FRANZ BOAS (1858-1942)

Franz Boas as a Linguist

Murray B. Emeneau

When one begins to examine the contribution of the late Franz Boas to science in the linguistic sphere, the striking fact that leaps to the attention at once is that he is the guru, the ancestor in learning, of all those in this country who work in descriptive linguistics. If we drew up a genealogical table, the guruparamparã (as the Hindus call it), the succession of teachers and pupils, would come clear. It is only necessary, however, to mention Sapir and Kroeber, Jacobs and Andrade as his own pupils, and Leonard Bloomfield as a devoted admirer of his, and to note that practically without exception the younger generations are pupils of one or more of these men and often of Boas himself as well, due to the habit of young linguistic students in this country, better developed in them perhaps than in the apprentices in other disciplines, of indulging in Wanderjahre. The succession even goes over into the ranks of those who work in comparative linguistics, for all of those interested in the more exotic families and some of those in Indo-European and Semitics have sat at the master’s feet.

What shall we say are the reasons for this extraordinary influence? It was not merely Boas’ longevity that made the paramparā of so many linguistic students include his name. Longevity itself is not enough to bring students to a teacher’s door. It was indeed what he had to offer that drew them, and we must attempt to estimate his place in the history of linguistics in order to find out what this was.

Central in the estimate of his linguistic achievement must come his approach to the languages with which he was concerned and his success in dealing with them. Boas was not by any means the first to find that exotic, i.e., non-Indo-European and non-Semitic, linguistic material had scientific importance. Work on such languages had been done before him and, to be fair, had in a few cases been well done and in rare cases by Indo-Europeanists. Had Boas not lived and worked, some other agent or agents would have focused the attention of scholars on those unfamiliar languages that are the vehicles of the lowly and despised cultures of ‘ our primitive contemporaries.’ Nevertheless, he was, in cold and sober fact, the agent who brought this about more than any other, through his long years of devoted work and through his pupils and his pupils’ pupils. He set forth the ‘ manifesto ‘ for this study in the famous introduction to Bulletin 40, part 1, of the Bureau of American Ethnology, the Handbook of American Indian Languages, published in 1911, though he had used the method and taught it from the beginning. In the section on ‘ Linguistics and Ethnology ‘ he stated with insistence his creed that ‘ a command of the language [of a tribe] is an indispensable means of obtaining accurate and thorough [ethnological] knowledge, because much information can be gained by listening to conversations of the natives and by taking part in their daily life, which, to the observer who has no command of the language, will remain entirely inaccessible ‘ (p. 60). He admitted that this is an ideal impossible of attainment, though he himself and a few of his followers almost attained it in the case of some languages. But he conceded that approximations to this ideal are of the utmost value, both in allowing the investigator to establish rapport by means of the few phrases that he really commands and to bluff the tribesmen into believing that he already knows a great deal more than he really does.

However, this somewhat tricky fieldprocedure based on imperfect linguistic knowledge was for Boas of less value in the last resort than the ability of the investigator to record in writing in the native language extended series of texts dictated by tribesmen. Subsequent procedure involved linguistic analysis in the strict sense and then a process of expressing from the texts all the information that they could yield about the culture of the tribe and of correlating this with information derived otherwise. Moreover, only through the medium of such texts can the investigator approach the native rationalizations and evaluate them for what they are worth. This process, closely akin to part at least of what classical philology does for the cultures with which it is concerned, is supplemented for Boas, again meeting classical philology in method, by the study in the native language of the cultured verbal art-forms, poetry and songs, myths and tales, ritual texts, oratory, and the like. As he pointed out, while some ethnological material can be presented, though not studied, without use of the native language, for the art-forms and for any of the deeper problems of ethnology the native language is indispensable both in study and in presentation. Perhaps it was only Boas himself who ever presented a series of texts in the ideal way, with all the apparatus of ‘ philological ‘ handling, in order to derive from them all they could be made to yield for a study of the culture. It required the master’s hand to use the difficult method to its best advantage. I refer, of course, to his Tsimshian Mythology, where this method is used and combined with another technique of ethnological study, that of comparative folk-lore. But many of his followers have used the method successfully, if less explicitly. Many others have given it at least lip-service by publishing volumes of texts in native languages. Some of these, at least, have thereby performed one of the duties which Boas kept before himself, the rescuing from total oblivion of languages on the verge of extinction.

