Franz Boas spent a lifetime arguing that the study of human beings could be a science without pretending to be a physics of society. Race, Language and Culture gathers essays, addresses, reviews, and technical papers written across four decades and organizes them by theme rather than by date: race and bodily form, the growth and spread of culture, language, and a closing miscellany. Boas prepared the collection at Columbia University in late 1939, and the earliest pieces reach back to the 1890s, so the volume doubles as a record of how American anthropology remade itself in the first half of the twentieth century.
“Knowledge of the life processes and behavior of man under conditions of life fundamentally different from our own can help us to obtain a freer view of our own lives and of our life problems.”
What the Book Argues About Race
Boas’s central claim is deflationary and, for its time, radical: race is a category of populations, not of essences. He defines “race” and “racial” as the assembly of genetic lines represented in a population, and he insists that racial characteristics be described the way a botanist describes a variable plant—under stated environmental conditions, without prejudging how much change environment might produce. From there the argument turns statistical. Individuals within any race vary enormously, and neighboring European types overlap so heavily that many individuals of one type correspond to individuals of another; only the most distant extremes, such as the blond north European and the dark Sudanese, fail to overlap at all. To say one population averages five centimeters taller than another is not to say anything about any particular pair of people, and Boas blames a whole literature of racial difference on the failure to keep those two senses of “difference” apart. He is equally skeptical of eugenic selection: in a heterogeneous country like the United States, apparent selection by bodily form is usually selection by social class, since Scandinavians drift to the northwest and Hungarians to the Pennsylvania mines. The body, he notes, runs adequately at sea level and at twenty thousand feet, on meat or on vegetables—so within elastic limits, selection by form is unlikely to amount to much.
The Preface and the Purpose of Anthropology
The book opens with its title page and copyright notices—Macmillan, 1940, with a fourth printing in 1953—before Franz Boas’s Preface, dated at Columbia University, November 29, 1939. There he takes aim at the notion that anthropology is merely a pastime for the curious, a discipline that satisfies idle interest in early human history without bearing on present problems. Growing up inside our own civilization, he writes, we know little of how our bodies, our language, and our habits of thought and action are conditioned by environmental limits. Knowledge of human life under fundamentally different conditions can give us a freer view of our own lives and problems. Boas values the dynamics of life over mere description, though he grants that description is indispensable material for conclusions. The collected writings are meant to validate this viewpoint, and they are organized thematically rather than chronologically: general discussions first, then special investigations. Statements are largely left as first published, except that scattered material on racial stability and growth has been combined, with mathematical problems omitted and diagrams substituted for numerical tables; reviews and controversies are included where relevant. “Race” and “racial,” he explains, mean the assembly of genetic lines represented in a population. Earlier papers lack current data, but Boas made no changes, believing the fundamental theoretical treatment still valid, adding only a few footnotes on new investigations or criticisms. Two early general papers appear at the end as indicators of his underlying attitude. He thanks Dr. Alexander Lesser for help in selecting material, and a table of contents and list of illustrations follow, outlining sections on Race, Culture, and miscellaneous topics.
Defining Racial Characteristics and the Problem of Selection
Boas proposes defining racial characteristics as one defines those of a variable plant: by stating that under definite environmental conditions a race’s bodily form and functioning are as observed, without prejudging how far modifications may result from changing environment. The actual problem is to determine whether and how far bodily traits may be so influenced. It is conceivable, he notes, that congenital modifications in form may exist that are nevertheless not hereditary; constitutional changes in the mother’s body may modify prenatal growth and give a superficial impression of hereditary change. These considerations must be weighed in any thorough investigation of the American population.
