22
On Comparative Methods in Social-Cultural Anthropology and in Linguistics Author(s): Gopala Sarana Source: Anthropological Quarterly, Vol. 38, No. 1 (Jan., 1965), pp. 20-40 Published by: The George Washington University Institute for Ethnographic Research Stable URL: http://www.jstor.org/stable/3316722 Accessed: 09/10/2009 19:29 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ifer. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The George Washington University Institute for Ethnographic Research is collaborating with JSTOR to digitize, preserve and extend access to Anthropological Quarterly. http://www.jstor.org

3316722

  • Upload
    soyo

  • View
    214

  • Download
    0

Embed Size (px)

DESCRIPTION

ioi\]\i

Citation preview

Page 1: 3316722

On Comparative Methods in Social-Cultural Anthropology and in LinguisticsAuthor(s): Gopala SaranaSource: Anthropological Quarterly, Vol. 38, No. 1 (Jan., 1965), pp. 20-40Published by: The George Washington University Institute for Ethnographic ResearchStable URL: http://www.jstor.org/stable/3316722Accessed: 09/10/2009 19:29

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=ifer.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

The George Washington University Institute for Ethnographic Research is collaborating with JSTOR todigitize, preserve and extend access to Anthropological Quarterly.

http://www.jstor.org

Page 2: 3316722

ON COMPARATIVE METHODS IN SOCIAL-CULTURAL ANTHROPOLOGY

AND IN LINGUISTICS

By

GOPALA SARANA

Peabody Museum, Harvard University

Often even the most elementary type of empirical (social) re- search is supposed to be based on experimental inference. But experimentation presupposes certain conditions. To experiment means, it seems, to tamper with the natural situations. The idea of control is basic to it. There are certain logical premises, which, in the form of models, have been guiding the empirical-experi- mental inference. John Stuart Mill called them 'the four methods of experimental inquiry.' They are the Method of Agreement, the Method of Difference, the Method of Residues and the Method of Concomitant Variations. Beginning with Dr. Whe- well, who was Mill's contemporary, this scheme has been sub- jected to various criticisms.

The Method of Residues is not a pure method of induction. Deduction plays an important part in arriving at conclusions through its application. Moreover, in Mill's own words, it is "in truth a peculiar modification of the Method of Difference" (1911:260). A casual relationship between two phenomena may be inferred when two or more instances of these have only one circumstance in common (the Method of Agreement) or have every circumstance in common save one (the Method of Dif- ference). These 'methods' are difficult to apply in the case of social phenomena. Here controls are difficult. In many cases they are undesirable too. In the Method of Concomitant Vari- ations, on the other hand, circumstances need not be artificially produced at the will of the observer. Instead of the qualitative elimination of the circumstances only their quantitative co- variation is required. Mill has pointed out that in order to be able to apply the Method of Concomitant Variations fruitfully, the concomitance itself must first be proved by the application of the Method of Difference (1911:263).

20

Page 3: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

In the social sciences experimentation is not possible in the same way and to the same extent as it is in the case of the natural sciences. The social science methods can at best be called 'quasi- experimental.' The method which can approximate experi- mentation in the natural sciences is usually called, to quote Nadel, "somewhat loosely, the Comparative Method" (1951:222). For

establishing sociological proofs, according to Durkheim, "the method employed is that of indirect experiment, or the com-

parative method" (1938:125). Parsons also feels that 'the in-

dependent variation' of the values of the analytical elements cannot be empirically demonstrated without comparative method (1937:743). But Mill tells us much more than this. Without a

comparative method even dependent-variation or co-variation cannot be effected. Our contention is strengthened by the fact that Mill's four Experimental Methods use one sort of comparison or the other. In other words, all of them are comparative methods.

If we were to follow Oscar Lewis, technically speaking, anthro-

pology could not be called a science. He discusses comparisons in

anthropology because "there is no distinctive 'comparative method' in anthropology" (1955:259). For him 'comparison' is rather a generic aspect of human thought than a special anthro-

pological method. In that case there is no justification in even

talking of 'comparisons in cultural anthropology.' But Lewis does so. In spite of all these commissions one will readily agree with him that to identify 'the comparative method' with a particular anthropological research design or school of thought is un- fortunate. This was recognized by Schapera and others at least two years before Lewis. They concluded thus: "that there is no single method of comparison in anthropology, that method is largely determined by problem, and that a method appropriate for the comparison of kinship systems is not necessarily most

appropriate for other types of cross-cultural comparisons" (Singer 1953:362). This should not be taken as a support for Lewis's contention. The Symposium very rightly felt that it was not possible to discuss the problems of comparative method apart from the aims and problems of comparison. In fact, the first includes the last two. One might think that Oscar Lewis did not like the term 'comparative method'-as was the case with

21

Page 4: 3316722

ANTHROPOLOGICAL QUARTERLY

many Boasians-and wanted to substitute 'comparison' for that. But he specifically denies that the change is only semantic (1955: 259). He rather wants to highlight the fact that "the method of a comparison is only one aspect of comparison." He is not correct in his use of the word 'method.'