Of his success in dealing in the strictly linguistic way with the unfamiliar and difficult languages which he handled, we need say little. On perusal of the grammars which he wrote, we are struck by the combined ingenuity and soundness of his analysis in a field where the pedestrian had reigned for too long before him and where the fantastic and far-fetched can too easily creep in. He was at times at the beginning of his work thought by old-school linguists to be a peddler of the unbelievable, if not definitely mistaken in what he thought he had found. But his work brought it about that we can now calmly accept on good authority much that would have been laughed out of court a half-century ago. It is due to him that it would be impossible nowadays for an editor to reject a paper on such trivial grounds as those once used against him (in anecdote at least), that ‘ it is well-known that no word can exist which does not contain a vowel.’ He brought to his descriptive work an almost complete freedom from preconceptions, at a time when linguistic scholarship had not yet quite freed itself from the last preconceptions of prescientific lin- guistic study and when, rather worse, it was beginning, after the contribution from ancient India had been assimilated and con- solidated, to harden in a new set of preconceptions. He was able, because of this freedom of approach, to admit that in linguistics anything is possible and, consequently, to analyse exotic material without forcing it into the straitjacket of the familiar. This was perhaps the most valuable single lesson in analysis that he taught his pupils and they theirs. It has been the characteristic of American linguists that they have been unstaggered by the remarkably unfamiliar material with which most of them have dealt, that they have succeeded in bringing order into it and at the same time in refining techniques and methodological concepts, and that they have been able to return from exotic domains with methods which have yielded striking results when applied to the long-known, familiar material. Other linguistic schools there are, to be sure, but few of them have developed entirely untouched by the Boasian principles, even when they have not owed some of their being and growth to direct contact with Boas and his school.

We ought to note in passing the succinctness of Boas’ presentation of linguistic data. The seeds of the stringency and quasi-mathematical type of formulation seen in most of the recent American linguistic work are already there in his work, though the immediate urge in this last direction has come from another source. Most of his pupils were able to imitate him in this feature in a greater or less degree, though at times the imitation degenerated into barrenness and was quite devoid of the elegance with which Boas and his best pupils managed to invest their presentations.

In historical linguistics Boas’ influence was not so great as in the other part of the linguistic field. He observed and published corre- spondences between closely related dialects, but hardly attempted to construct on a strict basis of’ phonetic law ‘ groupings of the order of the Indo-European family. This was, in part at least, due to the tremendous differentiation between the languages that he handled. In greater part however, one should attribute it to his attitude to historical studies in general, a matter that may be left to the ethnologists to evaluate. It issued at the end in his doctrine of language mixture, stated thus:

We should rather find a phenomenon which is parallel to the features characteristic of other ethnological phenomena—namely, a development from diverse sources which are gradually worked into a single cultural unit. We should have to reckon with the tendency of languages to absorb so many foreign traits, that we can no longer speak of a single origin, and that it would be arbitrary whether we associate a language with one or the other of the contributing stocks. In other words, the whole theory of an ‘ Ursprache ‘ for every group of modern languages, must be held in abeyance until we can prove that these languages go back to a single stock and that they have not originated, to a large extent, by the process of acculturation.¹

Although this statement admits the validity and necessity of the genetic attack, yet it seems to warn off those who would attempt to apply that attack where its applicability does not strike the student as obvious at first glance. Such a position would probably have prevented some of the achievements of scholars in the Indo-European field. For the Indo-Europeanist Boas’ statement seems dangerously like manufacturing unnecessarily a dilemma in order to impale oneself on one horn of it, whereas a better view of the situation should recognize a combination of factors. It did however lead Boas to look searchingly for evidence of diffusion of phonetic, morphological, and syntactical features between languages unrelated genetically (at least so far as is known), while in similar circumstances others had on the whole been content when they had pointed out the diffusion of lexical items; and he found many items that carry conviction. The Boasian view, moreover, if it had been thoroughly assimilated, would have warned many of his ethnological followers in this country that the famous Sapirian reduction of American Indian stocks, or that of Schmidt in Indo-China and Malayo-Polynesia, or other similar ones, can be accepted only as sketch-maps for further exploration ; no far-reaching hypotheses in extra-linguistic fields should be based on them.