How far, he asks, do selective agencies determine population movements—immigration and emigration, the settlement of the western United States by eastern inhabitants, migration from country to city—and mating, mortality, and fertility must also be considered. Selection by bodily form can be studied only in a homogeneous population; but homogeneous populations do not exist anywhere, and heterogeneity in modern civilization is always connected with social stratification. In a heterogeneous population like that of the United States, determining a direct relation between selective influences and bodily form is almost insurmountable: Scandinavians attracted to the northwest, Hungarians to the mines of Pennsylvania, Mexicans to the southern borderland, French Canadians to the New England states and northern New York—each group represents a physical type, producing an apparent relation between selection and physical type actually based on social factors. Mating depends on social contact; nationalities clustering together, denominational or racial endogamic restrictions, produce selective mating of similar types due to social heterogeneity, though some selection may unite tall with tall. Social heterogeneity also affects mortality and fertility: more recent immigrants are less well-to-do than earlier immigrants and their descendants; since fertility relates to economic well-being, recent immigrants have more children, shifting population distribution in their favor. Earlier immigrants represent the northwestern European type, later ones the south and east European types, so selection according to bodily form appears—due not to a direct relation between physical characteristics and fertility but to economic group composition. Boas concedes some direct relation between bodily form and selection: settlement of new western countries required bodily and mental vigor; it has often been pointed out, though never proven empirically, that emigrants from New England villages represented a physically superior type—highly probable though not based on observation. Similarly the supposed greater susceptibility of slightly pigmented individuals to certain diseases versus greater resistance of brunets; Boas is not convinced incontrovertible proof exists, but if the blond constitution is weakened by intense sunlight, selective influence might operate when a people moves from cloudy temperate zones to the brilliant sunlight of southern arid climes. He stresses the body’s organs operate adequately under widely varying circumstances—lungs at sea level and at 20,000 feet, heart in sedentary life or high-altitude activity, digestive organs on vegetable or meat diet, the central nervous system adjusting to varied conditions—so as long as environment stays within elastic limits, selective influences are not probable to any great extent, at least not solely determined by form and functioning of organs.
Racial and Individual Differences
Individuals of a race differ considerably in every feature—pigmentation, hair form, body size and proportions, physiological reactions—and these differences are measurable, expressing the race’s variability. A complete presentation of a race’s characteristics would state the relative frequency of each bodily form. Comparing European racial types constituting the bulk of the American population, the range of variation is such that many individuals of one type correspond to individuals of another; certain forms are common to all European populations. Strongly contrasting head forms exist in northern Italy and Sardinia, yet 27 per cent of the population may belong to either. There is considerable overlapping of bodily form between neighboring types; only fundamentally different races show non-overlapping characteristics—the blond north European White and the dark Sudanese Negro show no overlapping in pigmentation, hair form, nose form, lips—but proceeding by steps from northern Europe to the Sudan yields many intermediate overlapping steps, so only extremes are entirely separate. Even where two races are distinct in certain features, other features differ so slightly that assigning an individual to either race is doubtful.
Boas criticizes expressing racial differences by differences between averages or modes as misleading: two individuals measuring 170 cm and 165 cm differ by 5 cm, but two populations averaging 170 cm and 165 cm cannot be said to differ by 5 cm, because with wide variability many individuals of the taller population equal individuals of the shorter. Differences between races are not necessarily differences between individuals; modern literature on racial differences invites misinterpretation through failure to distinguish these senses of “difference.” The generalization that one population is 5 cm shorter is often interpreted as a characteristic of all individuals, while a single selected individual of the shorter race may be much taller than one of the taller race. This holds for all anatomical, physiological, and psychological characteristics exhibiting overlapping—and even for non-overlapping ones, since differences between selected representative individuals vary widely. If Whites are said to have larger brains than Negroes, this means only that the Negro average is lower, not that every White has a larger brain than any Negro; for many such characteristics the difference between racial averages is insignificant compared to within-race variability. Most anatomical characteristics are stable throughout adult life until senile degeneration, whereas physiological and psychological functions vary strongly with environmental conditions and demands on the organism; their variability is therefore much greater, combining differences among individuals with variations of response under varying conditions. When comparing racial types, differences of types must not be expressed simply as differences of averages.
The Tempo of Growth
Physiological development, Boas continues, has an individual “tempo” that is largely hereditary but modifiable by outer conditions.