By its very definition, method implies a goal or an end to be attained. It is a system, or a complete set of rules of procedure, for achieving a given goal. By appropriate choice the suitable research 'techniques' are included in any such set of rules of

procedure. The techniques are the tools included in the frame- work of the method concerned for the attainment of the object. [Century Dictionary and Cyclopedia, Shorter Oxford English Dictionary, Lazarsfeld and Rosenberg 1957 and Parsons 1937]. It is true that the 'technique' of comparison is only one aspect of a comparative method. Lewis is not sure about his use of the term 'method.' In one of his usages he makes method equivalent to technique. It is interesting that after denying earlier that there is any distinctive comparative method in anthropology Lewis is found talking of "the recent writings on comparative methods and problems in anthropology" (1955:262). Thus in spite of himself he echoes the proper usage of the term method as out- lined by us above, and by Herskovits's statement: "the term method implies more than the actual procedures employed in

prosecuting a given research project" (1954:5). During the nineteenth century comparative anatomy was the

'glamour science' of the day. Its methodological influence on the social studies was immense. But much before the establishment of

comparative anatomy as a discipline F. Lafitau (1670-1740) used a type of comparative approach for his 'anthropological' work. Evolutionary ideas in social fields antedate those in the

biological sciences. The former is said to have stimulated the latter. By the end of the eighteenth and the beginning of the ninetenth century anthropological studies were no longer in-

spired by their forebears, such as Condorcet and Voltaire. In- stead, the method shaped by the immensely popular discipline of

comparative anatomy was borrowed. Anthropology had not yet been subdivided into physical and cultural sections. Blumebach, Pritchard, Foster and others were the chief contributors. A

majority of the anthropologists then were either physicians or

22

Page 5: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

comparative anatomists by training. By the middle of the nine- teenth century, ethnology or cultural anthropology became sep- arate but remained comparative. The fruit of the decades of patient work in comparative anatomy was the theory of organic evolution. Its method, (i.e., comparative method) became the method of anthropology when both physical and social-cultural anthropology turned evolutionist. 'The' comparative method itself had become identified with evolution. It would be wrong to assume that its nature was the same as that of the pre- evolutionary biology.

Before Linnaeus the knowledge of the organic world was not well-organized. He made a gallant attempt to work out its 'ranks and order.' He gave us the first organized system of biological classifications. It was descriptive. Moreover, his species concept was static. But his overall system was "based on comparison- exact comparison, and a marshalling according to principles" (Kroeber 1954:275). So thorough was the Linnaean systematics that even Darwinism did not uproot taxonomy or necessitate basic taxonomic rearrangements. It was, rather itself, grounded on that taxonomic work (Mayer et al. 1953). In the pre-Darwin era Cuvier further elaborated the work of classification by his 'pattern' or 'type' concept. In fine, there was a pre-evolutionary comparative method in biology different from the comparative method of the latter half of the nineteenth century.

Marett's contention that 'anthropology is the child of Dar- winism' is rejected outright today. It is debatable if the theory of organic evolution even served as a catalytic agent in crystal- lyzing evolutionary approach in the socio-cultural fields. But the impact of comparative anatomy on the anthropologists of the later nineteenth century was apparent. Tylor spoke of the need to 'dissect' civilization/culture into details and to classify these in their proper groups. He compared these details of culture with 'the species of plants and animals' (Tylor 1877, 1:7-8). One of the chief goals of the nineteenth century unilinear evolutionists was to arrange human societies or elements thereof in social series. This was done by "comparing races near the same grade of civilization. Little regard need be had in such comparisons for the date in history or for place on the map" (Tylor 1877, 1:6). Starke (1894) points out that 'the' comparative method was used for ascertaining the relationships between the different

23

Page 6: 3316722

ANTHROPOLOGICAL QUARTERLY

races as well as to discover the definite causes of the origin of myths and institutions. As indicated by Tylor, this method was used for dissecting cultural materials. His famous definition of culture as 'that complex whole' was not utilized properly either by him or by his contemporaries. Cultural traits and complexes were torn out of context to be placed in a hierarchical evolu- tionary order. Tylor considered 'survivals' as a vital point in his ethnographic research. These, we are told, are processes, customs, etc., which are examples of an older condition of culture out of which a newer one may have evolved. 'By force of habit' they come i .to a new state of society although they originated else- where. 'The' comparative method was to extract and pin-point the supposed 'survivals' in modern societies all over the world. These were then to be placed into stages of sequences so as to lead to the origin. If need be, the connecting links were made up by the anthropologist's imagination.

At the turn of the century attacks on evolution-both biolog- ical and socio-cultural-became rampant. The former was based on the classification of animals and plants which the natural historians had worked upon for at least two centuries. No such systematic classification of peoples and their customs and insti- tutions existed. By the early second decade of the present century evolutionism had ceased to be a living force in social-cultural anthropology. 'The' (evolutionist) comparative method also fell into disrepute. But it is interesting to note that the so-called his- torical-diffusionist schools, in fact, did not stop comparing, al- though for purposes different from those of the evolutionists. In the heliolithic as well as in the German-Austrian schools com- parison was the basis for establishing relations to study the growth of culture through diffusion. Graebner's Ferninterpretation, i.e., the interpretation of borrowing despite distance, disregards spacio-temporal factors as did the evolutionists before him. The criteria of 'form' and 'quantity' were not to be limited to con- tiguous regions. Graebner used them for establishing far-reaching connections between the Old World and the New World cul- tures. His Kulturkreis concept was a cultural type or a block of cultural material. Koppers (1955) and Heine-Gelder (1954) find it difficult to justify Kulturkreise today. But both of tiem use Graebnerian comparative method for the attainment of those very goals. Further, the concept of survivals was retained by

24

Page 7: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

early diffusionists. Rivers was using the 'survivals' of kinship terminology among the primitive peoples to reconstruct their past. Father Schmidt used the so-called 'survival' among the

contemporary primitives to trace the 'origins' of religion. Like the unilinear evolutionists the diffusionists also failed to see cul- ture as integrated parts of life. Only recently some among them like Haekel have made pleas to fall in line with the prevailing trend in anthropology (Fiirer-Haimendorf, 1955).