Finally, it remains to be said, what could be said better by one of his own students, how Boas encouraged the student by interest and forbearance, how he moved heaven and earth to obtain funds and backing for the field-work of linguistic projects, how successfully he maintained outlets for the presentation to the world of work in progress or completed. Not least of these outlets was the monthly Sunday meeting of linguistic scholars and students held usually at Columbia University during recent years under his aegis and attended regularly by scholars from points as far apart as Boston and Phila- delphia. American linguistics would have had a different flavor, had he not lived. Above all, it would have lacked that personal flavor that was given to it by the warm relationship of pupil and teacher that existed between all the epigonoi in the descriptive field and the grand old man who was called by the younger generations at least, Papa Franz.

Source : Murray B. Emeneau, ‘ Franz Boas as a Linguist,’ Memoirs of the American Anthropological Association 61.35-38 (1943). By permission of the Editor of the American Anthropological Association and the author.

¹ American Anthropologist 22, pp. 374-375. The rigid application of this statement of principle is not at all softened in the later article in Language 5, pp. 1-7.

Franz Boas’ Approach to Language

Roman Jakobson

When trying to sum up the linguistic heritage of Franz Boas, I vividly relive our long conversations on the science of language— talks or rather delightful lessons, where the great master initiated me into problems which absorbed him during the last years of his life. How he loved this science! In the autumn of 1942 a telephone call from Boas informs me that he has been ill, but is feeling better today and asks me to visit him.—In the afternoon?—’ No, at once, later it may be worse, and 1 would like so much to have a linguistic talk.’ And in order to justify such haste he adds : ‘ It was so hard to spend ten whole days without scientific work.’ There is something of Marcus Aurelius in this sentence as well as in Boas’ whole life.

Linguistics was often erroneously thought to play a subordinate role among his manifold activities. It is true that he came to the humanities from an entirely different field ; at first Boas specialized in physics and geography, and he always declared himself a self-made- man in ‘ the science dealing with the mental phenomena,’ particularly in linguistics. The only linguist he met in his student years was Steinthal, but Boas was not yet interested in language and afterwards he regretted never having attended the lectures of that enquiring thinker. Self-instruction can become a danger, but in Boas’ case it was his great power : he remained free of the various prejudices and antiquated survivals which weighed heavily on linguistics and ethnology. He came from the natural sciences with a demand for reliable and rigid method but he had no ambitions to force naturalistic habits on the humanities. On the contrary, he asserted and espoused the autonomy of the humanities, and just because he knew perfectly both domains—the natural and the social sciences—he could never confound them and carefully distinguished ‘ human language, one of the most important manifestations of mental life,’ and cultural phenomena in general from their ‘ biological premises.’ He repeatedly insisted upon the impossibility of explaining a linguistic or some other cultural structure as a product of the natural environment, and he confessed both his former exaggerated belief in the importance of geographical determinants with which in his youth he had started his first expedition (1883-84), and ‘the thorough disillusionment in regard to their significance as creative elements in cultural life,’ a resolute disillusionment which is reflected already in his first piece of ethnological work—The Central Eskimo (written in 1885).

It is worth mentioning that this very trip to Baffinland definitely turned the interest of the scientist from geography to ethnology, and the leading place in his wide ethnological work belongs to linguistics. The first Boas study on American Indians and his first contribution to Science (1886) was devoted to language. Curiously enough it is a ‘ letter from Berlin ‘ : his field research with ‘ language of the Bella Coola in British Columbia ‘ took place at a Berlin exhibition to which some natives of this tribe were brought. Thereafter, the languages of British Columbia became a favorite field of Boas’ exploration. On one of these languages, Kwakiutl, he continually worked more than a half century, and his last finished manuscript, which occupied the final years and days of his life, is a comprehensive linguistic analysis of the Kwakiutl (Grammar ; Dictionaries of Suffixes and Words ; Texts with Translations). In the field of Indian languages it is now the most exhaustive and in many respects exemplary description, which should be published as soon as possible.