Powell’s Classification and the Turn to Comparison
Boas begins with the state of American linguistic classification. Powell’s scheme, published in the Annual Reports of the Bureau of American Ethnology and revised in the Handbook of North American Indians, rested on scanty vocabularies, many of them drawn from old and inadequate sources; neither Powell nor collaborators like Gatschet and Dorsey would have called the map final. Since then, students have attended to single languages rather than classification. Largely through Sapir, with Kroeber, Dixon, and especially Radin, attempts revived to compare apparently distinct languages by vocabulary and to prove far-reaching relationships.
Boas’s own position dates from 1893: comparison should move from closely related dialects outward to more diverse forms. He once treated striking morphological resemblances between neighboring stocks—unaccompanied by vocabulary similarities—as proof of relationship on the Indo-European model, but now doubts that reading. The present wide distribution of a few stocks is late; earlier areas were small and many languages have vanished. Indo-European, Chinese, Turkish in Siberia, Bantu, Arabic in North Africa, Algonquian, and Creek superseding isolated southeastern languages all illustrate expansion and extinction. Early diversity among peoples of one physical type was probably greater than we imagine, and languages may have differentiated so far that interrelation cannot be proven without historical knowledge.
Phonetics, Grammar, and Vocabulary as Separate Problems
The real problem is not classification but tracing the history of speech; classification is only a means. Phonetics, grammar, and vocabulary must be studied separately. Phonetic traits—i-sounds and laterals on the North Pacific Coast, nasalization on the central and eastern plains, labialization in the extreme Northwest—often belong to morphologically distinct languages, while great phonetic differences develop within a single stock. Morphological areas likewise fail to coincide: reduplication on the Great Plains, in the Eastern Woodlands, and on the Pacific Coast south of the British Columbia–Alaska boundary, unknown among northern families; incorporation in Shoshonean, Pawnee, Kutenai, and Iroquois; instrumentals in Kutenai, Shoshonean, and Sioux; cases among Shoshonean and its neighbors; nominal and verbal, neutral and active distinctions scattered irregularly.
Loan words seem fewer than in European languages, yet word categories resemble those of neighbors where cultural conditions are alike—reciprocal relationship terms on the western plateaus, for instance. Eskimo lacks preposition-like adverbs and supplies them with special verbs. Similar structure yields similar stem categories: verbs of existence and motion differ by the form of the object (round, long, flat) in Athapascan, Tlingit, Kwakiutl, and Sioux.
Diffusion Against Independent Origin
If areas do not coincide, languages must have influenced one another—acculturation in language. Phonetic influence is intelligible: small tribes, intertribal marriages, the abduction and enslavement of women, alien women transmitting foreign pronunciation; southern Bantu adopted Bushmen and Hottentot clicks despite hostility. Similar word categories arise through shared social, political, and religious forms. Morphological transfer is harder but Boas believes it occurs: nominal cases in upper Chinook dialects under Sahaptin influence; the second and third person in Kutenai through Algonquian contact; Eskimo–Chukchee parallelism in semi-participial verb forms. Yet irregular worldwide distribution shows some similarities are psychological or physiological—limited categories, articulatory limits. Haida instrumental prefixes and Shoshonean, Kutenai, or Sioux parallels are probably not historically related; glottalized sounds in Chile and on the Northwest Coast, neutral and active verbs in Maya, Sioux, and Tlingit, three genders in Indo-European and Chinook—all rash to associate. Borrowed words do occur, as in Comox with Kwakiutl admixture and Bella Coola borrowing from Kwakiutl and Athapascan.
If acculturation affects language, similarities need not mean a common family; we should expect development from diverse sources into a single cultural unit, and the “Ursprache” theory must be held in abeyance. Even Indo-European comparison—Armenian and English—would be difficult without history. Three tasks remain: study dialect differentiation (Siouan, Muskhogean, Algonquian, Shoshonean, Salishan, Athapascan); study the distribution of phonetic, grammatical, and lexicographical phenomena, including principles of concept-grouping; and study dissimilarities equally.
