Spedkiii^G^Fffonheeld Tongifiss

Home , Mass comparison

praisal of the Dinantian Floras at Oxroad Bay, East Lothian, Scot- land" [Transactions of the Royal Society of Edinburgh B, in press). IVES GODDARD is a Curator in the Department of Anthropology Spedkiii^g^ffFonHeeld at the Smithsonian Institution, where he has worked since 1976. He was born in 1941 and received his A.B. from Harvard College and his Ph.D. in linguistics from Harvard University in 1969. Af- Tongifiss ter a postdoctoral year at the Smithsonian, he taught at Harvard from 1970 to 1976. His research interests are the linguistics and ethnohistory of Native North Americans. His publications include "Algonquian, Wiyot, and Yurok: Proving a Distant Linguis- The Feasibility of Reconciling tic Relationship," in Linguistics and Anthropology: In Honor of C. F. Voegelin, edited by M. D. Kinkade et al. (Lisse: Peter de Human Phylogeny and the Ridder, 1975), Delaware Verbal Morphology (New York: Garland, 1 1979), and (with K. J. Bragdon) Native Writings in Massachusetts History of Language (American Philosophical Society Memoir 185, 1988). RICHARD O'GRADY has been a postdoctoral fellow in the Smithso- nian's Department of Invertebrate Zoology since 1987. He was born in 1956 and received a B.Sc. from the University of British by Richard Bateman, Columbia in 1978, an M.Sc. from McGill University in 1981, and a Ph.D. from the University of British Columbia in 1987. His pub- Ives Goddard, Richard O'Grady, lications include "Ontogenetic Sequences and the Phylogenetics of Parasitic Flatworm Life Cycles" [Cladistics 1:159-70), (with D. R. Brooks and D. R. Glen) "Phylogenetic Analysis of the Di- V. A. Funk, Rich Mooi, genea with Comments on Their Adaptive Radiation" [Canadian Journal of Zoology 63:411-43), and "Historical Processes, Evolu- W. John Kress, and Peter Cannell tionary Explanations, and Problems with Teleology" [Canadian Journal of Zoology 64:1010-20]. v. A. FUNK is Curator (Research Scientist) in the Smithsonian's Department of Botany. Born in 1947, she was educated at Murray State University (B.A., 1969, M.S., 1975) and Ohio State Univer- We address the problem of reconciling human phylogeny and lin- sity (Ph.D. 1980). Her research interests include the phylogenetic guistic history and conclude that its resolution requires (1) devel- systematics of the Compositae, speciation in high-elevation South opment of a valid objective method of quantifying linguistic rela- America, and the flora of the Guianas. She has published "The tionships, (2) delimitation and subsequent characterisation of Systematics of Montanoa" [Memoirs of the New York Botanical human populations and languages by large-scale demographic cen- Garden 36:1-133), (with C. J. Humphries] "Cladistic Methodol- sus, (3] integration of genetic and linguistic data with other types ogy," in Current Concepts in Plant Taxonomy, edited by V. H. of information, (4) parallel analyses of the relationships between Heywood and D. M. Moore (London: Academic Press, 1984), and genetic and linguistic entities using specifically phylogenetic al- (with D. R. Brooks] "Systematics and the Basis of Comparative gorithms, and (5) clarification of the biological and philosophical Biology" [Smithsonian Contributions in Botany 73, in press). relationship between human lineages and potentially dependent cultural phenomena such as speech. Also, increased discourse be- RICH MOOI joined the Department of Invertebrate Zoology at the tween linguists and biologists is needed to distinguish homolo- Smithsonian as a postdoctoral fellow in February 1988. Born in gous from analogous processes in the two disciplines and thereby 1958, he received his B.Sc. (1981) and M.Sc. (1983) in zoology and standardise terms and concepts. Even if these criteria are even- his Ph.D. (1987) in evolutionary biology from the University of tually satisfied, different processes and rates of evolution and ra- Toronto. His research interests include the paleontology and phy- diation in human populations and languages will continue to logenetic relationships of clypeasteroid sea urchins. Among his complicate attempts to recover "racial" phylogenies. Although publications are "Structure and Function of Clypeasteroid Miliary profound, these difficulties may not be insuperable, particularly if Spines" [Zoomorphology 106:212-23), "Non-respiratory Podia of initial studies sample a regional rather than a global catchment. Clypeasteroids" [Zoomorphology 106:21-30, 75-90), and "Living and Fossil Genera of the Clypeasteroida" [Smithsonian Contribu- tion to Zoology, in press). RICHARD BATEMAN is Lindemann Research Fellow in the Depart- ment of Paleobiology at the Smithsonian Institution (NHB E318, w. JOHN KRESS is Associate Curator of Botany at the National Washington, D.C. 20560, U.S.A.) He was born in 1958 and was Museum of Natural History. He was born in 1951 and educated at Assistant Scientific Officer in the Section for Quaternary Studies Harvard (B.A., 1975) and at Duke (Ph.D., 1981). His interests in- of Rothamsted Experimental Station, Harpenden, U.K., from 1977 clude the taxonomy of tropical plants, plant reproductive biology, to 1984. He holds B.Sc.'s in life sciences, from Luton College of and speciation in tropical angiosperms. He has published "Self-In- Higher Education (1982), and earth sciences, from Birkbeck Col- compatibility Systems in Central American Heliconia [Evolution lege of London University (1984), and a Ph.D. in Dinantian palaeo- 37:735-44)/ "The Phylogeny and Classification of the Zingi- botany from London University (1988). His activities include pro- berales," in Systematics and Evolution of the Monocotyledons, moting "demographic" systematics and investigating the roles of edited by J. W. Walker (Annals of the Missouri Botanical Garden, saltational macromutation and pollination biology in the in press), and "The Systematic Distribution of Vascular Epi- megaevolution of higher plants. Among his publications are "A phytes," in Phylogeny and Ecophysiology of Epiphytes, edited Reappraisal of the British and Irish Dactylorchids" (with I. Den- by U. Luttge (Berlin: Springer-Verlag, in press). holm) {Watsonia 14:347-76, 15:321-55, 17:319-49) and "A Reap- PETER CANNELL is Research Associate in the Division of Birds of the National Museum of Natural History. Born in 1954, he received his B.A. from Bowdoin College in 1977 and, through a joint program with Queens College, City University of New York, and 1. We thank R. B. Potts for information on early human evolution, the American Museum of Natural History, his M.Phil, in 1984 W. A. DiMichele for helpful discussion, and J. T. Temple for a and his Ph.D. in 1986, both in evolutionary biology. His thesis critical reading of the manuscript. G. McKee kindly scanned the work and subsequent studies as a postdoctoral fellow at the Na- literature in search of human genetic data. tional Museum explored the value of the anatomy of the syrinx 2 | CURRENT ANTHROPOLOGY Volume 31, Number 1, February 1990