Language was considered by Boas not only as a part of ethnological phenomena in general but even as ‘ one of the most instructive fields of inquiry,’ and his motivation is thoroughly remarkable : ‘ The great advantage that linguistics offers in this respect,’ Boas tells in his magnificent introduction to the Handbook of American Indian Languages (1911), ‘ is the fact that, on the whole, the categories which are formed always remain unconscious and that for this reason the processes which lead to their formation can be followed without the misleading and disturbing factors of secondary explanations, which are so common in ethnology. . . .’

This statement seems to us one of the most daring, most fertile and pathbreaking ideas ever uttered by Boas. As a matter of fact, just this unconscious character of linguistic phenomena has made and still brings so many difficulties to the theoreticians of language. Even for the great Ferd. de Saussure there was an insoluble antinomy. In his opinion every state in the life of a language is a ‘ fortuitous state ‘ inasmuch as ‘ individuals are in a great degree unconscious of the language laws.’ Boas proceeded from exactly the same starting point : although ‘ the fundamental ideas of language ‘ are in constant use by a speech community, normally they do not emerge into the consciousness of its members. But the traditional doctrine has become permanently inhibited by ‘ the unconsciousness of linguistic processes,’ whereas Boas (and Sapir also in this respect truly con- tinues his path) knew how to draw the due conclusions from such premises ; the individual consciousness usually does not interfere in the grammatical or phonemic pattern of language and consequently does not ‘ give rise to secondary reasoning and to re-interpretations.’ The conscious individual re-interpretations of fundamental ethnic habits are capable of obscuring and complicating not only the real history of their formation but also their formation itself. Meanwhile the formation of linguistic structures, as Boas emphasizes, can be followed and unfolds itself without these ‘ misleading and disturbing factors.’ Linguistic elementary units function, without the necessity of each unit itself entering into consciousness and becoming a separate subject of unschooled thought. They can hardly be isolated one from another. And consequently this relative non-interference of the individual consciousness in language explains the rigid and imperative character of its pattern—a whole where all parts hold firmly together. The weaker the consciousness of the customary habits, the more their devices are stereotyped, standardized and uniform. Hence the clear-cut typology of the diverse linguistic structures and above all the universal unity of their fundamental principles which repeatedly impressed Boas’ mind : ‘ relational functions ‘ presenting necessary elements of every grammar and phonemics all over the world.

Among the various ethnological phenomena the linguistic processes (or rather operations) exemplify most strikingly and plainly the logic of the unconscious. For this reason—Boas insists—’ the very fact of the unconsciousness of linguistic processes helps us to gain a clearer understanding of the ethnological phenomena, a point the importance of which cannot be underrated.’ The place of language with regard to the other social systems and the meaning of linguistics for a thorough insight into the diverse ethnological patterns had never been stated so precisely. And modern linguistics will still give some suggestive lessons to the searchers in the various branches of social anthropology.

In accordance with these general views Boas endeavors ‘ to subject the whole range of linguistic concepts to a searching analysis ‘ and in his descriptive studies of various Indian languages he tries to seize their ‘ inner form ‘ and to attain the most objective, most literal and least distorting translation of their unconscious categories into the language of scientific formulations. From the start he distinctly saw the cardinal task in its double aspect : as a limited number of sounds and sound-clusters have been selected by each single language (and moreover by human language in general) from the infinitely large mass of possible sounds and sound-clusters, so the infinitely varied range of ideas has been reduced by classification in each single language (and moreover in human language in general) to a lesser number. Thus the search has to specify the bilateral ‘ selection of material used for expression ‘ : the choice of sound features to discriminate concepts and the choice of conceptual features to be ‘ recognized by the symbol of the same sound complex.’ This two-sided selection converts foreign bodies into linguistic values ; it creates fixed PHONEMIC units from the sound matter and fixed SEMANTIC units from the conceptual matter.

