, Two, or Three or More Fields of Activity, by Region of Birth One Field Two fields Three fields. Region of birth a 356 249 77 J A Per cent of literati of the region ' E 3 1 16 Per cent of literati of the region ' Per cent of literati of the region' Total Canada New England Middle Atlantic . . . South Atlantic... Fast ^outh Central 18 73 79 78 .... 20 18 16 36 7 3 487 316 99 •5 4 S3 5 3 West South Central 4 East North Central 37 West North Central ! 3 Other 3 70 12 2 23 4 8 75.8 185 18.5 S7 Total 758 5.7 1000 duction of prominent citizens by diflferent sections of the country. Cf. James McKeen Cattell, American Men of Science (New York, 1910) ; George R. Davies, " A Statistical Study in the Influence of Environ- ment," Quarterly Journal of the University of North Dakota, vol. iv, no. 3 ; and Scott Nearing, " The Geographic Distribution of American Genius," Popular Science Monthly, vol. 85, p. 189. ^ Not given when very few persons are concerned, for the ratio would be spuriously accurate and therefore misleading. The totals in these columns are based on the complete absolute figures. 55] ANALYSIS AND INTERPRETATION OF DATA 55 TABLE XI American Literati Classified by State of Birth and Period of Birth State of birth 8 >-* lU s 4 5 1 f o 00 1 t^ ON 1 00 8 00 7 2 6 5 5 29 I 18 18 3 1-1 1 1 8 1 5 M w 00 r .It 13 f2 1 3i I J 1' t t 2 2 37 I 10 29 2 17 4 I 6 I 4 3 I 6 3 .. "■ 6 •7 ig 54 46 28 246 '3 100 205 26 85 3 31 6 29 5 12 13 I 2 3 I 2 10 ?8 TWfliTip .......... > "! 1 New Hampshire. . . . •• I •• 2 6 *-6 I 31 3 6 22 Q s 9 49 3 15 49 s ¥ T 3 1 27 17 21 1 10 9 43 36 4; 4 19 3 II 1 2 1 1 2 2 ** Massachusetts Rhode Island Connecticut New York New Jersey Pennsylvania ■>.■•■ 6 3 I I I I 3 2 I 2 .. 3 2 2 7 4 I 5 3 2 I 2 12 I S I 2 r 18 I II 2 2 c Delaware 5 " I 2 I ^ 2].. 4' 3 -- ! 2 5 - .. "1" District of Columbia Virginia ■• ■■ [ 2 - - '3 I North Carolina South Carolina Georgia Alabama Mississippi Louisiana Arkansas Tennessee Kentucky Ohio * ' ■ ■ •• ;; I I • • I 2 • • 2 I 3 ■i 2 I 2 I 2 I I 2 2 I 2 9 S I 1 3 I ■3 I I I 2 I 12 Indiana Illinois Michigan Wisconsin Missouri Unknown Total .. •• ■■ 1 i ,. ' .. 1 .. 1 1 '■ 1 " *■ I .. i . . 1 I 1 1 1 1 1 1 I I 5 3 4 2 ? 1 13 4 5 3 5 3 1000 sented in the third of the tables on geographic environment (Table X). This table shows the region of birth of literati; classified as of one, two and three or more fields of activity. In this table, the New England and East North Central states again appear appreciably in the lead/ ' An unpublished classification of literati by sex, according to the group of states in which they were born showed remarkable uniformity 56 AMERICAN MEN OF LETTERS [56 In the fourth place, classification of men of letters by decade and state of birth (Table XI) shows plainly that the relative imp jrtance of the states was not constant. The full extent of the changes in the relative importance of the states in the production of men of talent is not apparent, however, till one considers the results of a further analysis which yielded the fifth and last of the tables on geographic environment, on the literary fecundity of each group of states in proportion to white population (Table XII). The dechne in the number of literati made manifest in Table II is here shown (Table XII) to be no local phenomenon. During the latter decades studied there was a marked diminution in the relative number of men of letters born in every group of states which possessed enough authors to make figures significant. Table XII also shows that the center of Ameri- can literary activity was slowly but surely shifting. In the decade 1841-50, New England was still supreme, but its lead had been appreciably reduced. The East North Cen- tral states showed the least relative decline in literary fecun- dity, a fact which may indicate that the future literary leadership of the country is to be theirs. When the history of the nation as a whole is considered, however, it seems that New England's predominance during the period studied was little short of marvelous. The group produced in pro- portion to population more than twice as many literati as did the Middle Atlantic states, and more than six times as many as did the South Atlantic group, or any of the other groups of states. in the proportion credited to each. No group of states appeared to possess conditions particularly favorable to the development of liter- ary talent in one sex, rather than in the other. 