HISTORICAL INFERENCES FROM ETHNOHISTORICAL DATA: BOASIAN VIEWS

KENNETH W . PAYNE A N D STEPHEN 0. M U R R A Y

Just before Boasian anthropologists turned from trying to reconstruct North American prehistory to synchronic work on social organization and on the psy- chological integration of intact cultures, they considered the validity of traditional/folk accounts. Robert Lowie, editor of the official journal of the American Anthropological Association, categorically rejected any use of such data, twisting the carefull? delimited claims made for them by the anthropologists engaged in disen- tangling American ethnic groupings, who were unwilling to throw out all data except those provided by professionally trained (that is, Boasian) anthropologists. Rather than the "refutation" (in the Popperian sense) this episode has been taken to repre- sent, it exemplifies a problematic along with the appropriate methods for the problematic being dropped as other problematics and methods become fashionable.

Particularly under the stimulus of Jan Vansina's oeuvre,' the possibility of using oral traditions to draw historical inferences-earlier rejected out of hand by func- tionalist2 and structuralist anthropologistss-has been reopened in the last decade or so. In a recent collection of essays assessing this predominantly Africanist work,' there is only one mention that the first generation of professional Americanist anthropologists considered the validity of inferring history from ethnohistorical data, and that mention misses the major advocate^.^

THE CULTURE ELEMENTS PARADIGM Franz Boas, the prime mover both in the institutionalization of American

anthropology and in the revolution overthrowing the paradigm of nineteenth-century unilinear evolution, viewed anthropology as

history, pursued in rather positivistic terms through the study of the distribution of elements . . . to reconstruct the history of human variability . . . completing the Powellian program of basic ethnographic description and mapping of the North American continent.'

I n the first decades of this century he directed his students to chart the geographical dis- tribution of institutions and material culture to infer patterns of migration of people and diffusion of parts of culture.' Gathering data and refuting theories were more congenial to Boas than using the data for the purposes for which they were ostensibly gathered, and when his former students began to draw inferences about prehistory, Boas did not sup- port their efforts and instead shifted to another kind of particularistic study of single cultures, their psychic integration and reproduction.* However, those already pursuing

K t N N t T H PAYNE earned a Ph.D. i n anrhropologj~ from /he Universiry of Calt$mtia. Berkeley, and is crtrrenrli ii visiring projessor in the Deparrmenr of Anthropology ar Washingion Universiiy. Sainr I.o~ti.s, M u . 63/30, Research on Bagoho rrhnohisrory srimulared his inreresf in rhepioneer Boa- sion ethnographer Lauro Warson Benedicr

STtPHth M U R R A Y earned a Ph. D. i n sociologj, Jroni rhe Universily oJToronro and has been par- ricularlj. inrererrrd in rhe nerworh of'socia1scienri.rr.s prominenr in rhejirsr third o f t h e rk'enrierh cen- ruri'. HIT puhlicarions include Group Formation in Social Science and numerous arricles, including OW preLious otie\ i n JHBS. Currenrly afiiliared wirh San Francisco C'ouniy Menral Healrh Services. arid rhc. 6'hrrgtiri ln.srirure, 1360 De Haro. Sari Francisco, C a l g 94 /07 . he was mpporred bl; a posr- docrorul fellor ship ,froni rhe ,Yalional Insrirurr of Menial Healrh while wiring [his arricle.

335

336 K E N N E T H W . P A Y N E A N D STEPHEN 0. MURRAY

the Boasian “normal science” in which they had been trained at the turn of the century continued to try to solve the kinds of historical problems abandoned by Boas and never considered by his later students.