In Boas’ synthetic sketch ‘ Language ‘ written for General Anthropology (1938) there is clearly put forward the system of phonemes and grammatical devices as a necessary base of linguistic analysis. He wholly understood that the linguistic problem of sound selection was fruitfully solved by the ‘ phonemic principle ‘ and he appreciated Sapir’s important achievements in this matter. The necessary ecoпоту of energy forbade him in the decline of life to alter the technics of his work, but he excellently familiarized himself with this new field of research. Once, a year before his death, Boas delivered an instructive talk to some linguists on the structure of Kwakiutl. He mentioned in passing that the Kwakiutl accent is on the first syllable, if it is long, and on the second syllable, if the first is short. Since a short syllable comprises one MORA and A long syllable two MORAS, I proposed to reduce both these conditional judgments to a common denominator : the accent falls on the syllable containing the second mora of the word. Then we hearers went home under the impression that the question remained strange to Boas, but some months later in explaining to Lévi-Strauss and me another Indian prosodie pattern, ‘ as you would say ‘ he suddenly added and immediately translated the accent rule in question into my combined count of moras and syllables. He wonderfully grasped and understood the language of others.

As a matter of fact, ethnology and particularly linguistics was for Boas first and foremost a means to understand the other and perceive oneself from without. Soon after Boas had entered the field of ethnology he tried to define its aims (1888) : ‘ The data of ethnology prove that not only our knowledge, but also our emotions are the result of the form of our social life and of the history of the people to whom we belong. If we desire to understand the development of human culture we must try to free ourselves of these shackles.... It is impossible to determine a priori those parts of our mental life that are common to mankind as a whole and those due to the culture in which we live. A knowledge of the data of ethnology enables us to attain this insight. Therefore it enables us also to view our own civilization objectively.’ The Copernican struggle for the scientific overcoming of the stubborn, narrow-minded, self-confident egocentrism still continues. We have come to know that our space as well as our time is only one among the innumerable varieties both of space and of time. Boas’ task in the development of linguistics could be compared with the historic role of a Lobschevsky, an Einstein and other fighters against self-centered tradition. In the Introduction cited, Boas opens the ‘ Discussion of Grammatical Categories ‘ with the following lucid statement : ‘ Grammarians who have studied the languages of Europe and western Asia have developed a system of categories which we are inclined to look for in every language. It seems desirable to show here in how far the system with which we are familiar is characteristic only of certain groups of languages, and in how far other systems may be substituted for it.’

Of course, this trend in linguistics could be traced as far back as Humboldt’s time. But nevertheless the ‘ Indo-European imperialism ‘ (as Russian linguists joke) even recently continued either deliberately or unconsciously to sway the study of’ exotic ‘ languages. And only with Boas’ attacks there began a veritable change of approach. He did not content himself with declaratory slogans and with criticism of older grammars ‘ modeled strictly on the [quasi-] Latin scheme, which obscures the characteristic psychological categories of Indian languages.’ Above all, he convincingly showed ‘ that each language has a peculiar tendency to select this or that aspect of the mental image which is conveyed by the expression of the thought.’ Different languages differently select those aspects of experience ‘ that must be expressed.’ Such ‘ obligatory aspects are expressed by means of grammatical devices,’ whereas some other aspects are taken as non-obligatory and are expressed by lexical means. And each language in its own way chooses the concepts to be expressed by single simple terms or by combinations of distinct terms, by entirely heterogeneous or by related terms.

Numerous grammatical characterizations of various Indian Ianguages by Boas and his followers still more successfully try to shake off and eliminate every linguistic prejudice, every familiar and school-acquired habit of the authors and every bent to impose categories derived from our languages upon foreign languages. Already in his earliest linguistic reflections (On Alternating Sounds, A A 1889), Boas endeavored to prove that everyone ‘ apperceives the unknown sounds by the means of the sounds of his own language ‘ and therefore in linguistic records we must carefully work off our usual misspellings ‘ due to the phonetic system of our native language.’ The modern outlines of Russia’s various languages pre- sent the closest and most valuable counterpart to these fruits of the American school. Boas saw the necessary sequel of such work in exhaustive scientific dictionaries which should with the greatest accuracy interpret the value of native words without any of the ordinary concessions to our own semantic patterns. He emphasized this problem of semantic accuracy in his last public address (before the American Ethnological Society, May 13, 1942), and he tried to embody his program in the aforementioned Kwakiutl dictionary.