57] ANALYSIS AND INTERPRETATION OF DATA 57 TABLE XII Relative Litekati Productivity of the Several Groups of States Abbreviations : Pop., white population of the region at the end of the period, in thousands; Lit., white hterati born in the region during the period; Ratio, number of literati born per million of white population living in the region at the end of the period.' Group of States New England. Middle Atlantic. South Atlantic East South Central. West South Central (■Pop .. J Lit . . . ( Ratio . fPop.. . -^ Lit . . . ( Ratio . (Pop.. . 4 Lit . . . ( Ratio . (Pop.. . ] Lit . . . ( Ratio . ■2 M 2.55° 43 17 1.747 18 2,421 7 3 17 (■Pop ]Lit...i ( Ratio . East North Central . (Pop . -^ Lit . . ( Ratio West North Central. United States . fPop. ■^ Lit . . ( Ratio (Pop . J Lit . . (. Ratio . 6,735 68 730 20 27 638 8 13 844 6 7 37 ^ so ^4 IX 0^ r* 00 —1 1,214 M i,4';i 64 70 53 48 991 36 36 908 1,338 1,933 9 27 32 1,178 1,426 4,' 3 93 2,249 34 15 3.170 49 15 277 2 17 1.593 <3 563 2 34 I 5° 4.305 lOI 23 270 '7 S,86i 118 20 1,638 94 57 2,610 65 25 1.787 II 6 902 3! 1.934 61 32 3.484 48 14 2,117 19 9 1.304 I 87I "5 7861 1,453 2 9 3! 6 2,212 55 25 4.407 66 15 2,329 19 I.74S 3 235 I 56 7,866 '75 22 "5, 2,896 16 6 367 3 2,704 44 16 S.772 43 7 2,819 II 4 2,241 4 571 2 4.478 26 6 790 2 Sum- mation and Total 15.424 487 32 22,837 316 14 16,514 98 6 7.179 15 2 1,042 4 4 9.933 53 5 1.345 5 4 10,522 14,191 19,37s '74.274 138; 163 132 '978 13 " 7 »3 'The ratio was not calculated in the case of those states where both the population and the number of literati produced was very small, for such a ratio would be inaccurate and misleading. 'The population summation here given differs from that given in Table VIII because this figure includes the population of several southern and western states which produced no literati before 1 85 1 and which, therefore, were not included in the earlier table. ' This figure does not include eighteen Canadians, two white residents of the United States whose exact place of birth was unknown, and the two negroes, for one of whom the place of birth w'* also unknown. ^8 AMERICAN MEN OF LETTERS [58 All five tables which have been presented under the sub- ject of geographic environment thus show this same fact, namely, that there were great dififerences in the literary pro- ductivity of different sections of the country during the entire period studied. There are, of course, two possible ways in which the dif- fering literary productivity of the several states can be ex- plained, namely, in terms of nature or nurture. If one believes that nature is greatly predominant over nurture he may hold that this difference was due to the fact that the northern states were inhabited by persons of superior stock. If, on the other hand, he thinks that nurture is much more important than nature, he will explain the high literary fecundity of the north in terms of some environmental in- fluence. To prove the first of these theories it is necessary to prove two things, namely, that a great diversity of population elements was found in the several states, and that there also existed considerable differences of innate ability in the different population elements. There is no evidence that either of these conditions existed. In the first place, the relative nimibers of persons of different nationalities found in the several states were fairly uniform in 1790, and it was not till after 1840 that large numbers of immigrants began to come to America and congregate in the north.