Ackerknecht accuses functionalists of disregarding comparative method completely. He specially mentions Malinowski in this connection. The latter's monographs are certainly not essays in comparison. They are not required to be so. In such cases only re-studies can be comparative in their nature. After his Argonauts (1922) Malinowski remained primarily engaged in

writing his field material. But his attempt to compare the oedipus complex in the primitive and in the modern societies indicates that he was not opposed to a comparative method. In fact, Mali- nowski could in no case oppose a comparative method of the legit- imate variety. He was one of the first authors of a comparative method which later became the keynote of what Radcliffe-Brown called the method of 'structural' or 'sociological' analysis. In 1913 Malinowski published a book entitled The Family among the Australian Aborigines. It was exclusively based on library work. Its conclusions were not in line with the Frazerian comparative approach. It differed also from the then diffusionist approaches. Here Malinowski anticipated Radcliffe-Brown's life-work. The latter hailed the book as a landmark of the legitimate scientific

approach in anthropology. That very year Radcliffe-Brown pub- lished his analysis of the Kareira system in his famous essay, "Three Tribes of Western Australia." It is true that after his Trobriand visit Malinowski did not do any large-scale com-

parative work. It is difficult to say how far it was due to the

fact, as asserted by the French sociologists, that two similar cus- toms may have different functions in the societies in which they exist, and are therefore not properly comparable (Radcliffe- Brown 1958:162). At the same time, it is since the advent of functionalism that field work has yielded more reliable data. The

opportunities of comparison have also increased (Fortes 1953). Radcliffe-Brown advocated a process of 'abstractive generaliza- tion' based on analysis and comparison (1950).

25

Page 8: 3316722

ANTHROPOLOGICAL QUARTERLY

Eggan (1954) feels that 'the' comparative method has been more continuously utilized in British than in American anthro-

pology. Franz Boas pointed out 'the limitations of the compara- tive method' as early as 1896. He proposed his historical method instead. Eggan says that 'the' comparative method has long been in disrepute in the U.S.A. It is difficult to agree with him if he means to say that a comparative approach had no place in the anthropology of Boas and his students. Boas himself advo- cated a modified use of comparison. The historical method "duly recognizes the results obtained by comparative studies" (Boas 1940:277). He refused to accept that the occurrence of the same

phenomena is always due to the same causes (1940:275). He

recognized the existence of laws governing human cultural

growth. He also aimed at discovering the processes of the de-

velopment of culture. But to begin with he wanted to restrict his attention to a well-defined, small geographical, territory. A detailed study of the customs in relation to the total culture of the tribe concerned is the first step. Then, one should investigate their distribution among the neighboring tribes. Boas very specifically mentions that "its comparisons are not extended be- yond the limits of the cultural area that forms the basis of the study" (1940:277) [Italics mine]. It may be noted that Boas's historical method is itself a comparative method. It aims at work-

ing out 'histories of the cultures of diverse tribes.' The general laws, whose existence Boas did not deny altogether, were to be formulated with "a thorough comparison of the manner they became manifest in different cultures." In other words, it is by comparing histories of growth that general laws may be found

(Boas 1940:279). As early as 1920 Lowie attempted a com-

parative approach in his Primitive Society. He clearly went be-

yond Boas's prescriptions by not restricting himself to one geo- graphical region. He did not treat 'history' either. He was interested in showing the shallowness of some of the claims of the unilinear evolutionists. His method of 'text-book' comparison apparently served his end well. Another student of Boas, Ruth Benedict, wrote her famous Patterns of Culture in 1934. That was an essay in comparison. It had a more set and positive goal than Lowie's book. Much water from functionalism had flowed under the bridge of anthropology in the preceding decade to

bring about this difference.

26

Page 9: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

Radcliffe-Brown had an unshakable faith in comparative method. He started using his method of 'sociological analysis' in 1913. He became a functionalist later. But he did not abandon his comparative studies. This is unlike Malinowski. Eggan's attribution of a continuous use of 'the' comparative method to British anthropology stems to a very great extent from Radcliffe- Brown's efforts in this direction, particularly since the 1920's. The latter traces the 'legitimate' use of 'the' comparative method back to Frazer's works in the last decade of the last century (Rad- cliffe-Brown 1951). After Malinowski's The Family (1913) the most important work on comparative lines was Radcliffe-Brown's Oceania Monograph (1930-31) entitled "The Social Organiza- tion of Australian Tribes." Ever since, it has served as a model of comparative work, planned and limited in scope. Besides this regional or geographical-area comparison, Radcliffe-Brown wrote several essays in 'topical' or institutional comparison. His Huxley Memorial Lecture (1951), besides its methodological dimension, is of the second variety. Fortes's and Evans-Pritchard's (1940) African Political Systems and Radcliffe-Brown's and Forde's

(1950) African Systems of Kinship and Marriage are generally regarded as examples of regional comparison of two different sets of institution. To my mind they remain merely compilations of essays although general 'Introductions' by editor(s) have a comparative tinge. Schapera hails the first work as a model and is critical of the latter. But the members of the Symposium, Mil- ton Singer reports, did not join Schapera in his exhortation of Fortes's and Evans-Pritchard's work. It is apparent that under Radcliffe-Brown's patronage 'the' comparative method, or its variant(s), was used to discover regularities or 'laws.' But he never doubted that comparisons could aim at reconstruction of history too. How far the two aims can be combined into one is yet to be shown conclusively. This remains so despite Murdock's tribute to Eggan (and the latter's acceptance of that) that he worked out a highly productive and creative synthesis.