for elucidating the phylogenetic relationships of birds. He has treeness, and population-admixture tests (all p. 6004). more general interests in the development of complex vocal or- Also, the reader is prevented from interpreting the rela- gans and its correlation with complex communication. tionships of five European populations, which are pooled The present paper was submitted in final form 5 vn 89. a posteriori without explanation (p. 6003, fig. 1). The omission of data and sampling details from the paper One reason for our tardiness in treating sociocultural (and from all potential primary sources that we have evolution as a selection process is that most of us been able to consult) precludes assessment of the rigour know a great deal about the vagaries of sociocultural of their data collection and analysis. transmission and have an overly simple view of biological transmission. D.L. HULL, Science as a Process Delimitation and Characterisation of Genetic The instant fame of Cavalli-Sforza et al.'s (1988) "Recon- Entities struction of Human Evolution," which received the rare accolade of rapid review in Science (Lewin 1988a), Na- The limited data that are presented by Cavalli-Sforza et ture (Diamond 1988), and Natural History (Gould 1989), al. do not inspire confidence. The data-matrix of 42 en- reflects the fascination of the question that it raises, tities (putative "racial" groups) x 120 attributes (non- namely, whether human phylogeny can be recon- DNA polymorphic alleles) contains 23.7% gaps, reflect- structed by integrating genetic and linguistic informa- ing the authors' desire to maximise the number of tion within the broad temporal context provided by the attributes in the analysis ("The number of genes is of few relevant dated assemblages of fossil humans. We paramount importance for the accuracy of conclusions" have argued elsewhere (O'Grady et al. 1989) that its [Cavalli-Sforza et al. 1989:1128]). This contravenes the main conclusion, that there is "considerable parallelism general theoretical principle that no single row of en- between genetic and linguistic evolution" (p. 6002), is tities or column of attributes in a phenetic data-matrix undermined by analysis of an inadequate database by should have more than 5 % missing values (e.g., Temple inappropriate methods and by several conceptual flaws 1982). in subsequent interpretations. (For their response, see The usefulness of the genetic entities (operational tax- Cavalli-Sforza et al. 1989.) In this paper, we take the onomic units sensu Sneath and Sokal [1973:69]) is also opportunity to discuss our reservations in greater detail, questionable. Cavalli-Sforza et al. (1988) repeatedly term in order to explain how anthropological studies of such these entities "populations," eschewing the term "race" potentially profound significance may be made scientifi- on the grounds that it is "a concept which, for humans, cally rigorous. is devoid of any useful scientific definition" (1989:1128). They delimit all 42 populations a priori, most by partitioning global landmasses into geographic regions and Constraints on Critical Appraisal practicing "pooling" within these regions to form geographically delimited groups of organisms (topodemes The crux of the paper by Cavalli-Sforza et al. is a pheno- sensu Gilmour and Gregor [1939]). They report having gram (a tree reflecting overall similarities of entities) "avoided as much as possible the use of individual popu- that orders "aboriginal" human populations by sum- lations, unless they were unique, and pooled them with mary of selected "non-DNA" (i.e., enzyme) allele fre- others of similar age and used averages [of allele frequen- quencies using Nei's genetic distance (their fig. 1). Lin- cies]. A few unique, isolated populations ... were ... the 2 guistic "phyla" and "superphyla" (extracted from average of many samples" (p. 6004). Hence, the geo- Ruhlen [1987] and Greenberg [1987]) are superimposed graphic entities termed populations in the data-matrix on the phenogram, which we have redrawn as fig. 1. are aggregates of lesser geographic entities that are also Critical appraisal of the phenogram (and the paper in termed populations. Even when 5 European data-matrix general) is hampered by the omission of much crucial populations are further amalgamated, the result is still information. On five occasions, the reader is referred to described as "one population" (p. 6003). Thus, three dif- methodological procedures to be published elsewhere: ferent levels of geographically delimited entity are en- (1) treatment of "biological or statistical weaknesses" compassed by the unqualified term "population," and no in the genetic data (p. 6002), (2) tests of statistical justification is offered for the geographic locations of the significance of the dichotomies in the phenogram, (3) putative population boundaries at any of these levels. statistical evaluation of the "treeness" of the phenogram Geographic pooling suffices for 36 data-matrix popula- to infer evolutionary rates, (4) statistical evaluation of tions, but the remaining 6 (although possessing discrete "population hybridisation" and its implications for geographic distributions) are delimited primarily by lin- the phenogram, and (5) integration of bootstrapping, guistic affinity (they may be termed "glottodemes"). By opting to use genetic data as attributes, Cavalli-Sforza et al. implicitly assume that these topodemes and glot- 2. The 16 linguistic entities listed by Cavalli-Sforza et al. are a mixture of "families" and "phyla" as defined by Ruhlen (1987). todemes are also composed of interbreeding individuals Since they treat families and phyla as equal in status, for simplicity (gamodemes) and exhibit significant genotypic differ- we refer to all of these linguistic entities as "phyla." ences (genodemes sensu Gilmour and Heslop-Harrison BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 3

POPULATIONS LINGUISTIC PHYLA

NIGER-KORDOFANIAN" X NILOSAHARAN- O 3 — J KHOISAN H >• < I a. a. AFROASIATIC- IV e- a. ui u -a a. z 3 en INDOEUROPEAN-

(DRAVIDIAN LING.J-E-DRAVIDIAN- VI z URALIC-YUKAGHIR- U 3 - J < I — a. ui ce ALTAIC- VIII < ui os a. 3 3 ui in ESKI MO-ALEUT IX CHUKCHI-KAMCHATKAN-

AMERIND-

• N.V. AMERIND (NA-DENE LING. )2S-NA-DENE XII -#S. CHINESE- 2 SINOTIBETAN — MON KHMER- ^ AUSTROASIATIC- THAI- DAIC -AUSTRIC-

AUSTRONESIAN—'

INDOPACIFIC- XVI AUSTRALIAN—

0.030 0.024 0.018 0.012 0.006 0.000 GENETIC DISTANCE FIG. 1. Phenogram of human populations (1-37) based on average-linkage analysis of Nei's genetic distances representing 120 non-DNA polymorphisms. Linguistic phyla (I-XVI) and superphyla of Ruhlen (1987) are superimposed, and six population aggregates (A-F) recognised by Cavalli-Sforza et al. (1988) are labeled. (Redrawn from Cavalli-Sforza et al. [1988: fig. 1], where it is argued that Melanesians [population 35] possess two linguistic phyla [XIV, XV]).