^ In the second place, it is still unproved, as will presently be shown, that the different nationality strains in the country varied widely in innate ability." Hence it appears that some factor of the environment must be sought to explain the differing literary productivity of the different sections of the country. ' Cf. A Century of Population Growth. Diagram 11, p. 118. ' Cf. infra, p. 89. jg] ANALYSIS AND INTERPRETATION OF DATA e,g It might possibly be inferred from a certain degree of correlation between different kinds of geographic environ- ment and different types of literary talent, brought out by Table VII, and also by the facts discussed in the last few pages, that geographic environment was causally related to the distribution of men of letters. A little study will show, however, that so far as correlation exists it was due chiefly to other factors. One might infer, perhaps, that the south- ern states were handicapped by heat, humidity and dis- ease. This is probably true to a considerable extent. It is evident, however, that climatic conditions cannot be con- sidered the predominant influence, because adjacent states possessing practically identical topographical and meteoro- logical conditions varied widely in literary fecundity. Some other influence must be sought to explain why, for example, Alabama ranked far below Georgia, and Rhode Island below both Massachusetts and Connecticut, while the District of Columbia stood far above either Maryland or Virginia. Thus one is led to conclude that though no doubt geographic environment did play a part in the production of literary talent, it was far less important than other factors.^ Density, also, might appear to have had an important in- fluence, since, for example, the relatively crowded New England and Middle Atlantic states were relatively more productive of men of letters. A superficial examination of the census reports, however, shows that there was not a uniform relation between the two conditions. If there had been such a uniform relation, Rhode Island, for instance, would have headed the list of states, instead of ranking* seventh, and Delaware would have been above Maine, New ' This conclusion is identical with that of Odin. He conceded that geographic environment has some influence in the production of men of letters, but considered that its influence is too small to be measured. Odin, op. cit., pp. 439 et seq. 6o AMERICAN MEN OF LETTERS [60 TABLE XIII American Literati Classified by Rank and by Character of Birth Place (St/ Capital, Chief City of State, County Seat, and other Places '") Abbreviations: M., merit; T., talent; To., total. State or Province Capital M. 49 3 5 8 I I I I Nova Scotia .... New Brunswick. . Quebec Ontario Maine New Hampshire . Vermont Massachusetts.. . Rhode Island . • ■ Connecticut .... New York New Jersey Pennsylvania . . . Delaware Maryland I i DistrictofColumbia' 4 Virginia | 2 North Carolina . . South Carolina. . Georgia Alabama Mississippi Kentucky Tennessee Louisiana Arkansas Ohio Indiana Illinois Michigan Wisconsin Missouri Unknown T. To. 20 5 2 I Total. 82 30 69 8 7 9 I I I I 5 2 Chief City M. T. To, 9 69 40 'S 112 161 47 ic 84 57 ie" 208 130 County Seat M. I I 4 3 4 30 I II 29 3 5 I I To. I 2 S 4 7 51 2 17 32 S 7 I 2 53 183 380 All others M. T. To 92 34 3 I 2 35 34 21 126 3 66 80 20 20 I 12 I 20 2 3 7 "7 497 753 Tota M. 5 4 2 4 43 36 18 171 7 77 172 16 58 3 26 4 22 5 9 II I 2 9 2 2 I 20 8 2 5 3 3 2 2 II 10 10 75 6 23 33 ID 27 5 2 7 3 2 2 I 247 ' Frequently a city was both the capital and chief city of a state, and both capitals and cities were usually county seats. When born in such a community, literati were credited ' capital rather than to the chief city, and to the chief city rather than to the county seat. location of a few county seats changed during the period studied, but the result of the investi can be affected only slightly by the fact that these changes were ignored. The list of ca chief cities and county seats used is that of 1850. 6i] ANALYSIS AND INTERPRETATION OF DATA gi Hampshire and Vermont, instead of ranking far below them. • Nevertheless it may still seem as though density were in some way connected with literary fecundity. In order to investigate further the relation of the two phenomena, a separate study of various aspects of the local environment, the third of the nine environmental influences, was made (Tables XIII to XV). Table XIII shows the men of letters classified as born in a state or provincial capital, the chief city of a state or province, a county seat, or elsewhere'. From this table it appears that the Capitals produced 11.2 per cent of the literati, the chief cities 20.8 per cent more, and the county seats