The above appraisal does not warrant us joining'with Acker- knecht in applauding the appearance of G. P. Murdock's Social Structure (1949) as the sign of 'a renaissance of the comparative method' (Ackerknecht 1954:117). Kroeber had a much clearer vision in remarking: "I would say that it ['the' comparative

27

Page 10: 3316722

ANTHROPOLOGICAL QUARTERLY

method] has never gone out; it has only changed its tactic" (1954:273). It is true that the late forties and the early fifties of the current century saw a spate of theoretical appraisals of the comparative approaches in social-cultural anthropology. Mur- dock's attempt is distinctive because he emphasizes the necessity of sampling and uses statistical techniques for the analysis of his data. This amounts to the introduction of a new mode of com-

parison in anthropology. Murdock deserves credit for it. But most British anthropologists, and many others too, are very skeptical of the claims of the statistical comparative method. It is not possible to accept that it is 'the' comparative method.

Schapera has voiced concern and has pointed out several short-

comings of Murdock's approach (1953). Schapera also holds that far-flung and/or global comparisons are not of much value. Rather, he would prefer intensive regional comparisons to begin with. He would, thus, be able to avoid problems of sampling- at least he thinks so. But he does not convincingly show why comparisons outside a region will not yield useful results. Eggan's controlled comparison is very similar to Schapera's intensive reg- ional comparison. Unlike the latter Eggan does not doubt the usefulness or validity of wide-range comparisons. But in order that they may be controlled, he would like the comparisons to be limited in their scope, both geographically and theoretically. Save Murdock's statistical comparison the other two approaches mentioned above fall in the same category as Radcliffe-Brown's 1930-31 study, which may well have been their source of in-

spiration. Comparative method has a very important place in anthro-

pology. But what are its limits? Is it the sole method of discovery as well as of proof? Can useful hypotheses emerge from non-

comparative studies also? Redfield and Julian Steward (Singer 1953) hold that the last one is possible. It seems quite reasonable. A similar situation arose when Dr. Whewell, in his Philosophy of Discovery, lashed out against Mill's methods. The most dif- ficult thing, said Whewell, was to reduce the phenomena to the

type of formulae which Mill has so easily taken for granted. He doubted if anything could be discovered or proved by Mill's Methods. Mill argued that wherever observation and experiment were used for making discoveries, his four methods were methods of discovery. It is worth noting that Mill did not say they were

28

Page 11: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

the only methods of discovery. Probably he was not very sure on this point. He specifically states that "even if they were not methods of discovery, it would not be the less true that they are the sole methods of proof" (1911:284). The same holds good for the methods of comparison in social-cultural anthropology.

Radcliffe-Brown considered classification a most vital task. To- day he is being accused of 'butterfly collecting.' It should not be forgotten that a similar butterfly collection by Linnaeus, Cuvier, and others-at what Kroeber calls the phenomenological-class- ificatory level-gave a solid foundation to the theory of organic evolution. With the latter biology entered the dynamic processual level of 'genetic-historical implications' (Kroeber 1954; 274 ff.). There was no solid cumulative work of 'assemblage, analysis, comparsion, and classificatory organization of knowledge' about the forms of social life. The attempts of our nineteenth century forebears to work an impeccable social evolution theory was sure to end in a failure. Their discipline did not have a base similar to that of biology. Even today, though a century has elapsed since then, we are without any systematic classification. Both Kroeber and Radcliffe-Brown were aware of this deficiency. The former has very aptly remarked that the fully flourishing his- torical science is still ahead of us and "comparison-the con- textual comparative method, I should like to call it-is neces-

sarily involved as a foundation" (1954:281).

According to Hoenigswald (1963) scientific linguistics began around 1800. Fourteen years before, in 1786, Sir William Jones made his famous observation. He saw more than accidental af- finity between Sanskrit, and Greek and Latin and suspected that

they had some extinct common ancestor. But Jones did not him- self give lead in accomplishing what he preached. Friedrich

Schlegel opined that Sanskrit was itself the ancestor of the Indo-

European languages. Hoenigswald does not subscribe to the view that Jones observed and Schlegel violated "a comparative method that was still far in the future" (1963:4) [Italics mine]. This 'a Comparative Method' has been given the status of 'the

Comparative Method' of linguistics. It is hard to understand why Schlegel's approach is denied the credit of having introduced a comparative point of view in linguistics. 'Comparative gram- mar' was first conceived by him. For working out the 'language genealogies,' the grammatical features of a group of languages

29

Page 12: 3316722

ANTHROPOLOGICAL QUARTERLY

should be examined, said Schlegel. The archaic ones are retained because they are present by definition in the ancestor language. The latter is the language having the phonetic shape on which the archaic features of a number of related languages agree. The

premises of this approach are not correct; nevertheless, it is based on a type of comparative method. The nineteenth century unilinear evolutionists similarly confused the (contemporary) primitives with the primordial man. But we do not-in fact, cannot-deny that they used a comparative method. Moreover, Cuvier's 'comparative' approach had great influence not only on Schlegel but also on Grimm (Ackerknecht 1954) and on

Bopp (cp. Hoenigswald 1963). In biology 'the' comparative method was non-evolutionist in the days of Cuvier.