[1954]). Thus "population" is used indiscriminately for others (e.g., Cann, Stoneking, and Wilson 1987) have four fundamentally different types of deme (topodemes, elucidated the numerous difficulties associated with the glottodemes, gamodemes, and genodemes), demonstrat- sampling of human genodemes. Samples are generally ing that "population" and "race" both require thought- small, their geographical coverage local and idiosyn- ful qualification when used in a biological context. cratic, and their genetic "purity" (i.e., absence of admix- What is more serious, subsequent genetic characteri- tures of genes recently acquired from other populations) sation of the populations sensu lato is essentially ty- questionable. Furthermore, different individuals of par- pological; samples are too small to permit determination ticular populations tend to be analysed for different at- of the range of genetic variation exhibited by the tributes by different workers. These variations in sample genodemes that they supposedly represent. Although size, attributes measured, and, especially, population Cavalli-Sforza et al. provide no information on the num- concept represented result in data-sets that are concep- ber (or nature) of individuals analysed per population,3 tually (though not methodologically) incompatible. In other words, any pair of data-sets sharing attributes can 3. We are unable to penetrate the ambiguities of Cavalli-Sforza et be compared statistically, but such comparison is of al.'s (1989:1128) subsequent statement that "the average size for doubtful value if the entities represented by the data-sets gene frequencies is well above 100 individuals. Small samples are rare and were avoided." Sample sizes should always be reported, occupy different levels in the demographic hierarchy together with ranges and/or sample standard deviations if sample (see Bateman and Denholm 1989a, b). We assume that size varies. such inconsistencies underlie the "population means" 4 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

that comprise Cavalli-Sforza et al.'s basic data-matrix. 1987:130, 148, 180, 221-27), are awarded equal status Certainly, they lacked the large volume of fully compat- with generally accepted phyla such as Sino-Tibetan and ible data per population necessary to identify the zones well-established (and probably relatively recent) phyla of positively correlated depressions in allele frequencies such as Indo-European and Dravidian. As Greenberg that would allow objective delimitation of genodemes, concedes (in Lewin 1988b: 1632), there is little support the most appropriate intraspecific entities for phylo- among specialists for his Amerind hypothesis, which genetic analysis.4 This is most readily achieved by mul- unites all but the northwesternmost Western Hemi- tivariate ordinations, using individual humans as en- sphere languages. Criticisms extend throughout the lin- tities and algorithms that do not require a priori guistic hierarchy within Amerind, encompassing its six assignment of entities to groups (see Gower 1984). Once constituent branches (e.g., Northern Amerind), most of genodemes have been delimited, intragenodeme varia- its constituent stocks (e.g., Almosan-Keresiouan), the tion in alleles can be accurately assessed. constituent units of some of the stocks (e.g., Almosan), The consequences of the typological approach to char- and even some of the subunits (e.g., Mosan). Ruhlen's acterisation of populations would be less deleterious if Austric phylum (credited to a personal communication the resulting mean data-sets were less similar. Cavalli- from Greenberg [Ruhlen 1987:153]) consists of the Au- Sforza et al.'s phenogram is, however, rooted at a small stroasiatic and Austronesian families (a combination Nei's genetic distance (0.030), and the most similar pair originally proposed by Wilhelm Schmidt) plus the Daic of populations is linked at a minimal Nei's genetic dis- (Kam-Tai) and Hmong-Mien (Miao-Yao) families, which tance of only 0.001; given this value, the a posteriori lack generally accepted affiliations with each other or pooling of the European populations could be objectively with any other families. The original concept of Altaic justified only if the distances among them were all less (comprising Turkic, Mongolian, and Tungusic) is now than 0.001 (i.e., if they were virtually genetically identi- widely regarded as discredited; in contrast, an affiliation cal). In this phenogram, 20 of the 26 first-order di- of Japanese, Korean, and possibly Tungusic is increas- chotomies (i.e., those directly subtending at least one ingly favoured (Doerfer 1985, Unger n.d., Whitman n.d.). population) occur at or below 0.006. We suspect that Other problems may prove more common on detailed intrapopulation variation would approximate this figure investigation. For example, those posed by geographi- and that individuals of the same population can there- cally separate branches of some phyla such as Munda fore be less similar to each other than individuals of (the branch of Austroasiatic spoken in India) and the different populations in these attributes. This conclu- forms of Uralic spoken by North Europeans and by very sion is supported by the PAUP5-generated parsimony localised languages such as those of Papua (Wurm 1982, tree for mitochondrial DNA data presented by Cann, Diamond 1988) are obscured by the coarse resolution of Stoneking, and Wilson (1987: fig. 3) (see Spuhler the phyla. [1988:36-38] for a critical evaluation of this study). Al- Much of the controversy surrounding the linguistic though individuals from different continents exhibit entities can be attributed to substantial variation in the significantly higher frequencies in certain clusters, over- criteria by which they are delimited. To permit valid all clustering is poor, reflecting considerable intraconti- comparison, both racial and linguistic entities should in nental variation in gene frequencies. Representation of principle be delimited objectively and independently. samples of individuals from a population as a single However, although several techniques allow objective data-set of mean values for attributes fails to take ac- delimitation of genetic races, there are no generally account of such intrapopulation variation and thus seri- cepted analogous techniques in historical linguistics. ously compromises the apparent interpretative value of Perceived linguistic similarities (whether they reflect the phenogram that forms the basis of Cavalli-Sforza et phonology, grammar, semantics, or syntax) have a al.'s discussions. highly problematic relationship to historical connections to the extent that they may be less likely to reflect shared inheritance than dissimilarities that can be ex- Delimitation and Characterisation of plained as the results of divergent historical processes Linguistic Entities (e.g., Hoenigswald 1987:260). Hence, historical linguists methodologically distinguish convergent (ahistorical Cavalli-Sforza et al. list 16 of the 17 linguistic "phyla" of stochastic), "genetic" (historically inherited), and diffu- Ruhlen (1987), together with two "superphyla" dis- sional (historically borrowed) resemblances by con- cussed by Ruhlen (1987) and Greenberg (1987). Ruhlen's structing hypothetical models of the histories of sets of book is a synthesis of both well-established and contro- related languages (or, more typically, selected parts of versial work. Phyla that are subject to serious criticism, such languages) (e.g., Ruvolo 1987). The history of lan- such as Altaic s.L, Austric s.L, Indo-Paciflc, Amerind guage is regarded as a sequence of changes (innovations) (sensu Greenberg 1987), and Na-dene s.L (Ruhlen between two points in time. Innovations can develop spontaneously within a language or be borrowed from another language. They are considered historically valu- 4. Many systematists consider operational taxonomic units less inclusive than the species inappropriate for phylogenetic analysis able only if they become ubiquitous within a population (cf. de Queiroz and Donoghue 1988, Wheeler and Nixon n.d.). (e.g., Wiener 1987:218-^). The prototypical historical- 5. D. Swofford's "Phylogenetic Analysis Using Parsimony." linguistic hypothesis postulates some features of a BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 5