added another 18.3 per cent. Thus it may be said that half (50.3 per cent) of all American men of letters were born in places which were relatively metro- politan, even though their actual population may not have been large. Further calculations showed that although, during the period studied, the capitals and chief cities of the several states had never contained over nine per cent of the total population of the United States, they had been the birth- place of approximately thirty-two per cent of the men of letters.'' Thus it appears that in proportion to population cities have been very rich in men of letters. ' Cf. Ward, Applied Sociology, pp. 169 et seq., and Davies, loc. cit., p. 232. ' County seats were not included because their population figures were not readily available. The population of 1850 for the cities under consideration was found by adding together their respective populations, as given in the Com- pendium of the Seventh Census, pp. 338 et seq. The total population for these cities constituted nine per cent of the total population of the United States. Since the urban population of the nation had increased from the founding of the Republic, this proportion was a maximum for the entire period considered. In this phase of the study the Can- 62 AMERICAN MEN OF LETTERS [62 The results of a more detailed study of the influence of cities is given in Table XIV, which shows the literary fecun- dity of the fifty leading cities of the country in 1850, and of six other cities which produced five or more literati.' In this table the cities are arranged according to the size of their population base, a figure obtained by summing fig- ures for the white population in each census year for the period 1781 to 1850 (Column I). Columns II and III give the number of literati of merit and of talent who were born in each city, and Column IV combines these two classes. Column V gives the number of literati born in each city between 1781 and 1850. This figure had to be used for comparative purposes, for population figures were available for this period only. Column VI, headed Index, contains a figure calculated by dividing the number of literati born in a city between 1781 and 1850 by the population base for that city. It indicates the relative literary fecundity of the city. Figures for cities having a population base of less than fifty [thousand] are not given, as they would be spur- iously accurate and therefore misleading. Enough figures are given, however, to show significant differences among cities. ' adian literati were not considered, because population estimates for Canadian cities were lacking. No correction was made for the influence of the colored population, which was overwhelmingly rural before 1850. If the study had been of white literati and white population only, the relative fecundity of the cities would appear somewhat smaller, but the general conclusion of the study would be the same. ^ Actually only fifty cities appear on the list. Figures for the five which have since been annexed to Philadelphia and for the area which has been annexed to Brooklyn are combined with the figures for the annexing cities. ' It is worthy of note that most of the cities which produced rela- tively large numbers of authors were also relatively productive of men of talent. 63] ANALYSIS AND INTERPRETATION OF DATA 63 TABLE XIV Absolute and Relative Numbers of Literary Persons Born in Important Cities City New York Philadelphia Baltimore Boston Brooklyn New Orleans . . . Cincinnati Albany Providence St. Louis Pittsburgh Salem Charleston Louisville Troy Newark Washington Buffalo ' Rochester New Haven . . . . Portland . Richmond Lowell Charlestovra. . . . Hartford New Bedford . . Portsmouth Newburyport . . Roxbury Lynn Utica Cambridge Reading Worcester Norwich San Francisco . Chicago Allegheny Norfolk Detroit Litchfield, Conn Syracuse Dorchester. . • • Bangor Columbus. ..