Holger Pedersen (1931) regards Rasmus Rask chronologically first among the greatest specialists in comparative linguistics. Ac-

cording to Otto Jespersen, Rask gave the best expression to "the true method of linguistic research written in the first half of the nineteenth century" (1950). He correctly determined and sub- stantiated the relationships between Icelandic and other lang- uages. The proper linguistic method is not to pick up a few details or words for comparison but to examine carefully the total structure of a language. He emphasized the significance of

comparing the morphology of languages. He preceded Grimm in

indicating that sound laws are a proof of relationship between

languages. He also had the remarkable insight to point out that 'the most essential, concrete, indespensable words (are) the foundations of a language.' If two languages show such a fre- quent agreement in these, such that rules regarding 'letters'

(sounds)-shift between the two can be formulated, then the two are related (cp. Waterman 1963). But Rask was primarily a comparative grammarian. For him grammatical agreement was a much more certain indication of original identity or rela-

tionship between languages. His was the first sketch of com-

parative Indo-European grammar. Jacob Grimm was not the first to perceive the sound shifts in

Germanic. His contribution lies in positing the theoretical prin- ciple underlying the apparently discrete sound changes and in

determining their etymological value. In Waterman's words he thus paved the way for the science of historico-comparative linguistics and-in our own day-structural linguistics (1963:

30

Page 13: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

30). It is not fair to call him the founder of historical rather than comparative linguistics (Pedersen 1931). Rask did not under- take any thorough comparisons. Although more limited in scope than the works of Rask and Bopp, Grimm's Deutsche Gram- matik was comparative in nature and was a most needed work in the development of comparative linguistics as a science. Water- man informs us that he came to be regarded as 'the founder of

comparative linguistics.' Grimm was influenced by Cuvier. His

comparative Germanic grammar was on the lines of the then new historico-organic method which influenced the development of what came to be regarded as the comparative method of

linguistics. Franz Bopp was also a comparative grammarian. In contra-

distinction to Rask's and Grimm's interest in phonetic change Bopp concentrated on morphology. The most valuable contribu- tion which he made through his book, Conjugationssystem, was in the field of the inflection of verbs. He broadened the base of

comparative Indo-European grammar by the inclusion in it of Sanskrit. Many regald this as the 'real' or 'true' beginning of

comparative linguistics and consider Bopp as the greatest among its founders (cp. Jespersen 1950). With regard to what consti- tutes linguistic kinship, he could not improve upon Rask. As a Sanskritist he introduced a new technique of analysis. He failed in his attempt to trace the ultimate origin of inflectional elements but in its place discovered Comparative Grammar.

Wilhelm von Humboldt is considered by many as one of the soundest linguistic thinkers of the nineteenth century. To him each language is a picture of 'the original aptitute for language.' He introduced relativism in historical-comparative linguistics by recommending that each separate language be looked upon and treated as an organic whole. He did not approve of deductive 'universal' or 'general' grammar. He believed that grammatical form or 'inner form' was a universal feature. He would like to see an inductive general grammar which showed how the same

grammatical notion was expressed in a number of different

languages. August Friedrich Pott had admiration for von Hum- boldt. But he was on the Rask-Grimm line in showing the in-

creasing importance of phonology in the Indo-European linguistic field. He had a mature understanding of the sound-change. He considered phonology the surest and the most important key to

31

Page 14: 3316722

ANTHROPOLOGICAL QUARTERLY

etymology. He attempted to prepare an etymological dictionary which even now forms "the warp and woof of the Indo-

European 'comparative' linguistics" (Hoenigswald 1963). With August Schleicher the first or the older period of Com-

parative Linguistics ended and a transition to a new one started.

Hoenigswald says that his Compendium marked a step from

'comparative grammar' to 'the Comparative Method' (1963). This has a special meaning. According to Hoenigswald these

early grammarians' main interest was in grammatical structure. Their approach was typological but may be called 'comparative' as well. In spite of it, we are told, there was "neither comparative 'linguistics' nor the Comparative 'Method' " as that time (Hoen- igswald 1963:5). Most linguists today will probably accept the second part of the above contention, but not the first part. To us neither of them seems correct. For instance, Waterman points out that as early as 1814, Rask formulated "certain basic prin- ciples and methods of modern comparative linguistics" (1963: 18). This statement is almost the exact opposite of Hoenigswald's.

In fact, the comparative grammarians presented their own ver- sion of a comparative method for studying the grammatical structure of the Indo-European. They erred at many places. It is interesting that Hoenigswald himself says that in a certain sense some of the concepts of the early grammarians have come back today (1963: 10). It does not mean that comparative linguistics has ceased to exist because of this resurgence. As we shall see later, Schleicher's greatest merit lies in presenting, for the first time, a proper Comparative Method for reconstructing the ancestor language. Today, while it is possible to use some technical aspects of his comparative method (of reconstruction) -which has been subsequently considerably elaborated (Hoe- nigswald 1960)-hardly any serious modern linguist will accept his 'life-cycle' interpretation of the growth of language. Nobody today will venture to write-or will even think of writting- a fairy tale in the reconstructed ancestral Indo-European as Schleicher did. These points clearly indicate the difference in the aim, the scope and the purpose of the use of a method of reconstruction. There is hardly any justification now for calling Schleicher's comparative method 'the' comparative method of

linguistics. It is also difficult to share his concept of linguistics as a natural science. In spite of his magnificent methodological

32

Page 15: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

contributions he held that language evolved only in prehistoric times (Pedersen 1931).