reconstructed ancestral language (protolanguage) and including an evolving understanding of the extent to specifies the innovations that have accrued in each puta- which general linguistic principles can predict the direc- tive descendant via a complex of unique and interdepen- tionality of changes. Because languages typically do not dent historical events. retain unaltered earlier attributes in sufficient quantities For example, superficial evaluation of the second- to permit outgroup comparison (e.g., Wiener 1987:220- person pronomial prefixes (Sioux) ni-, (Cheyenne) ne-, 21), a linguistic tree can be rooted only by untestable a and (Ojibwa) gi-6 suggests a match between Sioux and priori assumptions of historical relationships. Cheyenne to the exclusion of Ojibwa. However, applica- Historical linguists therefore regard languages as pre- tion of historical-linguistic techniques to the same data conceived conceptual entities that cannot be quantified demonstrates that Cheyenne ne- and Ojibwa gi- are his- in isolation; attributes can only be identified and scored torically identical. Cheyenne ne- and Ojibwa gi- (in any by comparison with other putatively related languages. function) are the expected continuations of Proto- An a priori historical-linguistic phylogeny determines Algonquian *ke- in the respective languages,7 and the the character states, and a unique line of reasoning sup- superficial resemblance between Sioux ni- and Chey- ports each postulated event (Hoenigswald i960, 1973). enne ne- is a historically spurious accidental conver- Phenetic evaluation of linguistic data will succeed only gence. in straightforward cases of close relationship resulting These innovations illustrate four constraints on quan- from equal rates of change. Consequently, the relation- tification: ships of miscellaneous languages cannot be determined 1. Languages lack the widespread but nonuniversal at- simply by tabulating similar traits. tributes that characterise major groups of organisms Thus, techniques developed to classify languages ob- (e.g., the backbones of vertebrates). The linguistic inno- jectively, such as glottochronology and lexicostatistics vations outlined in n. 7 have no broader application. (e.g., Gudschinsky 1964, Hoenigswald 1987), rely on sev- 2. Once adopted, phonological innovations can be eral controversial a priori assumptions and yield results masked by further phonological changes or by functional that are at best ambiguous (Bergsland and Vogt 1962; simplification, termed analogical leveling. Both types Ruvolo 1987:194; Rankin 1988:647). Other attempts to of change tend to remove the evidence for the initial estimate the number or percentage of various types of phonological innovation and thereby obscure homology matches between languages that would rule out chance (Wiener 1987:222). similarities on probabilistic grounds (e.g., Bender 1969) 3. Homoplasies (parallelisms, convergences, and rever- have not achieved general acceptance. Unfortunately, as sals [e.g., Farris 1983]) are commonplace (e.g., Ruvolo is noted by Hoenigswald (1987:257-58), rejection of 1987:194,- Wang 1987:252) and can covary in greater these techniques precludes objective delimitation of lin- numbers than do homologues. guistic entities of the kind advocated earlier for genetic 4. Languages rarely satisfy the criteria used in biosys- entities and confines the linguist to typological charac- tematics to establish the direction of evolution (polarity) terisation followed by analysis using the somewhat within a series of related attributes. Polarity assessment, tautologous historical-linguistic methodology. which temporally "roots" a tree, requires that the earlier Of the 16 phyla used by Cavalli-Sforza et al., 8 (Afro- attributes of a historical series survive unaltered in at Asiatic, Amerind, Austric s.L, Indo-Pacific, Khoisan, least one of the entities being compared. In biosystemat- Na-Dene s.L, Niger-Kordofanian, Nilo-Saharan) are pos- ics, this polarity assessment is called outgroup analysis tulated by Greenberg (1963) and Ruhlen (1987) using a (e.g., Farris 1983); its implementation depends not on phenetic method of classifying languages: "mass com- assumptions of increasing or decreasing complexity of parison" of selected words and grammatical elements of attributes (or of other transformational dines) but on a languages and the subsequent tabulation of perceived maximum-parsimony explanation of the extant distribu- similarities. This approach assumes that sound corre- tion of the attributes of an evolutionary series. Hypoth- spondences will be lost after a given period following eses of polarity are tested against one another by the initial divergence (Greenberg 1987:1—37) and produces parsimony criterion. Because outgroups can themselves results similar to those of historical linguistics only at evolve (Wiley 1981), it may not be possible to polarise shallow time depths (Goddard 1987, Campbell 1988). every attribute in the analysis, and it may be necessary In contrast, the Nostratic superphylum is based on to use more than one outgroup. In evaluating hypotheses proposed sound correspondences between languages of linguistic history, the parsimony criterion is comple- (Kaiser and Shevoroshkin 1988). If the sound correspon- mented by a considerable body of comparative evidence dences proposed for Nostratic are correct, the premise of for the relative probabilities of different types of change, Greenberg's method is false, and the Amerind superphylum (and many of its component phyla) must be rejected 6. These prefixes differ in the particulars of the ways in which they because no sound correspondences have been demon- they are used but share characteristics such as the inalienable pos- strated. Alternatively, if Greenberg is correct, the pro- session of nouns by a second-person singular. posed Nostratic sound correspondences must be spuri- 7. The Proto-Algonquian second-person prefix *ke- is directly ous; they would merely demonstrate the extent to confirmed by Fox ke-. Proto-Algonquian *e became Ojibwa i and *k became g. Proto-Algonquian *ke- became pre-Cheyenne *kye-; which accidental similarities between languages can be word-initial *k- was lost; *y became Cheyenne n (Goddard 1988). discovered if the concept of similarity employed is Sioux ni- is from a similar Proto-Siouan form. sufficiently broad. Thus, the link between Nostratic and 6 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