-■■ Popula- tion Base' 1346 1038^ 421 412 274' 213' 199 139 117 III 95 91 80 80 74 73 7-2 71 67 63 60 60 60 54 53 S3 52 48 46 43 42 40 39 38 36 35 34 34 33 33 29 29 28 28 26 Rank of literati Merit Talent 69 40 15 49 4 I 10 5 I 2 4 3 I 9 6 2 I 5 S I 5 "5 «7 I 20 I I Total 57 16 69 5 2 9 8 4 I 14 7 I 2 I781- 1850 5 ! I i 2 I 10 i 13 i I' 6 I 7 i 7 i 11 I 80 49 16 53 5 2 9 8 4 I 12 7 I 2 5 3 2 9 '3 2 I 4 7 I 7 9 I Index * I 3 I 5 I 4 59 47 38 129 18 9 65 68 36 II 132 88 13 27 69 42 30 143 217 33 17 74 132 19 140 1 Cf. supra, p. 51. The figure for the population base was in each case derived by summing figures given for the population of a city, 64 AMERICAN MEN OF LETTERS TABLE yi\yi— Concluded [64 City Popula- tion Base Kank of literati Total 2 2 I ! i 1 1 1 Merit Talent Savannah Cleveland » 25 25 21 21 20 2 I I I 1 \ : Milwaukee ^ 1 I 1 Mobile 1 Hin^liELtn • • . • . , . , . 3 2 5 ' I:::::::: by decades, correct to the nearest thousand. Most of these figures were found in the Compendium of the Eleventh Census, Section on Population, Table 4a. The base is given in thousands. Three zeros [000] are omitted in each case. In the cases of Bangor, Portsmouth, Newburyport, Salem, Charles- town, iRoxbury, Dorchester, Hingham, Norwich and Litchfield, the population figures were compiled from figures given in the reports of each decennial census. In a few cases the data were incomplete, and population estimates had to be made for Cambridge, 1830; Charles- town, 1790; Litchfield, 1790 and 1830; Norwich, 1790 and 1830; Hart- ford, 1790; New Haven, 1790; and Albany, 1810. It would have been desirable to use figures for the white population only, but the colored population was not reported separately in the earlier censuses. A rough and partial correction was made for the colored population in cities south of Mason and Dixon's line and the Ohio River. It was assumed that the colored population had always constituted the same proportion of the population of these cities that it did in 1900. The appropriate figure was then subtracted from the total population in each decade, to obtain the figure given as the population base. Since the proportion of the colored population of these cities has tended to increase, the resulting figure is somewhat smaller than it should be in reality. The figure for the literary productivity of these cities is therefore correspondingly larger. In the case of the northern cities for which no correction for the colored population was made, the figure for the population base is of course somewhat too large, and the index is correspondingly small. 2 In determining the population base of Philadelphia and Brooklyn, figures for areas which have since been annexed to them were included, and literati born in those areas were of course also credited to the annexing cities. ' The population base of New Orleans contains no figures from censuses prior to 1810. ' No index is given for cities having a population base of less than fifty, as it would be spuriously accurate and therefore misleading. 65] ANALYSIS AND INTERPRETATION OF DATA 65 It appears from the table that some of the old cities of New England, such as Portland, Portsmouth, Newburyport, Salem, Boston, Cambridge, Hartford and New Haven, ranked very high. Other northern cities, such as Lowell, Lynn, Cincinnati, Newark, Brooklyn and Pittsburgh, ranked relatively low. Most of the southern cities also ranked low ; Charleston seems to have been a notable exception. How are these differences to be explained? Ward and Odin were of the opinion that the superior literary fecun- dity of cities in general is due to their superior educational advantages, and they explained differences among cities in the same way.^ On the other hand. Professor Thorndike points out the danger of assuming that educational opportunities entirely account for the high rank of cities when he says : " That cities give birth to an undue proportion of great men does not in the least prove that city life made them great ; it may prove that cities attract and retain great men, whose sons are thus city born." ^ It seems reasonable to believe that the theory suggested by Thorndike partially explains the differences existing among cities. For instance, the birth- place of those authors who were the sons of Yale and Har- vard professors was obviously determined by the fact that New Haven and Cambridge had attracted their fathers. This theory may also explain the low rank of the industrial cities of the north, which contained little to attract persons of literary taste. Again, this theory seems to explain ade- quately the low rank of most southern cities, when it is remembered that the cities of the south were almost exclu- sively commercial centers, and that the leisure classes of the south were very fond of country life. Finally, the rank of ' Cf. Applied Sociology, ch. ix, and Odin, op. cii., pp. 511 et seq. « Edward L. Thorndike, " A Sociologist's Theory of Education," The Bookman, vol. xxiv, p. 290. 