Thus Indian languages came to be first scientifically treated, but the significance of Boas’ linguistic investigation goes much farther. We familiarized ourselves with unfamiliar languages, we looked at them from within in order to find an objective approach to our own languages, and already this Boasian PRINCIPLE OF RELATIVITY has—in Emeneau’s words—’ yielded striking results when applied to the long-known, familiar material.’ Moreover this principle modifies our ideas not only on this or that language, but on language in general. We had learned that each language is arbitrary in its classifications, but this traditional (particularly Whitney’s and Saussure’s) statement is subjected by Boas to an essential restriction: indeed, the Introduction says, each language may be arbitrary, but solely ‘ from the point of view of another language ‘ in space or in time. In a mother-tongue whether ‘ primitive ‘ or ‘ civilized ‘ no classifications are arbitrary for its speakers. Such classifications develop ‘ in each individual and in the whole people entirely sub-consciously ‘ and build a kind of linguistic mythology which may direct the attention of the speaker and some mental activities of the given speech community in definite lines. Thus linguistic forms exert an influence not only upon poetry and beliefs but even upon speculative thought and ‘ scientific views, which are apparently based entirely on conscious reasoning.’ In itself every grammatical pattern, a ‘civilized’ as much as a ‘primitive’ one, is in permanent conflict with logical reasoning, and nevertheless every language is at the same time ‘ sufficiently pliable ‘ to any terminological needs of culture and ‘ to more generalized forms of thinking,’ which ‘ give a value to new, formerly unidiomatic expressions.’ Civilization requires only an adaptation of vocabulary and phraseology, while grammar may remain intact.

Thus ‘ it is not justifiable to consider languages as hindering or favoring cultural development.’ When a native of the far North saw a camel for the first time, he put it down as a distorted horse. Likewise we are sub-consciously inclined to take unfamiliar, remote linguistic structures as backward, defective or perverted. We are subconsciously living still in the Ptolemaic Universe and we still believe we stand in the center of the world. ‘ It is somewhat difficult for us to recognize that the value which we attribute to our own [form of] civilization is due to the fact that we participate in this civilization.’ By numerous examples Boas demonstrated how utterly unwarranted is such an overestimation of our own linguistic habits. For instance he pointed out our first person plural as a ‘ logical laxity ‘ avoided by many exotic languages, where we find ‘ the more logical distinction ‘ between ‘ I and you ‘ and ‘ I and he.’ And in his last published linguistic study Language and Culture (1942) Boas wittily remarked that we could read our newspapers with much greater satisfaction if in the same way as Kwakiutl our language, too, would compel them to say whether their report is based on self-experience, on inference, or on hearsay, or whether the reporter has dreamed it.

Descriptive linguists found in Boas one of its leading representatives, but description was never for him a self-sufficient, ultimate aim. ‘ In order to give each language its proper place,’ Boas requires us ‘ to compare as well the phonetic characteristics as the characteristics of the vocabulary and those of the grammatical concepts.’ He truly belongs to comparative linguistics in the broadest sense of this term.

The first anniversary of Boas’ death falls fifty years after the death of William D. Whitney. Both were enabled to have their say to the end, whereas the last of the three great men lost by American linguistics, Edward Sapir, could be commemorated by the sad Kwakiutl verb wibãlisEm to perish without reaching the end. The work of these three prominent representatives of American linguistic tradition was equally essential in descriptive and in comparative linguistics.

Whitney, whose ideas exercised a mighty, decisive influence over international linguistic thought, was particularly praised by European scientists (as for instance Leskien, Masaryk and Saussure) for his convincing demonstration that linguistics belongs to social and not to natural sciences. Both Boas and his pupil Sapir continued to battle with the old naturalistic deviation in the science of language. One of the most tenacious survivals of the traditional naturalism was Schleicher’s genealogical tree of languages, an idea or rather a myth which in spite of all criticism still pressed heavily on comparative linguistics. Boas began by embracing this tenet, and in ‘ Classification of the Languages of the North Pacific Coast,’ a paper read at the Chicago International Congress of Anthropology (1893) he taught: ‘The structural resemblance of the two languages [Tlingit and Haida].... can be explained by the assumption of a common origin,’ and although in vocabulary ‘ the similarities are doubtful... nevertheless, the structural resemblance must be considered final proof.’ At this point Boas, contrary to Major J. W. Powell’s classification of the North American languages, has even surpassed the precepts of orthodox Indo-Europeanists who derive the evidence of cognation exclusively from the MATERIAL likeness of words or morphemes. These precepts were transgressed by many observers of exotic languages, who often found an impressive STRUCTURAL similarity in their grammatical or phonemic pattern but almost without common roots and affixes. The more deeply Boas delved into indigenous linguistic life, the more clearly he saw, that side by side with DIFFERENTIATION stands INTEGRATION, another and opposite factor which works widely. This time is was ethnology—and especially its notion of DIFFUSION— which served as an example for linguistics.