Restating the Case in 1929
The 1929 paper restates the 1920 argument. Morphological types spread over large areas while vocabulary differences complicate any common “Ursprache”; early small units had different conditions of mixture; the answer must come through mutual influence, especially in morphology. Nez Perce, an eastern Sahaptin dialect, has vocalic harmony, an s with raised tongue margin, a dental t series, and a voiced affricative dl; eighteenth-century Sahaptin penetration into Washington and across the Cascades, with intermarriage into Salishan tribes, produced a dialect phonetically identical with neighboring Salishan—same vocalic system, no vocalic harmony. Vocabulary comparison is difficult. Eskimo is conservative, Alaskan and Greenland dialects only slightly differentiated after more than a thousand years, while Aleutian is divergent; Aztekan lost its higher literary style and adopted Spanish syntactic types but is otherwise unchanged, its dialect areas reconstructable from sixteenth-century grammars; Salishan languages of British Columbia and Washington show great instability. “Primitive languages” change rapidly only to a limited extent. Hybridization in morphology remains the question, and the transfer of grammatical categories has never been directly observed.
Culture Against Geography and Economics
No culture has sprung from immediate response to geographical conditions; location matters only because it enables or blocks contact with neighbors of varying culture, and response to foreign stimuli is a key element of change. Spatial relations give the opportunity for contact; the processes themselves are cultural. Economic interpretations fare little better. Morgan’s early attempts to link social organization to economic conditions proved fallacious, and recent purely economic readings are equally unsuccessful. Economic conditions lie closer to culture because they form part of cultural life, but they are both determined and determining; nothing economic makes man an agriculturist or herder—those arts grow from experience of contact between man and plants and animals only indirectly related to economic conditions. Intricate social forms, religious ideas, and art styles cannot be explained by economic needs; mental attitudes of a different order are the determinants. Economics sets the medium in which those attitudes play out, furthering or hindering them, not their forms: lack of leisure prevents artistic industry, a roving life forbids bulky property; leisure and stability favor production but do not create a particular industry or an artistic style.
Attempts to find general laws of cultural integration yield little. Religion and art may seem closely associated, but comparative study shows only that art forms may express religious ideas—of no particular value; religious significance may stimulate a higher style, induce slovenly execution, or forbid representation altogether. In each specific case the integration may be recognized as an important social feature.
Myth as Borrowed Property on the Fraser River
Boas turns to the Salish tribes of the Fraser River delta, close kin to the Thompson Indians, where each village claims a mythical ancestor, some of them animals. One tribe near the river’s mouth traced itself to Beaver-Man, who quarrelled with the transformer and, being weaker, was turned into a beaver; another tradition on the Harrison River looks almost like remembered history, telling how a chief invited the ancestor of a neighbouring tribe—descended from marten and mountain goat—down from the mountains, after which the two lines merged into one people. The explanation lies in social organization. Coast tribes north of the Fraser are divided into totemic clans, each guarding a clan tradition that explains its privileges and is defended as property. When Salish groups met these coast peoples, the absence of family traditions was felt as an inferiority, and the desire to possess such myths arose. But the Salish were organized by village, not clan, so each village produced a single mythical ancestor. The Bella Coola, whose mythology is more thoroughly reshaped by coast influence, show the same process at work.
These cases carry a general lesson for reading primitive myth: its material is of mixed origin, much of it adopted ready-made, and the way it is woven together depends on social conditions and habits. Ritual moulds the explanatory tale; myth conforms to social rank; discrepancies survive as relics of earlier custom or fragments of older traditions under changed conditions, proof of gradual assimilation. Diffusion and modification together obscure original meaning. Explanations of nature are the primary source of myth, but the earliest form cannot be reached without tracing historical change—mythological worlds, Boas writes, have been built up only to be shattered again, and new worlds built from the fragments.