66 AMERICAN MEN OF LETTERS [66 a small town may be profoundly affected by the influence of a single family of great ability, as was the case with Litchfield. It is apparent, on the other hand, that cities which, before 185 1, ranked high in men of letters, did possess superior educational opportunities, as Ward maintained. If not actually the seats of colleges, they were at any rate situated conveniently near them. In addition, they possessed an educational and literary tradition which must have been of no mean importance in stimulating the development of men of letters. Data are not at present available to show which of the two factors mentioned above was of more importance in the development of literati in cities. Facts are available, however, which show the importance of education in the development of men of letters in general. It must be ap- parent to the most casual observer that the states which ranked highest in literary productivity were those which possessed greatest educational opportunities. Their literary fecundity carmot be explained on the theory that they were inhabited by persons of superior stock, for reasons to be noted on a subsequent page.^ Some environmental influence has therefore to be credited with the differences observed, and educational opportunities are the most conspicuous and apparently significant factor in which the north and east dif- fered from the south and west.^ The influence of educa- tion, the fourth of the environmental factors to be consid- ered, will now be indicated in Tables XV to XVIII. '■ Cf. infra, p. 89. ' This is not the place for a discussion of why certain sections of the country furnished better educational opportunities than did others. Greater economic surplus, the superior energy of the northern people and the momentum of the Puritan educational tradition may, however, be suggested as among the more important reasons why some sections of the country were particularly liberal in their patronage of education. 67] ANALYSIS AND INTERPRETATION OF DATA 67 TABLE XV Education Received by American Literati, Classified According to Field of Chief Activity Abbreviations: G. S. P., partial grammar school course, or less; G. S., lull grammar school course; H. S. P., partial high school course; H. S., full high school course; A. B. P., partial college course; A. B., full college course. The symbol means either the formal education stated, or its equivalent. Field of chief activity d Patrons .... Librarians ' 4 Actors 3 Orators ; i Publicists 6 Narrators 3 Erudite 3 Popularizers . Speculative . . I^ose C/3 d 3 I 4 Poets 1 5 Dramatists Total . 36 4 3 7 7 3 12 8 I I I a, X w 60 2 2 3 S II 9 12 I 17 12 3 80 I I 2 2 6 12 18 21 2 44 21 m 133 I 2 8 13 II 23 I 18 17 4 9 2 15 35 24 104 16s 41 52 54 I 506 40 39 6 63 5° 34 66 II 82 31 41 7 50.6 2 IS 4 4 5 c8 1 19 IS S 88 23 33 24 71 70 '57 249 50 166 132 15 Table XV shows the education received by American hterati, classified by field of chief activity.^ From this table it appears that, with the exception of the two classes, actors and dramatists, there were more literati in each group who received a full college course than there were literati who received any other amount of education.^ It is obvious that ' In the following tables on education the equivalent of a given amount of formal training, when received during childhood and youth, is counted the same as that formal training. ' It is true that the facts were not available in the case of every author. The figures for the lower education groups would therefore probably be somewhat increased if the education received by all the literati studied were known. Presumably the majority of those authors whose education could not be learned received relatively little formal instruction, for education received by an individual is more likely to be recorded when ample than when scanty. The possible error can not be serious, however. 