Many significative data prevented inferring from any striking similarity in neighboring languages to the community of their origin: evidently unrelated but contiguous languages frequently manifest a range of common features in their grammatical and phonemic structure. Grammatical and phonemic peculiarities are distributed over large continuous areas and spread over one part of some related languages (or even over one part of a single language) without extending to the other part. Certain grammatical and phonemic types have a wide continuous distribution without corresponding lexical similarities. Some neighboring languages with similar phonemic features are morphologically quite distinct and vice versa. The areas of single grammatical or phonemic features do not coincide, so that one and the same language happens to be linked by different features with quite different languages.

In America Boas discerned various vast areas which have common grammatical or common phonemic characteristics. As early as the Introduction he envisaged the diffusion of phonetic, syntactic and even morphological traits ‘ beyond a single linguistic stock,’ and the evidence seemed to him ‘to be in favor of the existence of farreaching influences of this kind,’ but he considered them ‘ still obscure.’ Meanwhile convincing examples still accumulated and the fact of the frequent occurrence of morphological or phonemic assimilation between contiguous but unrelated languages became indisputable for Boas who, since his introductory paper to the International Journal of American Linguistics (1917), again and again returned to the subject. But there remained still a certain wavering in his explanation of these surprising facts. So in The Diffusion of Cultural Traits (Social Research, 1937) he appeals even to an ‘ actual intermingling of tribes,’ although nearly at the same time in The Mind of Primitive Man (1938) Boas soundly stresses that ‘ assimilation of cultures occurs everywhere without actual blood-mixture, as an effect of imitation. Proof of diffusion of cultural elements may be found everywhere. Neither differences of race nor of language are effectual barriers for their spread.’ Let us add, that differences of language are particularly no hindrance to the diffusion of phonemic or grammatical devices. The integration is the natural trend of linguistic intercourse and this trend naturally does not remain confined to the bounds of a single language or linguistic family.

At first Boas did not realize the full implications of his discovery. He surmised that the conditions favoring such diffusion must have been much more frequent in primitive America, Africa or Oceania than in Europe or Asia. Meanwhile just in Europe and Asia wide continuous areas of single phonemic as well as morphological features which spread without regard to the genetic relation of languages were disclosed by some Old World linguists quite independently of Boas’ inquiry. Then a phonemic atlas of the world was planned, and already the first preparatives showed that the wide continuous distribution of pivotal phonemic and grammatical features is generally typical of linguistic life.

Boas gradually generalized his Americanistic experience and his attitude toward the ‘ genealogical tree ‘ became still more critical. ‘ The whole theory of an Ursprache for every group of modern languages must be held in abeyance until we can prove that these languages go back to a single stock and that they have not originated, to a large extent, by the process of acculturation ‘ (1920). For a long time Boas thought he was isolated in these views. Only in the last years did he learn about parallel investigations in Prague, Copenhagen, Oslo and elsewhere in Europe. He came to read Trubetzkoy’s Gedanken über das Indogermanenproblem in Acta Linguistica 1939. This posthumous paper brilliantly develops the following idea : ‘ Es gibt eigentlich gar keinen zwingenden Grund zur Annahme einer einheitlichen indogermanischen Ursprache, von der die einzelnen indogermanischen Sprachzweige abstammen würden. Ebenso gut denkbar ist, dass die Vorfahren der indogermanischen Sprachzweige ursprünglich einander unähnlich waren, sich aber durch ständigen Kontakt, gegenseitige Beeinflussung und Lehnverkehr allmählich einander bedeutend genähert haben, ohne jedoch jemals mit einander ganz identisch zu werden.’ ‘ It’s all right,’ Boas told me cheerfully and heartily in returning the book. The bitterness of loneliness disappeared.