The Growth of Indian Mythologies
The North Pacific coast, drawn from his collection Indianische Sagen, supplies the test case. Its peoples differ in language but share a culture so uniform that origins are hard to detect; acculturation has made myth travel freely, and a large body of legend is common property. Moving away from the area of highest development, arts and customs change gradually and the common stock thins—an elaborate cycle dwindles into mere adventures or episodes within adventures. The raven tradition shows the pattern: most characteristic among the Tlingit, Tsimshian and Haida, it loosens and shrinks going south, surviving fully only to the north end of Vancouver Island, where thirteen of eighteen tales remain at Nahwittee, eight among the Comox, six among the Nootka, three among the Coast Salish. Of the daylight story—raven as cedar spike, swallowed by the daughter of daylight’s owner, reborn, breaking the box—the Nootka keep only the transformation, the Coast Salish only the ruse that forces the owner to release daylight, and the tale reaches Grey’s Harbor in Washington. The pitch adventure reappears far south in a new setting, attached to the origin of sun and moon, while some raven episodes, like the seal with greasy hands, range across all North America with no proof they ever belonged to the Alaskan cycle.
Distribution, then, is not governed by geography alone. The raven legend halts after northern Vancouver Island, absent among southern Kwakiutl-lineage tribes and the west coast of the island, though northern Kwakiutl speakers keep it; only fragments stray south. The cause is a lack of acculturation—difference of character, long hostility, or, most probably, recent shifts of location, since the Bella Coola, a Salish tribe living far north, still stand closest to southern Salish mythologies. Trade routes prove borrowing: the Columbia River road, a path from Athapascan country to the Bella Coola, a minor trail down the Fraser. Contact is strongest along these routes and fades with distance, and tabulations reaching as far as the Micmac, Greenland Eskimo, Ponca and Mackenzie Athapascan show extensive borrowing. Where two North American tales are considerably alike, Boas assumes dissemination over independent invention. A comparison of the Chinook culture hero with Kwakiutl ceremonial origins clinches the method: the Chinook divinity’s grandmother, carried off as a child by a monster, bears the hero’s mother, and mother and child destroy the monster; on Vancouver Island the cannibal spirit seizes a girl, is slain with her help, and their child becomes the new cannibal.
Tricksters, Animal Tales and Human Tales
Boas sorts tricksters by temperament. Raven is greed made cunning, cheating people of food and paying for it; Mink’s tricks run on erotic lines; Bluejay wants to outdo his betters at games, hunting and war; the Plateau Coyote combines all three, with rivalry tricks outnumbering the rest. Trickster and transformer are one: Raven frees the sun because he wants it, releases fresh water because he thirsts, liberates fish because he is hungry, raises the tides to gather shellfish. Where the wish to benefit mankind is pronounced, trickster and culture hero split into two persons, and the coast’s “one who sets things right” is no trickster at all—his tales are bare statements that something was once different, growing complex only in the south. Animal tales resist sharp characterization; Arctic stories barely have a plot, though in complex tales animals act by their natures—Kingfisher dives, Fox runs, Grizzly-Bear bullies—without much individual colour. Human tales live by plot: the faithless wife killed by her disguised husband marks the northern area from Siberia to the Mississippi basin; the deserted child shaming his parents, helped by animals, belongs to the North Pacific coast, while on the Plains he escapes through a protector and becomes a great hunter. Culture mirrors itself in story—Boas reconstructed Tsimshian life from myth alone, though the Raven cycle gives little and the late secret societies little, while potlatch and crests stand out.
Supernatural machinery is area-specific: ice giants in the east, stupid giants among Shoshoni and Kutenai, water monsters in the south, horned and double-headed serpents, the thunder-bird of Vancouver Island. Main plots follow interest—Kwakiutl tales authenticate privileges, Blackfoot tales explain ritual origins. Formal traits matter too: typical repetition, as when the brothers fail until the last succeeds; Tsimshian climaxes of ever more potent help; artificial speech, where Kwakiutl Mink cannot say ts and Kutenai Coyote cannot say s; the Northwest coast’s single-word lines against Southwestern metaphor. No tribe’s tales are purely local, and since actors, explanations, setting and form keep changing, modern explanations are recent adaptations, not integral parts. American folklore offers no case of myth as nature allegory; creative power now seems weak, tales being recombinations of old themes, and folk-tales are primarily works of primitive art, explanation a stylistic embellishment rather than philosophy. Lacking history, Boas compares how much lore is a tribe’s own against what it shares, and finds no tribe markedly more original than another.