68 AMERICAN MEN OF LETTERS [68 an actor's education does not need to be academic. The dramatist is also quite as likely to be well equipped by close relations with the stage as by working with books. The figures indicate that for all other classes of men of letters, however, higher education was a great aid in achieving suc- cess. Even poets, who are reputed to be born and not made, enjoyed at least a partial college course in more than half of the cases recorded. Over fifty per cent of all the literati studied received a full college education. No figures are available for the number of college graduates in that part of the American people which was born before 185 1. Certainly they did not number more than a few score thousand.^ Since this com- paratively small number of people produced more literati than the tens of millions of persons without a college de- gree, it is apparent that the man or woman with an academic education was several hundred times as likely to be a per- son who would achieve literary distinction as was the person without that training." Tables XVI and XVII show, by decades, the education received by literary men and women. It appears that, in spite of some fluctuation, the degree of education received by literary men remained on the whole constant. By decades, from fifty-three to sixty-nine per cent were college gradu- ates. This relatively small fluctuation was accompanied by no consistent tendency for the proportion to increase or diminish. On the other hand, the degree of education received by women increased remarkably. While very few women born even as late as 1850 enjoyed a college educa- tion, the proportion who graduated from high schools in- ' In 1850 there were less than twenty-eight thousand students en- rolled in the colleges of the United States. Compendium of the Sev- enth Census, table cxlv. ' C-f. Cattell, Davies, Odin, op. cit. 69] ANALYSIS AND INTERPRETATION OF DATA 69 TABLE XVI Education Received by American Male Literati, Classified bv Period OF Birth Period of birth t/5 d 2 4 4 8 S 7 2 32 4 en d pi H. S. A. B. P. < Ox PC g 3 (2 Before 1771 ... 1771-80 1781-90 1791-1800 i8ot-io 1811-20 1S21-30 1831-40 1841-50 3 2 6 6 9 6 9 10 3 8 2 3 5 lO 3 2 5 5 23 18 16 IS 3 3 10 6 19 10 13 18 45 18 29 56 69 87 68 70 SS 69 II S8 63 56 58 S3 51 I 4 3 10 7 10 2 8 7 6s 33 44 96 110 '55 117 108 Total ^a ^a 87 10 83 10 497 58 58 n 859 Total percent.. TABLE XVII Education Received by American Women of Letters, Classified by Period of Birth Period of birth d CO d to PQ < H.S.,A.B. P.&A.B., per cent 1 1 3 Before 1771 ... .... 3 3 I I 1 3 4 8 7 1771 00 1781-90 1791-1800 l8oi-io 1811-20 1821-30 1831-40 ,. 1841-50.. -jU.. 1 1 2 : 4 3 2 2 1 1 6 4 3 2 3 5 7 2 I I 3 4 11 14 12 20 29 30 ?l 59 s 7 12 1 3 2 6 2 2 2 3 23 23 38 29 Total .'- 27 19 46 33 14 10 7 s 48 37 26 141 Total percent.. creased from zero to about sixty per cent. This latter fact is of great significance. It was noted in the discussion of Table III that in all probability the chief reason for the in- 70 AMERICAN MEN OF LETTERS [70 crease in the number of literary women was the improve- ment of the social environment. This improvement had another aspect besides the disappearance of the ban of dis- approval which used to rest upon women who entered the field of letters. It included also the decline of the idea that women should not receive higher education. Since women promptly and successfully invaded the field of literature as soon as these two obstacles to their activity were removed, it seems evident that public approbation and education were necessary factors for the creation of American women of letters. The education received by literary men and women of more than one field of activity is shown in Table XVIII. TABLE XVIII Education Received by American Literati, Classified by Sex and by One OK More than One field of Activity Men.. . One field . . . Number . Per cent . . More than Number. . one field ■ ■ Per cent - . f .-. c ij Number. . I One field.... pgj.^g^j_ Women, -j jj^^.^ ^^^^^ Number.. [ one field.. Per cent. .. a P^ (^ f:^ t