Amerind advocated by Shevoroshkin (1988) and Cavalli- Pacific Islands (node = 0.010). However, objective parti- Sforza et al. connects two entities delimited by mutually tioning of a phenogram into groups of equal status re- exclusive (and therefore wholly incompatible) criteria. quires transverse dissection at a specific distance value Given these differences, we endorse Diamond's (e.g., Sneath and Sokal 1973:294-96). By this method, (1988:623) argument that quantitative characterisation the closest possible approximation to the six groups of of languages would be necessary in order to raise linguis- Cavalli-Sforza et al. occurs at a genetic distance of o.on tics to the same analytical and interpretative level as the (fig. 1), but this also requires further partitioning of the best of the current studies of genetic and morphological African group (San and Ethiopian are separated from the data. We recognise, however, that the quantification of remainder) and the New Guinea/Australia group (New language is fraught with hazard, as is the identification Guinean and Australian populations are separated), thus of historically significant similarities that are "inher- generating eight population aggregates. The two addi- ited" within populations rather than diffused among tional groups are rejected by Cavalli-Sforza et al. (1988: them (e.g., Ruvolo 1987:195; Wiener 1987:222). fig. 1) and Gould (1989: unnumbered figure) in favour of more interpretationally convenient but methodologically incorrect resolutions into six and seven groups re- Congruence of the Genetic Tree and spectively. the Linguistic Phyla We have attempted to quantify the congruence of the six population aggregates of Cavalli-Sforza et al. (A-F) In Cavalli-Sforza et al.'s innovative comparison of the with the 16 linguistic phyla (I-XVI). By definition, the 7 genetically based phenogram with Ruhlen's (1987) lin- linguistic phyla awarded parity with single populations guistic phyla, the phyla are treated as a second set of (II, III, VI, IX, X, XII, XVI) will agree with any tree. Five of entities and used as an ostensibly independent assess- the remaining 9 phyla are confined to one of the six ment of the integrity of the phenogram (though the population aggregates each (I to Aggregate A, V to B, VIII delimitation of six populations by language effectively to C, XI to D, XIV to E) while the other 4 occur in two precludes objective independence). Unlike the genetic aggregates each (IV in A and B, VII in B and C, XIII in C populations, the linguistic phyla are not represented by and E, XV in E and F). Thus, the 9 linguistic phyla ame- any quantitative data and therefore cannot be ordered nable to a test of agreement at the relatively coarse level algorithmically to form a tree. Also, most linguistic en- of the six population aggregates show 56% (5/9) corre- tities are awarded equal status as "phyla" s.l.; the only spondence. exceptions are three subphyla of the Austric phylum Correspondence between the genetic and linguistic (Austroasiatic, Daic, Austronesian) and the two largely entities can be assessed at a finer level of resolution by overlapping superphyla (Nostratic and Eurasiatic). Thus, determining whether a linguistic phylum is exclusively the linguistic "classification" is almost non-hierarchi- shared by inclusive clusters of populations within any cal, and the linguistic phyla can only be superimposed one of the six aggregates. Of the five phyla that are on the genetic phenogram of populations (see fig. 1), a confined to a single aggregate, only one (Amerind) corre- procedure that requires the dubious (and certainly un- sponds with an inclusive group of all the populations substantiated) assumption that boundaries of linguistic speaking languages of that phylum (XI with 23-25). The phyla coincide precisely with boundaries of populations other four phyla consist of populations that are not in- or aggregates of populations. clusively grouped in the genetic-data tree (I groups 1 Cavalli-Sforza et al. claim precise parity for seven lin- and 2, V groups 8-11, VIII groups 15, 17-20, XIV groups guistic phyla and genetic populations, while the remain- 28-35). Thus, correspondence at this more rigorous level ing nine phyla were associated with aggregates of up to of comparison is only 11% (1/9). seven populations each. In depicting the relationships of Neither of the two linguistic superphyla (Nostratic the phyla and populations, Cavalli-Sforza et al. exploit and Eurasiatic) precisely corresponds with any of the the mobile-like properties of a repeatedly branching tree population aggregates or groups of population aggre- that are succinctly reiterated by Gould (1989:25); nodes gates, undermining the "remarkable correspondence" of the phenogram are rotated to achieve maximum ap- perceived by Cavalli-Sforza et al. (1988:6002) between parent congruence of populations and linguistic phyla. the superphyla and the North Eurasian group of popula- Determination of true congruence is hampered by the tion aggregates (B-D). The North Eurasian group in- non-hierarchical treatment of the linguistic phyla. In cludes populations that speak non-Nostratic languages these circumstances, congruence is most appropriately (XIII in 16, IX in 21, X in 22, XII in 26), and a Nostratic assessed by observing whether a particular linguistic language occurs in a population (IV in 5) that is excluded phylum corresponds with inclusive clusters of popula- from the North Eurasian group. Overall, the Nostratic tions on the phenogram. Cavalli-Sforza et al. recognise superphylum encompasses 17% of aggregate A, all of B, six population aggregates (Africa, Caucasoid, Northeast and 67% of C. Inclusion of the Amerind phylum within Asia, America, Southeast Asia, New Guinea/Australia Nostratic (as advocated by Shevoroshkin [1988]) adds [A—F in our fig. 1], subtended by nodes that span a rela- 75% of aggregate D. Greenberg's (1987) Eurasiatic super- tively broad range of genetic distances (0.008-0.014). phylum largely overlaps Nostratic s.L, encompassing Gould (1989) recognises seven aggregates, further divid- 63% of aggregate B and 89% of C. Cavalli-Sforza et al. (p. ing Southeast Asia into Southeast Asia sensu stricto and 6005) argue that an amalgam of the two superphyla "in- BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 7