Tempo, Growth, and the Case Against Fixed Types
The most original material in the race section concerns growth. Every child develops at an individual tempo, largely hereditary but modifiable by circumstance: tall children mature earlier, girls spurt before boys, and early maturers compress their growth into a short, energetic period while late maturers stretch it out. Average stature in European populations rose markedly between the mid-nineteenth century and the 1930s, and experiments on rats—dried thymus gland, thymus extract injections—showed that development could be accelerated and fertility raised across generations, then lost when treatment stopped. Boas’s own data from an orphan asylum, where all children lived under identical conditions, showed siblings repeating one another’s tempo, which let him treat tempo as an inherited characteristic of individuals within a socially uniform group. He extends the question to whether rapid growth predicts rapid decay, citing Felix Bernstein’s insurance evidence linking early presbyopia to earlier death by arteriosclerosis, and proposes that schools and the officer corps keep the longitudinal records that would settle it.
Language as History, Not Genealogy
Boas treats language the way he treats skulls: as evidence of contact and drift, not of pure descent. Powell’s classification of American languages rested on scanty vocabularies, and Boas doubts that morphological resemblances between neighboring stocks prove common ancestry. Phonetic traits, grammatical habits, and vocabulary each spread on their own maps—reduplication across the Plains and the Pacific coast, instrumental prefixes in Kutenai and Sioux, glottalized sounds in Chile and on the Northwest Coast—and those maps do not coincide. Small tribes intermarry, abduct and enslave women, and adopt foreign pronunciation; the southern Bantu took up Bushmen clicks despite open hostility. If acculturation shapes language, then similarity need not mean family, and the search for an “Ursprache” must wait. Even Armenian and English, he observes, would be hard to connect without written history.
Culture Is Not Geography or Economics
Against geographical determinism Boas grants location only the power to permit or block contact; against economic determinism he grants economics the role of medium, not cause—leisure permits artistic industry but does not dictate a style. Attempts to find general laws of cultural integration, he says, yield only commonplaces, because anthropology is a historical science: like geology, it clings to specific forms, and its generalizations matter in proportion to how closely they adhere to them. The error of early anthropology was using uncritically gathered material for historical reconstruction; the error of modern anthropology is overemphasizing reconstruction at the expense of studying the individual living under his culture. His 1896 paper on the comparative method makes the point sharply: similarities of custom do not prove common origin, since like conditions produce like forms.
Totemism, Myth, and the Kwakiutl
The ethnographic essays apply the same skepticism to particular puzzles. Totemism, Boas argues, is a subjective rather than an objective unit: its contents differ by area, and the appearance of a single “totemic complex” is an artifact of retrospective systematizing. Exogamy is universal and ancient; totemism is neither, and it arose where a unilateral family needed a mark of recognition. Among the Kwakiutl, crests are individual privileges rather than clan property, so the numaym are not true totemic groups, and a chief’s identification with his animal ancestor is oratory, not theology. Secret societies, he shows, spread by imitation and suggestion, with ritual older than the myths that explain it; cannibalism reached the Kwakiutl from the Bella Bella around 1830. Myths travel along trade routes—the raven cycle thinning from the Tlingit southward, the daylight story surviving only as fragments among the Nootka and Coast Salish—and where a tribe lacked family traditions, as among the Fraser River Salish, it invented village ancestors to keep pace with prouder neighbors. “Mythological worlds have been built up,” he writes, “only to be shattered again, and new worlds were built from the fragments.”
The Legacy of a Skeptical Science
Boas’s closing essays carry the method into archaeology, primitive art, and the assimilation of Biblical stories into Thompson Indian myth, where Adam and Eve become Atam and Eem and the Fall shrivels every tree’s fruit. What holds the volume together is a single discipline of mind: describe variation before ranking it, separate the average from the individual, and treat every cultural form as the product of history and contact rather than of innate capacity. That stance helped dismantle the scientific racism of his era, and it also set the terms for the cultural relativism and fieldwork traditions that followed him. The prose is patient and unglamorous, but the cumulative effect is a quiet demolition of the idea that human beings come in fixed kinds.