Schleicher made three memorable contributions: classification of languages into types, a theory of language relationships and a 'comparative method' for reconstructing a parent language. He worked out both the phonological system as well as the gram- matical forms and the individual words of the parent speech. He demonstrated conclusively that none of the known Indo-Euro-

pean languages was identical with the parent PIE. But he would not consider his reconstructions as formulae only. He believed that the ancestral Indo-European was the 'original' and un-

decayed language whose words were undamaged. 'The' com-

parative method works 'as if the assumption were valid,' which, unfortunately, is not the case. It points to an ancestral language free from dialect variation. Also, 'the' method in itself does not

help us in inferring historical relationship, because the similarities between two languages may be due to mutual influence rather than mutual development. As Pedersen very aptly put it "the new method (i.e., 'the' comparative method) was far from per- fect in the hands of its founder" (1931:270).

A new trend of apathy toward Schleicherian comparative method began in the seventies of the last century. For some time the controversy about the 'wave theory' was prominent. Many contributions regarding the irregularity in the Indo-European sound laws were made in that period. Interest in a thorough study of the life and development in language began. The mod- ern languages and phonetics, or, the description of the sounds of language, became the focus of study. In the case of the latter the developments were different from Grimm's notion of sound

symbolism. August Leskian put forward the extreme neogram- marian view that "sound laws have no exception" and sparked off an interesting controversy. Whatever their mistakes may have been, the neogrammarians advocated a disciplined, rigorous and controllable methodology. They were for dealing only with the

physical phenomena of language (Waterman 1963). Their rigour implied that 'the' (Schleicherian) comparative method could be applied only in the case of certain language families, such as the Indo-European. Hoenigswald accepts that 'the' com-

parative method is not the only avenue to reconstruction known to modern linguistics. He also acknowledges that "there is a

33

Page 16: 3316722

ANTHROPOLOGICAL QUARTERLY

reality at the bottom of this curious (neogrammarian) notion" because in the reconstruction of the I-E, internal reconstruction probably played a more significant part than is normally ad- mitted (Hoenigswald 1963:9).

Ferdinand de Saussure thinks that 'the comparative school' opened up a new and useful field but did not establish 'the true science of linguistics.' Its exponents failed to seek out the nature of the object of linguistic studies. They never asked themselves "the meanings of their comparisons or the significance of the relations that they discovered" (de Saussure 1959:3-4). For any historical reconstruction, we are told, comparison is required. But this by itself does not establish the historicity. The method of 'the comparative school' was not historical. It was 'exclusively comparative'!

Further elaborations of 'the' comparative method of recon- struction came from a student of de Saussure, Antoine Meillet. He was primarily responsible for expanding the modern prin- ciple of linguistic comparison so that historical relationship be- tween different languages may be established. It was indicated that by a comparison of the forms of historically related lan- guages some idea of their parent language may be had. But this will only be a partial reconstruction. It is not possible to recon- struct a complete linguistic system. Meillet held that gram- matical forms were rarely borrowed. But even considerable identities in vocabulary is no proof of historical kinship. Par- ticularly in case of two languages of 'isolating' type, it would be almost impossible to prove historical relationship between them (Sommerfelt 1961:292). Meillet further pointed out that 'a Comparative Method' should deal not only with 'the phonetic shape of morphs.' Rather, it should be primarily applied to morphology and syntax. He gave a major methodological lead by insisting that a comparative method, in the historical per- spective, should be concerned with working out correspondences between dissimilar elements rather than with an exclusive occu- pation with similarities. This was a very important perspective and had revolutionary effects on the modern concept of 'the' comparative method (of reconstruction).

With the turn of the century a great change in emphasis took place in linguistics. The heyday of the historical-comparative linguistics was over. Linguistics moved towards descriptive and

34

Page 17: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

structural studies. Ferdinand de Saussure in Europe and Franz Boas in the U.S.A. affected this course of development. A similar change-over took place in social-cultural anthropology as well. It came in the wake of Boas's Jessup North Pacific Expedition and Haddon's Torres Straits Expedition and with an emphasis on field work. But the rules of contemporary anthropological field work were broadly outlined by Malinowski in the twenties. There is one important difference between linguistics and anthropology. The former did not experience any impact like Malinowskian functionalism. Many anthropologists today talk of structural

approach. They do not have anything to do with functionalism. In fact, it only seems so. They have incorporated the key points of functionalism, of course, minus its biological determinism, which Radcliffe-Brown shared in common with Malinowski.

Only in this sense 'structuralism' without functionalism is pos- sible in anthropology, but in linguistics it has been really so. Some linguists may say that de Saussure thought in functional terms. Probably Sapir and Whorf may also be credited likewise.

Except for some recent attempts like Chomsky's, visible marks of a long and reputable place for the concept of functional inter- relation of language are hard to find. The concept of culture or social structure as an integral whole is basic to anthropological theorizing today. But we wonder if linguists always treat a

language as a unit entity with interrelated parts (e.g., grammar, syntax, vocabulary, etc.), constituting a system, even though they may implicitly believe in it.