eludes all, and only, the languages spoken in our major phenogram and the linguistic phyla, we accepted pro Northeurasian cluster, with the exception of Na-Dene tern the authors' interpretation of the phenogram as [XII]." In fact, all or parts of two non-Nostratic/ analogous to a phylogenetic tree and examined the ori- Eurasiatic phyla are included (XIII in 16, XII in 26) and gin and persistence of the linguistic phyla among the part of one Nostratic/Eurasiatic language is excluded (IV racial lineages. Treating the 16 linguistic phyla as attri- in 5). butes (i.e., potential synapomorphic homologues) capa- Given 56% correspondence between linguistic phyla ble of distinguishing closest kinship groups of popula- and population aggregates at the coarse level of resolutions, we "mapped" them onto the genetic-data tree (fig. tion, 11% correspondence at the fine level, and the poor 2) in order to calculate a consistency index (Kluge and integrity of both superphyla, the parallelism between Farris 1969, Brooks, O'Grady, and Wiley 1986): the the genetic and linguistic entities does not strike us as minimum necessary evolutionary events10 divided by especially "remarkable." the actual events of character evolution required to explain the distribution of the linguistic phyla on the tree. Phyla confined to a single population are trivially consis- Phenetic and Cladistic Approaches to tent with any tree and were therefore omitted from the Phylogenetic Inference calculation. The resulting consistency index of 48%u suggests Cavalli-Sforza et al. (1988:6002; 1989) and commen- that over half the population-language associations tators (Diamond 1988, Lewin 1988a, Gould 1989) as- must be attributed not to a common history but to (1) sume that the tree of human populations (fig. 1) consti- the independent origin of a linguistic phylum in more tutes a phylogeny that represents a temporal sequence of than one population (e.g., IV in 6 and 7, VII in 13 and racial divergences. However, the tree is based on Nei's 14)—an untenably improbable hypothesis—or (2) the re- genetic distance, a measure of overall similarity, and is placement of one linguistic phylum by another in cer- therefore phenetic8 (i.e., a phenogram). It does not depict tain populations (e.g., II for I in 3, VI for V in 12). the positions of specific character transitions on the to- Cavalli-Sforza et al. (1988:6002, 6005) dismiss some of pology of the tree, nor does it distinguish homologous these discrepancies as "easily understood overlaps" characters derived through descent from a common an- reflecting loss of the "original language" (e.g., Mbuti cestor (synapomorphies) from those representing re- Pygmies) or its retention due to isolation (e.g., Basque), tained primitive characters (symplesiomorphies). Nel- genetic admixture ("hybridisation" s.l.), and/or docu- son and Platnick (1981) illustrate these concepts with a mented migrations. We accept the probable importance shark, a lungfish, and a human: although the shark and of these processes but note that they could also be re- the lungfish exhibit the greater overall similarity, the sponsible for generating the congiuities between popula- distribution of synapomorphies (e.g., lungs) indicates tions and phyla,- ad hoc explanation is a double-edged that the lungfish and the human possess a common an- sword. cestor not shared by the shark and are therefore more Whatever the reason, the poor fit between the ge- closely related by descent. Tree construction using syn- netic tree and the linguistic data can only reflect one of apomorphies only is termed cladistics.9 The resulting three explanations: (1) If the genetic tree is accurate, cladograms are more explicit and more specifically phy- the correlation between the evolution of human popula- logenetic than phenograms (e.g., Wiley 1981, Farris 1983). The topologies (branching patterns) of the two 10. This is the most parsimonious arrangement of characters. types of tree can be compared, though a phenogram Wherever possible, shared possession of a language is attributed to rarely exhibits the same topology as the most parsimoni- shared inheritance, thus accounting for the data with minimum reliance on ad hoc explanations. ous cladogram generated from the same data. n. Cavalli-Sforza et al.'s (1989:1128) criticism of our use of the In the absence of data, we cannot produce a cladogram consistency index to evaluate their study is unfounded. The index for comparison with Cavalli-Sforza et al.'s phenogram. counts the number of character changes on a tree, which represents To examine the relationships between the genetic a sample size of one. Measures of standard error are therefore in- applicable. In assessing the fit of a character to a particular tree, the tree is accepted as a statement of relationships of the operational 8. Cavalli-Sforza et al.'s (1989:1128) argument that their tree is taxonomic units, whether the data used to construct it are discrete, "genetic" rather than "phenetic" presumably reflects an erroneous continuous, or "quasi-continuous." Next, the characters (in this assumption that "phenetic" is synonymous with "phenotypic." case, the 16 discrete linguistic phyla) are mapped onto the tree in Rather than describing a class of data (i.e., genotypic versus order that the character-state changes which must be postulated to phenotypic), "phenetic" describes a class of data-analytical tech- explain the distribution of the characters on the tree can be niques that can be compared with "cladistic" techniques. Phenetic counted. Irrespective of the size of the tree, the consistency index trees reflect overall similarity; their topologies are dictated by of 48% means that over half of the linguistic data do not fit (i.e., do branch lengths, and their nodes cannot be related to particular at- not support the groupings of) the genetic-data tree. Our calculation tributes. In contrast, cladistic trees group only by derived homolo- of the consistency index incorporated the hierarchical structure of gous similarity (synapomorphy), and their nodes reflect explicit sub-phyla within the Austric phylum (XIVc ancestral to XlVa and character-state transitions. XlVb) by taking the mean of the consistency indices of all 9 linguis- 9. Widespread use of most pivotal cladistic concepts (albeit less tic phyla (4.33/9). If the sub-phyla are ignored, the consistency in- explicitly expressed) in textual and linguistic studies predated that dex of the tree decreases to 42% (11/26). If the autapomorphic in biological systematics (cf. Hoenigswald i960, Hennig 1966, Plat- states XlVa and XlVb also are omitted from the calculation, the nick and Cameron 1977, Ruvolo 1987, Wiener 1987). consistency index decreases further to 38% (9/24). 8 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

Nostratic Eurasiatic iW 11 Si ill j

0 op Qy m ED wi rximiv mm m m m] ffl [pixyn 0 Tz 3 4 '5 6 7''8 9 10 11' 12l3 14"l5l6 17 18 19 20'21 22 '23 24 25'26 27 28 29'3031 32 33 3435s 36 37

FIG. 2. Phenogram of fig. 1 redrafted as a cladogram to permit mapping of the linguistic phyla onto the tree as independent attributes and calculation of a consistency index. Putative historical relationships within the tree determine the pattern of language acquisition. Roman numerals on branches, acquisition or loss (negative sign) of linguistic phyla through time. Population 3 5 is considered to possess simultaneously linguistic phyla XIVc and XV.

tions and languages is poor. (2) If independent acquisi- through time (e.g., Kimura 1983, Levinton 1988). Or- tion of particular languages is untenable, the genetic tree ganic evolution is mediated and constrained by the ex- is inaccurate. (3) Both the genetic tree and the linguistic tremely conservative intrinsic mechanism of genetic phyla are seriously flawed. We suspect that explanation inheritance. Gene mutations (duplication, addition, 3 is most accurate. deletion, inversion, transposition, and substitution events), even if they confer great adaptive advantage, can only spread gradually through genodemes over many Comparative Evolution of Human generations. Genodemes and Languages Superficially, analogies between the evolution of life and that of language are attractive. The possession of Most biologists agree that there is a single history of life language per se and the evolution of particular languages (holophyly, the most inclusive form of monophyly), both appear amenable to explanation in terms of adap- characterised by synapomorphies (shared derived charac- tive fitness. Also, languages clearly undergo "descent ters) such as the genetic code and the universal molecu- with modification," manifested as changes in the shapes lar symmetry of metabolised sugars (e.g., Margulis 1982). and arrangements of words and other elements that are The subsequent history of life is envisaged as a re- analogous to genetic mutations. However, unlike genes, peatedly branching tree, each branching point repre- words are not intrinsically constrained to the reproduc- senting a speciation event. The topologies of most por- tive cycle via the genome. Like behavioural traits, they tions of this "tree of life" remain highly speculative, as can be acquired during the lifetime of the individual; does the degree of constancy of overall evolutionary rate transcriptional modifications are both more readily BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 9