John Reed Swanton and Roland Burrage Dixon were the first Boasians to affirm some historical value in folk traditions. Both received Ph.D.s in 1900 from Harvard, where Boas’s’ and Alfred L. Kroeber’s’’ patron, the archeologist Frederick Ward Put- nam, had established a center of anthropological instruction.” Swanton had studied with Boas at Columbia and would later aver, “Whatever I have done is due to the inspiration of our teacher Dixon, like Boas, was a veteran of the Jesup Pacific North Coast expedition and was interested in indigenous languages.” As Kroeber noted in Dixon’s obituary: “Almost alone among their major contemporaries, he and Swanton main- tained a sane and constructive interest in tribal and ethnic migrations.”“ This was the in- terest that motivated Swanton and Dixon at least to consider whether folk traditions might contain grains of history. I n their 1914 survey of the continent’s prehistory, they did not recommend uncritical acceptance of such traditions as offering transparent history; instead, they rather cautiously suggested,

I n investigating still existing peoples like the American Indian we can appeal in the first place to their traditions which, although sometimes noncommital and fre- quently misleading, gain weight when recorded by several different persons and when taken in connection with other data.16

This less-than-wholesale endorsement was too much for the Machian Robert Lowie, who was to succeed Swanton as editor of the American Anthropologist in 1923, and was in the late teens its book review editor. After claiming, “We are not concerned with the abstract possibility of tradition preserving a knowledge of events; we want to know what historical conclusions may safely be drawn from given oral traditions in ethnological practice,”18 Lowie appealed to E. Sidney Hartland’s exemplification of sound practice17 and then produced nothing other than a conclusion about “abstract possibility”:

I cannot attach to oral tradition any historical value whatsoever under any condition whatsoever. We cannot know them to be true except on the basis of extraneous evidence, and in that case they are superfluous since the linguistic, ethnological, or archeological data suffice to establish the conclusion in question. . . . From the traditions themselves nothing can be deduced.’*

Dixon warned, “Absolutely unqualified statements like that of Dr. Lowie’s are usually danger~us,”’~ while Swanton defended folk traditions as a valid basis for inference about the direction of migration (not cosmogenesis). For Swanton, an indicator that cor- roborated it in nine cases out of ten when other evidence was available could therefore be relied upon with some confidence in cases for which no other evidence was available.20 In Swanton’s view some data were better than no data, but for Lowie, “It is our duty to doubt till the facts compel us to affirm.”21

Boas’s Columbia colleague Alexander Goldenweiser, irritated by Lowie’s attempted reductio ad absurdum in shifting from accepting reports of the direction of migration to cosmogenesis myths, joined the fray in 1915, noting:

Dr. Lowie does not strengthen his case by citing creation myths as proof of the deficient historical sense of the Indians. Commonly enough, the Indians themselves distinguish between a myth and a historical tradition. But even were that not so, who would doubt the word of a woman who tells of having witnessed a child being run

IiISTORICAL I N F E R E N C E S FROM tTHNOHlSTORlCAL DATA 337

over by a street car, solely on the ground of his knowledge that the woman believes in ghosts?22

and providing some less colorful analogies: Poor evidence is poor evidence, but it is evidence, and the extent to which such evidence can be trusted is determined by the probability of its being true evidence, which again may be estimated from the frequency of agreement between such evidence of an intrinsically higher merit. Just as the physician is guided in his diagnosis of a disease by vague and doubtful symptoms until a positive one is forthcoming, just as the detective follows illusive and contradictory clues before es- tablishing convincing proof of the crime, so the ethnologist in the absence of better evidence follows the lead of tradition until data of higher evidential value serve to confirm or refute his preliminary conjectures or hypo these^.'^

I n 1916 Edward Sapir, another member of what Lowie considered the Boasian “super- intelligent~ia,”~‘ laid out the rationale for the age and area hypothe~is.’~ Rebelling somewhat against Boas’s and Lowie’s “methodological asceticism,”’“ he looked at data from sources Boas would have preferred to eliminate,27 and credited folk traditions with correctly indicating not only the direction of migration but also the source (that is, the tribe) from which a trait was borrowed.28

Lowie did not want to review Sapir’s book, but having failed to get either Boas or Kroeber to do it,” undertook the task himself. When he reached Sapir’s discussion of “native testimony,” Lowie produced the odd introduction, “Dr. Sapir’s position with reference to certain moot questions is of i n t e r e ~ t . ” ~ ~ Considering that Lowie’s position had recently been attacked in the profession’s core journal by four peers, including its editor, with no one weighing in on his side, “moot question” is a quite bold definition of the situation. Furthermore, one wonders how any answer to a “moot question” could be of interest.