Leonard Bloomfield has a unique place in twentieth century linguistics. By training he had some germs of historico-compar- ative linguistics in him. One can trace a direct linkage starting with Grimm to neogrammarians to Bloomfield (Waterman 1963). He regarded the principle of the regularity of sound-

change as the turning point in the history of linguistics. Although he sympathized with the neogrammarian plea of no 'exceptions,' he felt that that was a very imprecise way of putting things. In modern terminology it will mean that "conditioned sound

changes are purely phonetic" and "independent of non-phonetic factors, such as meaning, frequency, harmony, or what not, of

any particular linguistic form." According to Bernard Bloch, "Bloomfield's greatest contribution to the study of language was to make a science of it" (quoted in Fries 1961:199). He had

35

Page 18: 3316722

ANTHROPOLOGICAL QUARTERLY

no place for teleological 'explanations' of the 'residual forms' in his system. This is similar to the functionalist rejection of 'survivals' as an excuse for empirical explanation in social- cultural anthropology. Bloomfield acknowledges his debt to Boas for the interest in and the development of descriptive language study (Fries 1961:218).

Bloomfield regretted that great linguists like Herman Paul and A. Leskien neglected descriptive study of language. He re-

gards Leskien as one who laid the foundations of 'historical methods of research.' He credits F. Miiller, Fink and de Saussure with ha- ing the vision to perceive a natural relation between

descriptive and historical studies in linguistics. The necessity of

descriptive data as a prerequisite for comparative work became

apparent in the so-called historical-comparative treatment of the

languages belonging to families other than the Indo-European. This Bloomfield considered the merging of the descriptive studies with the main-spring of historical work. It certainly amounts to the broadening of the horizon of comparative work in linguistics with such aims, objects and procedures which were beyond the

comprehension of the father of 'the' comparative method (of re-

construction), i.e., Schleicher. Bloomfield himself points out that all historical study of language is based upon the comparison of two or more sets of descriptive data. Our power to make com-

parisons is dependent upon our knowledge of the things to be

compared. These experiences are derived from the comparative work done with non-Indo-European languages where the merg- ing of 'the historical-comparative and the philosophical-descrip- tive', as Bloomfield put it, was fruitfully demonstrated. All valid

generalizations about language can only be inductive generaliza- tions. He is not sure if language 'universals' are really universal. But he is not averse to their search. He hoped that some day linguistics would return to 'the problem of general grammar to

explain these similarities and divergences.' When such a study is undertaken it "will not be speculative, but inductive" (Bloom- field 1933:20). Chomsky has vindicated this prophesy.

Nineteenth century linguistics was a humanistic discipline with a background of classical scholarship. Linguistics, then, aimed at reconstructing the parent language. In those days social-cultural anthropology was interested in working out the

stages of development of human culture and in searching for

36

Page 19: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY

origins. The early grammarians regarded Sanskrit as the an- cestral Indo-European. Schleicher rectified this mistake but he did not go far enough. In his view the parent language was free from 'decay' and its words were still 'undamaged in all parts.' Moreover, Sanskrit reflected to him "a stage of linguistic development only slightly removed in excellence and purity from the mother tongue" (Waterman 1963:41). In anthropology the contemporary primitives were equated with the primordial man and the search for 'survivals' was intensified. All this was said to be done with and through 'the' comparative method in linguistics as well as in anthropology. But in the former its use continued even after a definite shift from historical to descriptive linguistics took place, at least in the relative emphasis, at the turn of the century. In anthropology the 'historical' reconstruc- tions were being abandoned and so 'the' comparative method fell into disrepute.

Today linguistics has reached a stage of development when it will find it hard to justify why the term "comparative method" be used only for "specifically a kind of comparison of related languages for the purpose of determining . . . their common ancestor" (Hockett 1958:469). There may have been some validity in calling this 'the' Comparative Method in the days of Schleicher. Though legitimate, the reconstruction of a parent language is a small subdivision of modern linguistics. Moreover, this is not the only part of linguistics where the logic of com- parison is applied. The experience of the biological sciences and that of anthropology shows that comparative method can be successfully and legitimately used for static (synchronic) as well as for dynamic (diachronic) purposes. Further, the so-called 'the' comparative method is not the sole method of reconstruction in modern linguistics. Both Hockett and Hoenigswald mention the method of internal reconstruction, besides 'the' comparative method, for linguistic prehistory. Hockett further accepts that there is 'comparison' outside the field of "comparative method," i.e., "in the sense of typological study of classification." He agrees that the term comparison does not mean the same thing outside linguistics that it means to a linguist. He opines that some other term has to be used for these other types of 'comparison.' I am afraid that this is not going to the root of the matter, but rather

37

Page 20: 3316722

38 ANTHROPOLOGICAL QUARTERLY

avoiding it. The fact is that there is more than one comparative method in linguistics as in social-cultural anthropology. The most logical solution is that 'the Comparative Method' be re- christened the method of external reconstruction in contradis- tinction to that of internal reconstruction. The latter, according to Hoenigswald, is 'non-comparative' (1963:9). In case this suggestion seems too drastic, then 'the comparative method of reconstruction' or 'the comparative method of (external) recon- struction' are alternatives to choose from. Linguistics was wise not to throw out the baby with bath-water. But now there is more than one baby (i.e., comparative method) to be dealt with. This fact needs to be duly recognized and taken care of.