achieved (either consciously or subconsciously) and language loss and migration, factors that must have had more readily rectified. As with mutant genes, the suc- a profound cumulative effect throughout human history. cess of the linguistic novelty is most appropriately mea- The existence of language isolates (each effectively a sured by its proliferation. monotypic phylum) also complicates resolution of the In contrast with genes, however, linguistic "muta- prehistory of the world's languages. Ruhlen (1987) recog- tions," because they are acquired, can be disseminated nises only 9 isolates that cannot be accommodated in his throughout a human population in a single generation. 17 "families and phyla" (Basque, Burushaski, Etruscan, Consequently, as is noted by Diamond (1988:622), there Gilyak, Hurrian, Ket, Meroitic, Nahali, and Sumerian), are many recent examples of the transfer of languages though many other linguists would considerably expand with little associated transfer of genes. Two or more lan- this list. These languages are either spoken by single guages can be maintained concurrently, both by an indi- isolated populations or long extinct and are probably the vidual and by a society, though in most circumstances remnants of a much larger number of extinct (and there- one language competitively displaces the other(s). More- fore largely undocumented) languages that would have over, languages need not be transmitted as complete sets been spoken by members (or forebears) of the 42 putative of words, analogous to genes in a gamete; each word can extant populations but would not have been assignable function as an independent entity. This characteristic to any of the 17 extant phyla. allows borrowing of linguistic elements, a process simi- The argument that language replacement in Europe lar to biological hybridisation. As in biology, geographi- was exceptional, reflecting the emergence of elites dur- cal proximity of the parental stocks aids this hybridisa- ing the last 5,000 years only (Renfrew 1987; Cavalli- tion (Hoenigswald 1987:262-63), but unlike biological Sforza et al. 1988:6005; Diamond 1988:622), is weak- entities, closely related languages are not inherently ened by the existence of widespread shallow linguistic more likely to hybridise (cf. Stace 1975; Wiener groupings in band-level societies such as Inuit-Inupiaq 1987:225). One of the few parallels between the evolu- (Beringian Diomede Islands-East Greenland [McGhee tionary constraints on human populations and those on 1984, Woodbury 1984]), Pama-Nyungan12 (most of Aus- languages that withstands close scrutiny is the shared tralia [Dixon 1980]), and Numic (Great Basin of North lack of intrinsic barriers to hybridisation and consequent America [Miller 1986]). Moreover, if the recent distribu- reliance on extrinsic (geographical and, especially, be- tion of languages does partly reflect dissemination by havioural) isolation. Whereas biological hybridisation in- elites, correlation between languages and populations evitably involves the transfer of a complete set of gene- cannot be expected. For example, Cavalli-Sforza et al.'s bearing chromosomes, hybridisation of languages tends European population includes speakers of Basque, a non- to be confined to specific elements and therefore has Indo-European language, and speakers of three branches more appropriate biological analogues in "jumping of Indo-European (Germanic, Italic, and Hellenic). Since genes" (e.g., Campbell 1983) and virus-mediated gene the spread of Indo-European languages across Europe transfer than in hybridisation of species. Opinions differ presumably postdated the establishment of the Euro- on the extent to which some words and concepts consti- pean population,13 available data suggest that Basque tute an inner core that is much less prone to being bor- rather than Indo-European should be regarded as the rowed and therefore much more likely to contain reten- original language of the European population. Even if tions (cf. Ruvolo 1987, Wang 1987). Although borrowing only the Indo-European-speakers in Europe are consid- generates homoplasies throughout language as a whole, ered, correlation between the branches of Indo-European it does create innovations within particular linguistic and any putative populations within Europe is poor. lineages. Hybrid words or, more commonly, phonolog- Furthermore, the English component of the European ical elements (mergers) can also function as innovations population has shifted its language twice within re- in speech communities (Hoenigswald 1987). However, cent millennia, thus further obscuring recognition of an large-scale borrowing can obscure relationships of lan- "original" European language. guages, especially if the borrowing is reciprocal. Ele- Similarly, the present languages of Australia probably ments of different parental languages may then become originated much later than the date of 40,000 years B.P. difficult to distinguish, reflecting a process analogous to estimated for the earliest human occupation (Dixon genetic introgression (e.g., Stace 1975:3). Borrowing and 1980:19); any distinct genetic and linguistic characteris- convergences perturb hypothetical, neatly dichotomous tics presumably largely reflect the relative geographical linguistic phylogenies and create anastomoses (Wang isolation of the island continent. Where geography con- 1987), certainly at the level of linguistic elements and strains migration to a narrow route (e.g., through the possibly at the level of languages themselves (Thomason and Kaufman 1988). 12. Dixon (1980) has suggested that Pama-Nyungan may not be a To summarise, most evolutionary processes in lan- bona fide linguistic entity, as some languages of divergent types guages appear fundamentally different from their ana- may be derived from the otherwise rather uniform Pama-Nyungan logues in living organisms. Hence, evolutionary patterns languages. If he is correct, the point made here is reinforced. may also be different, and the congruence between hu- 13. This would be true even if one accepted Renfrew's (1987, 1988) date for the introduction of Indo-European into Europe of ca. 6500 man genodemes and languages sought by Cavalli-Sforza B.C., regarded by Indo-European specialists (e.g., Jasanoff 1988) and et al. may be an unreasonable expectation. other critics (Gimbutas 1988, Greenberg 1988, Lamberg-Karlovsky Any primary congruence has been further disrupted by 1988) as untenably early. io I CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

PRESENT

DEVELOPMENT OF CURRENTLY RECOGNISABLE LANGUAGES T I M E

ABILITY TO SPEAK

FIG. 3. Hypothetical cladogram of human genodemes (A—J) in 'which languages evolved after the ability to speak. Dotted line, times at which different parts of the cladogram (solid black) developed languages independently.