Just as he used creation myths to dismiss delimited inferences about population movement in arguing against Swanton and Dixon, Lowie dealt with Sapir’s argument that borrowing credited to a specific named tribe possessed some plausibility by a reduc- tio ad absurdum of claims about invention:

I f the same ceremonial complex insists that it originated with them, e.g., is shared by the Sarsi, Blackfoot, and Crow and each of these peoples insists that it originated with them, it is difficult to reconcile these statements. The fact that one of them must be correct does not establish the methodological validity of accepting native traditions as history.31

AFTERMATH

Although Lowie’s critiques have come down (in at least the lore of the field) as refutations (in the ful l Popperian sense) of pseudohistorical methods, closer examination of the whole exchange shows that Lowie’s dogmatic condemnation of ethnohistorical data fit with the antihistorical wave of the future in American social science. Swanton’s and Dixon’s line of work was not refuted by Lowie but merely suspended, only recently to be continued several professional generations later. That such data were not much used by American anthropologists in the following decades has less to do with the cogen- cy of Lowie’s criticisms than with a shift from attempts at historical reconstruction to ahistorical work on modal personalities in intact cultures and synchronic fieldwork focused on social ~ r g a n i z a t i o n . ~ ~

338 KENNETH W . PAYNE AND STEPHEN 0. M U R R A Y

After completing his combination of linguistic stocks into pro to fa mi lie^,^^ Sapir was at the forefront of work on “culture and pers~nality,”~‘ and arranged for A. R. Radcliffe-Brown’s Chicago sojourn,36 which was a major stimulus to “social anthropology.” Lowie would eventually write a national character study and a social organization t e ~ t b o o k . ~ ” Yet at the end of his career, he would recollect, “It was the reconstruction of the ancient primitive life that interested me,”3’ despite his part in mak- ing such work seem neither possible nor worth doing.

Staggering under the responsibility of salvaging knowledge about the numerous, genetically diverse California tribes, Kroeber persisted in collecting checklists of culture elements into the 1930s. By statistical analyses of materials gathered by sometimes un- enthusiastic students, Kroeber and Harold Driver sought to draw inferences about diffu- sion and to correlate cultural and environmental areas long after Boas had decreed the study of diffusion ended, and after even Clark Wissler had given it up.38

Competing with his Berkeley colleague Frederick Teggart,38 Kroeber continued historical correlation^,'^ and after World War I 1 championed Morris Swadesh’s develop- ment of a new mechanical discovery procedure for genetic reconstruction, lex- i cos t a t i~ t i c s .~~

Kroeber was quoted above as noting that only Swanton and Dixon maintained an interest in tribal migration. Dixon’s interests shifted from American Indians to Polynesia. Swanton continued careful sifting of whatever could be recovered from ex- plorers’ accounts‘* and native American t radi t ion~.‘~

And, as noted earlier, interest in the possibility of lore containing some valid knowledge (of history, of herbal remedies, and so forth) has revived in recent years,“ as has scrutiny of accounts of vanished or greatly modified cultures made by non- professional (and still suspect) observers, making the methodological discussions of con- temporary rele~ance.‘~

NOTES

I . Jan Vansina, De la tradition orale (Teruvren: Musee Royal de I’Afrique Centrale. 1961); idem, “Once upon a Time,” Daedalus ‘I00 (1971): 442-468. A rather different collation of history and “legend” is proposed and exemplified by Marshall Sahlins in his Historical Metaphors and Mythical Realities (Ann Arbor: Univer- sity of Michigan Press, 1981).

Bronislaw Malinowski, Magic. Science and Religion (Toronto: Doubleday, 1954). p. 166, and George P. Murdock, Africa (New York: McGraw-Hill, 1959), p. 143.

Claude LCvi-Strauss, Structural Anthropology (New York: Basic Books, 1963). p. 281. Kenneth Brown, Using Oral Sources-Social Analysis 4 (1980). For related contexts see John R. Searle,

“The Logical Status of Fictional Discourse,” New Literary History 6 (1975): 319-332, and J . M. Cameron, “A Good Read,” New York Review of Books ( I 5 April 1982): 28-3 I .