NOTE

I wish to express my gratitude to Dr. A. Richard Diebold, Jr., for his encouragement, criticism and comments.

REFERENCES CITED

ACKERKNECHT, EDWIN H. 1954-On the comparative method in anthropology. In Method and

perspective in anthropology, Robert F. Spencer, ed. Minneapolis, The University of Minnesota Press.

BLOOMFIELD, LEONARD 1933-Language. New York, Henry Holt and Co.

BOAS, FRANZ 1940-Race, language and culture. New York, The Macmillan Co.

DE SAUSSURE, FERDINAND 1959-Course in general linguistics. New York, Philosophical Library.

DURKHEIM, EMILE 1938-The rules of sociological method. Illinois, The Free Press.

EGGAN, FRED 1954-Social anthropology and the method of controlled comparison.

American Anthropologist 56:743-63. FORTES, MEYER

1953-The structure of unilinear descent groups. American Anthro- pologist 55:17-41.

FORTES, M. AND E. E. EVANS-PRITCHARD, EDS. 1940-African political systems. Oxford, Oxford University Press.

FRIES, CHARLES C. 1961-The Bloomfield school. In Trends in European and American

linguistics 1930-1960, Christine Mohrmann et al, eds. Utrecht (Holland), Spectrum Publishers.

FiURER-HAIMENDORF, C. VON 1955-Culture history and cultural development. Ir Yearbook of

anthropology, 1955, W. L. Thomas, Jr., ed. New York, Wenner- Gren Foundation.

HEINE-GELDERN, ROBERT VON 1954-Die asiatische Herkunft der sudamerikanischen Mettalltechnik

(a 42-page abstract in English prepared by Dr. P. Phillips of Peabody Museum, Harvard University).

Page 21: 3316722

SOCIAL-CULTURAL ANTHROPOLOGY 39

HERSKOVITS, M. J. 1954-Some problems of method in ethnography. In Method and

perspective in anthropology, Robert F. Spencer, ed. Minneapolis, The University of Minnesota Press.

HOCKETT, CHARLES F. 1958-A course in modern linguistics. New York, The Macmillan Co.

HOENIGSWALD, H. M. 1960-Language change and linguistic reconstruction. Chicago, The

University of Chicago Press. 1963-On the history of the comparative method. Anthropological

Linguistics 5:1-11. JESPERSEN, OTTO

1950-Language. London, George Allen and Unwin Ltd. KOPPERS, WILHELM

1955-Diffusion: transmission and acceptance. In Yearbook of anthro- pology, 1955, W. L. Thomas, Jr., ed. New York, Wenner-Gren Foundation.

KROEBER, A. L. 1954-Critical summary and commentary. In Method and perspective

in anthropology, Robert F. Spencer, ed. Minneapolis, The Uni- versity of Minnesota Press.

LAZARSFELD, PAUL AND MORRIS ROSENBERG, EDS. 1957-The language of social research. Illinois, The Free Press.

LEWIS, OSCAR 1955-Comparisons in cultural anthropology. In Yearbook of anthro-

pology, 1955, W. L. Thomas, Jr., ed. New York, Wenner-Gren Foundation.

MALINOWSKI, B. 1922-Argonauts of the western Pacific. London, George Rouledge

& Sons Ltd. 1963-The family among the Australian aborigines (with an 'Intro-

duction' by J. A. Barnes). New York, Schoken Books [originally published in 1913].

MAYR, ERNST., ET AL 1953-Methods and principles of systematic zoology. New York,

McGraw-Hill Book Co. Ltd. MILL, J. S.

1911-A system of logic. London, Longmans Green and Co. MURDOCK, G. P.

1949-Social structure. New York, The Macmillan Co. MURDOCK, G. P., ET AL

1950-Outlines of cultural materials, (2nd edition). New Haven, Yale University Press.

NADEL, S. F. 1951-Foundations of social anthropology. Illinois, The Free Press.

PARSONS, TALCOTT 1937-The structure of social action. Illinois, The Free Press.

RADCLIFFE-BROWN, A. R. 1913-Three tribes of western Australia. Journal of the Royal Anthro-

pological Institute 43:143-94. 1930-31-The social organization of Australian tribes. Oceania 1:34-63,

206-46, 322-41 and 426-56. 1951-The comparative method in anthropology. Journal of the Royal

Anthropological Institute 81:15-22. 1958-Method in social anthropology. Chicago, The University of Chi-

cago Press. RADCLIFFE-BROWN, A. R. AND DARYLL FORDE, EDS.

1950-African systems of kinship and marriage. Oxford, Oxford Uni- versity Press.

Page 22: 3316722

40 ANTHROPOLOGICAL QUARTERLY

SCHAPERA, I. 1953-Some comments on comparative method in social anthropology.

American Anthropologist 55:353-63. SINGER, MILTON

1953-Summary of comments and discussion (of I. Schapera's above). American Anthropologist 55:363-66.

SOMMERFELT, ALF 1961-The French school of linguistics. In Trends in European and

American linguistics 1930-1960, Christine Mohrrnann et al, eds. Utrecht (Holland), Spectrum Publishers.

STARKE, C. N. 1884-The primite family. New York, D. Appleton and Co.

TYLOR, E. B. 1877-Primitive culture. New York, Harper and Brothers.

WATERMAN, JOHN T. 1963-Perspectives in linguistics. Phoenix Books (Chicago), The Uni-

versity of Chicago Press.