Perhaps the long inattention of his entourage to Boas’ favorite idea had been partly his own fault. He often presented his discoveries as a mere criticism of current theories. News on the discovery of America would be given by Boas as a disproof of the hypothesis on a shorter way to India, while data on the new part of the world would be mentioned only casually. He fervently insisted on ‘ limitations of comparative method,’ but he did not strive to make clear that in fact his outlook upon diffusion was designed first of all for enlarging the limits of historical comparison and for building up the historical geography of the linguistic world. Historical research, as Boas acknowledged perfectly well, ‘ remains equally valid, whether we assume purely genetic relationship or whether we ask ourselves whether by contact languages may exert far-reaching mutual influences ‘ (1936).

Among linguists Boas was often held to be interested solely in synchronic study and some approved of him, others did not agree. However the searcher himself never wearied of repeating that such a study is merely a way toward history. For him every social science was in the last resort a historic science : ‘ Anthropology deals with the history of human society ‘ (1938) and the study of languages purposes ‘ to unravel the history of the growth of human language ‘ (1920). This maxim evidently carries on Whitney’s tradition. In Boas’ opinion the diachronie, historic approach is superimposed upon the synchronic method not only as the aim of inquiry upon its means, but likewise as the explanation upon mere description. From such a viewpoint one would scarcely understand a phenomenon without knowing, ‘ how it came into being.’ And under these circumstances an exact classification of the now existing types ‘ can only be a substitute for the genesis and history.’ Boas does not look for general laws beyond the historical aspect of language, as if synchrony were merely a domain of casual particulars.

However Boas is rather disappointed by the searches for the general laws that govern and explain the historical sequences of culture and particularly the DEVELOPMENT of language : although linguistic data offer us many strikingly similar phenomena in remote parts of the world and disclose their independent origin, meanwhile any attempt to ground them on a similar, unilinear development breaks down. Quite similar structures may arise by fully unlike ways from dissimilar sources and may evolve in a multitude of divergent ways. A likeness of structures does not imply a similar line of development. And as this discrepancy became clearer, the search for evolutive laws of language (and of social life in general) seemed to Boas still less profitable and more hopeless.

Such an experience was indeed discouraging for some linguists who kept the faith that the only conceivable laws in language are evolutive. But Boas saw that the conditions determining the course of historical happenings ‘ are logically entirely unrelated ‘ (1930), and on the other hand his attention was attracted more and more by ‘ general forms ’ of language ‘ that are determined logically ‘ (1924). He had a clear inkling of these general devices which—either positively or negatively—underlie every linguistic pattern and which determine the typology of languages. A further step had to be made, and actually in the recent development of linguistic thought a new problem claims our attention : research on the laws that govern and explain the STRUCTURE of languages. Such synchronic or more correctly PANCHRONIC laws are superimposed on historical linguistics : without understanding a linguistic structure as such, we could scarcely explain, ‘ how it came into being ‘.

These structural laws prove to be both determinable and explainable if they are submitted to an internal linguistic test. For instance the question, why most languages do not carry out the same sharp, logical distinctions in the plural as they do in the singular, was quoted by Boas (1896) as ‘ difficult to answer.‘But recently Viggo Bröndal, the great Danish linguist who died on the same day as Boas, pointed out the general linguistic tendency to avoid an excessive complexity within a morphological formation : often forms which are complex in regard to one category of classification, are relatively simple in regard to another category. In accordance with this ‘ law of compensation י plural as a grammatical number, which is more thoroughly specified than singular, usually contains a smaller and never a larger set of distinctions.

The stubborn tradition which identified the scientific explanation with the genetic approach and reduced synchronic linguistics to a mere description influenced Boas, too ; nevertheless his linguistic theory came ever nearer to the fundamental problem of structural laws. And at this crucial point American linguistic thought of today and tomorrow will have to develop the legacy of the great teacher, the splendid heritage from which generations of linguists on two continents will still draw new problems, observations and suggestive ideas.

Source : Roman Jakobson, ‘ Franz Boas’ Approach to Language,’ International Journal of American Linguistics 10.188-195 (1944). By permission of the author.