Roy Willis, “The Literalist Fallacy and the Problem o f o r a l Traditions,” Social Analysis 4 (1980): 28-37. George W. Stocking, Jr . , “Ideas and Institutions in American Anthropology,” Selected Papers from the

American Anthropologist 2 (1976): 5-6. I t was believed that the center of these scattergrams was the point of origin, the peripheries where the

diffusion had most recently extended (rather than there being impermeable cultural boundaries for a particular trait). The wider the scatter, the older the trait was presumed to be.

Regna Darnell, “Hallowell’s ‘Bear Ceremonialism’ and the Emergence of Boastan Anthropology,” Ethos

George W. Stocking, Jr., Race. Cultureand Evolution (New York: Free Press. 1968), p. 279; Stephen 0.

Timothy H. H . Thoresen, “Paying the Piper and Calling the Tune,” Journal of’ the History of’ the

2.

3. 4.

5. 6.

7.

8.

9. Murray, Group Formation in Social Science (Edmonton: Linguistic Research, 1983). pp. 47-49. 10. Behavioral Sciences 1 I (1975): 257-275.

5 (1977): 13-30.

HISTORICAL INFERENCES FROM ETHNOHISTORICAL DATA 339

I I . 1980). 12. 13. 14. 15.

16. 17. Bushongo.” Journal o/ American Folklore 25 (1914): 428-456. 18. 19. 20. 2 I , ethnological theory: see Murray, Group Formarion, pp. 52-57. 22. 764. 23. Ibid., pp. 763-764. 24. Robert Lowie. Roberr H . Lowre. Erhnologisr (Berkeley: University of California Press, 1959). p. 133. He did not include Dixon or Swanton with Sapir, Radin, Benedict, and Goldenweiser, and there is a t least a hint of general contempt for Swanton in a letter from Lowie to Kroeber dated I3 March 1917: “Swanton’s recent effu- sion adds zero.“ However, see note 42 below 25. See note 7 .

26. Karl Mannheim (in “American Sociology” in Twenrierh Century Sociology. ed. George Gurvitch and W. E. Moore [New York: Philosophical Library, 19451, pp. 503-537) remarked on a general tendency in American social science to eschew any knowledge not derived from replicable methods. Leslie Spier’s review (American Anrhropologisr 31 (1929): 141-143) of Dixon’s The Building of Cultures (New York: Scribner’s, 1928) characterized Dixon as “stoutly conservative in his use of inferential methods” with “an almost Puritanical adherence to evidence and logic.” That such an assessment was generally held is confirmed by Fay Cooper Cole in “Personalities of the Half Century,” American Anrhropologisr 54 (1952): 161.

27. Sapir to Kroeber, I I February 1927. See Regna Darnell, “The Sapir Years a t the National Museum, Ot- tawa,” Proceedings ojrhe Canadian Eihnological Socieiy (1976): 15-25; Stephen 0. Murray, “The Canadian ‘Winter’ of Edward Sapir.” Hisroriographia Linguisiica 8 (I98 I ) : 63-68; idem, “The Creation of Linguistic Structure,” .4rnerican .4nihropologisr 85 (1983): 355-361. 28. Edward Sapir. Tivie Perspecrive in Aboriginal Culrure. [Canadian] Geological Survey Memoir 62 ( I9 16), p. 396. 29. Lowie to Sapir. 13 November 1916: Lowie t o Kroeber. 13 March 1917. 30. Robert H Lowie, “Review of Time Perspective in Aboriginal Culture,” American Anrhropologisr 2 I (1919): 76. 31. Ibid. 32. Darnell. “Bear Ceremonialism“: Stocking, “Ideas and Institutions.’’ 33. Regna Darnell, “The Revision of the Powell Classification,” Papers in Linguisrics 4 (1971): 70-1 10. 34. Weston La Barre, ”The Influence of Freud on Anthropology.” American lmago 15 (1958): 280-281.

35. “ I was responsible for having brought him to Chicago”-Sapir to Louis Wirth, 25 November 1931, Regenstein Library. University of Chicago. Sapir had been acting chair of anthropology the previous year, while Fay Cooper Cole was away in Washington, D.C. In our view, Sapir would not have made such an asser- tion i f his role had been only titular. Rather, we suggest. the shared esteem of Radcliffe-Brown’s mentor, W. H. H. Rivers, and Boas‘s first generation of students set the stage before any contact between Robert Redfield and RadclifTe- Brown. 36. Robert H. Lowie. Toward Undersranding Germanl. (Chicago: University of Chicago Press, 1954); idem, Social Organ/zarion (New York: Rinehart, 1948). 37. Lowie, Ethnolugi.rr. p. 169. 38. 39.

40.

Murray. Group Formation, pp. 36-38: Joan Mark, Four Anlhropologisls (New York: Science History,

Swanton to Robert Lowie, 30 July 1957, Bancroft Library, Berkeley. A. M. Tozzer, “Roland Burrage Dixon,” American Anrhropologisr 38 (1936): 291. Alfred L. Kroeber. “Roland Burrage Dixon,” American Anrhropologisr 38 (1936): 295. John R . Swanton and Roland B. Dixon, “Primitive American History,” American Anrhropologisi 16

Robert H. Lowie. “Oral Tradition and History,” American Anrhropologisr 17 (1915): 597-599. E. Sidney Hartland, “On the Evidential Value of the Historical Traditions of the Baganda and

Lowie, “Oral Tradition,” pp. 598-599. R. B. Dixon. “Reply,” American Anrhropologisr 17 (1915): 599-600. J . R. Swanton, ”Reply,“ American Anrhropologisr 17 (1915): 600-601. Lowie to Paul Radin, 2 October 1920. Facts rarely compelled either Boas or Lowie to cease doubting any

A. A. Goldenweiser, “The Heuristic Value of Traditional Records,” American Anrhropologisi 17 (1915):

(1914): 376-412.

Stocking. “Ideas and Institutions.’’ pp, 24-25. Stephen 0. Murray, “Resistance to Sociology at Berkeley.“ Journal ofihe Hislory o/Sociology 2 (1980):

Alfred I Kroeber. Configurations o/ Culiural Growih (Berkeley: University of California Press, 1944). 61-84.

340 KENNETH W . PAYNE AND STEPHEN 0. MURRAY

41. Alfred L. Kroeber, “Linguistic Time Depth Results So Far and Their Meaning,” International Journal of American Linguistics 21 (1955): 95-104; idem, “Semantic Contributions of Lexicostatistics,” Infernational Journal of American Linguistics 27 (1961): 1-8. 42. For example, J . R. Swanton, “Ethnological Value of de Soto Narratives,” American Anthropologist 34

43. John R. Swanton, “Some Neglected Data Bearing on Cheyenne, Chippewa and Dakota History,” American Anthropologist 32 (1930): 155-160 (which is centrally concerned with the danger of relying on available ethnohistorical data); idem, Source Materials on the History and Ethnology of the Caddo Indians, Bureau of the American Ethnology Bulletin 132 (1942). For appreciations of this work see A. L. Kroeber’s in- troduction to Essays in Historical Anthropology-Smithsonian Miscellaneous Collections I 0 0 (1940); and William N. Fenton, “John Reed Swanton, 1873-1958,” American Anthropologist 61 (1958): 663-668. Lowie recurrently used Swanton’s work on the lack of clans in some Amerindian tribes to refute evolutionary schema and presented it as the prime exemplar of Boasian ethnology in The Hisfory of Ethnological Theory (New York: Holt, 1937), p. 145. 44. Stephen 0. Murray, “Die ethnoromantische Versuchung,” Die Wissenschaftler und das Irrationale I (1981): 377-385; idem, “Recent LCvi-Strauss,” Contemporary Sociology 10 (I98 I ) : 222-223. 45. See, for instance, Stephen G . Bunker, “Latin American Migration and Migrants,” Contemporary Sociology 1 I (1982): 270-271: “Most societies incorporate tales of migration into their legends. . . . The myths are of course fundamentally correct.” For an example of ideological distortion in a migration myth, see Charles 0. Frake, Language and Cultural Descripfion (Stanford: Stanford University Press, 1980). On the more general necessity of caution in accepting folk models, see LCvi-Straws, Structural Anthropology, pp.

(1932): 570-590.

273-274.