Beringian bottleneck between Asia and North America), it emerged at the same time as the only subspecies character distributions are likely to be clinal. In both known to employ it. Laitman (1985), however, con- cases, language groupings are considered more recent cludes that the earliest H. sapiens (ca. 300,000 years B.P.) than the human occupation of the areas in which they possessed the anatomy necessary for human speech. If, are spoken. Furthermore, Takahata and Nei (1985) and as he suggests, the evolution of the organs of speech Spuhler (1988:24) argue that gene divergence appreciably conferred selective disadvantages (notably greatly in- pre-dates population (i.e., racial) divergence, suggesting creased vulnerability to choking), these must have been a consistent progressive order of temporally distinct outweighed by even greater selective advantages. It events: initial gene divergence -» population divergence could be argued that adaptation for another biological —> linguistic divergence. Increasing the time lags be- function preadapted vocal organs for subsequent linguis- tween events decreases the likelihood of generating con- tic use (thus increasing the likelihood that speech per se gruent phylogenetic and linguistic trees. is polyphyletic), but a convincing alternative role has Establishing a historical correlation of phylogeny and not been proposed (see Davidson and Noble [1989] for linguistic relationships presupposes the ability to estab- speculations concerning the origin of language). lish the absolute time depth of both phylogenetic and We agree with Cavalli-Sforza et al. (1989) that a large linguistic divergences. Unfortunately, even if the cur- temporal gap probably separates the evolution of the rent speculative divergence dates for the human popula- ability to speak and the emergence of what we now rec- tions are accepted, there is no generally accepted compa- ognise as distinct languages. If so, divergence of human rable method of dating linguistic divergences. genodemes pre-dated the appearance of languages in The physical ability to speak is generally regarded as a these populations and language is, by definition, poly- monophyletic trait (e.g., Greenberg 1987), though monophyletic (fig. 3). This would not preclude the construc- phyly cannot be assumed for language itself (Wiener tion of a tree connecting the independently evolved lan- 1987:220). Mellars (1988) and Cavalli-Sforza et al. (1988, guages such as that of Cavalli-Sforza et al. but would 1989) argue that human language emerged coevally with render attempts to discover historical relationships be- modern man [Homo sapiens subsp. sapiens) ca. 100,000 tween these unrelated entities entirely artificial. Even if years B.P., on the assumption that language as we know adaptation in early H. sapiens specifically for speech is BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | u

likely, the implied time depth for the origin of language tion and their role in evolution (e.g., Hull i988:chap. n). would place most of the history of the world's languages Dawkins (1976) has distinguished between the narrow far beyond the reach of available analytical techniques. concept of molecular genes and the broader concept of Many divergences are probably too remote to recon- evolutionary genes by terming the latter "replicators." struct their shared protolanguage, thus restricting inter- A replicator has structure, longevity, fecundity, and pretation to sister-group relationships (e.g., Platnick and fidelity and is defined by Hull (1988:408) as "an entity Cameron 1977, Farris 1983), which are very difficult to that passes on its structure largely intact in successive assess in the absence of a temporal framework. replications." It is contrasted with the concept of "in- One of the most important factors underpinning com- teractor" (Dawkins's [1982a] "vehicle"14), "an entity parison of the evolution of human populations and lan- that interacts as a cohesive whole with its environment guages is constancy of rate of change. Although the in such a way that this interaction causes replication to Nei-type genetic phenogram (fig. 1) is interpreted be differential" (p. 408). Hull then defines selection as "a phylogenetically by Cavalli-Sforza et al. (1988,1989), Di- process in which the differential extinction and prolifer- amond (1988), Lewin (1988a), and Gould (1989), they rec- ation of interactors cause the differential perpetuation ognise that it is dependent on a general assumption of of the relevant replicators" (p. 409). Replicators and in- constant evolutionary rates in human populations. Un- teractors are types of wholly interdependent entity that fortunately, Nei's genetic distance is non-metric and share many properties. Both are individuals (spatiotem- therefore cannot generate a phenogram with branch poral particulars) of finite duration and cohesive to the lengths that would accurately reflect constancy of rate extent that they are incapable of indefinite change (Farris 1981; Ruvolo 1987:200). Moreover, the assump- through time. Both also have active roles in evolution, tion of rate constancy remains controversial (cf. Farris though only interactors function in the selection pro- 1981; Kimura 1983; Gingerich 1986; Spuhler 1988:23), cess. Neither is restricted to a specific level of biological especially when the genetic entities being compared organisation. are not species but local or regional populations In complex, allogamous species such as man, the most (gamodemes) lacking intrinsic barriers to gene flow. obvious replicators are genes (e.g., Dawkins 1976), and (Both Cavalli-Sforza et al. [1988] and Gould [1989] selec- the most obvious interactors are individual organisms tively reject the constant-rate hypothesis for certain pos- (e.g., Hull 1988), though some biologists view entities tulated racial divergences on the phenogram that are lower (e.g., genes [Dawkins 1982a, £>]) or higher (e.g., contradicted by migration hypotheses based on dated avatars15 [Damuth 1985]) in the bio-organisational fossil assemblages.) The lack of intrinsic controls on the hierarchy as the prime interactors.16 As replicators pro- evolution of languages has undoubtedly precluded the duce sequences of themselves through time (faithfully or development of a "linguistic clock"; rates of change in with modification, with or without interaction), they languages cannot be assessed retrospectively. generate a lineage, a third type of entity that "persists Trees for both human populations and languages will indefinitely through time, either in the same or an al- lack most ancestral stocks, whose presence cannot be tered state, as a result of replication" (Hull 1988:409). A assumed. Many extinctions of both races and languages lineage is sufficiently incohesive to change indefinitely undoubtedly occurred. Thus, if repeatedly dichotomous but sufficiently cohesive to maintain its historical con- trees are a reasonable representation of the patterns of tinuity. evolution in human populations and languages, they These three concepts of biological entity (replicator, should be constructed using algorithms that are inde- interactor, lineage) provide a context for the similarly pendent of time and focus on topologies reflecting char- fundamental (though rarely critically analysed) concep- acter transitions at nodes rather than phenetically de- tual distinction between entity and attribute (cf. Temple termined lengths of branches. The application of such 1982). Most scientific research involves investigating cladistic algorithms would require the identification of the relationships between entities by recording their inherited homologous characters (preferably including attributes. In a data-matrix, the entities are represented autapomorphies for each entity) and their subsequent by data-sets of values for one or more (typically many polarisation into transformation series of increasingly derived character states (e.g., Platnick and Cameron 1977, Farris 1983). Difficulties inherent in applying 14. The term "vehicle" is ambiguous; it has been applied to other concepts in linguistics (Richards 1936:96) and to non-biological these techniques to genetic (e.g., Dunn and Everitt replicators in conceptual evolution (Campbell 1979; Hull 1982:131-37) and linguistic (e.g., Goddard 1975:255) 1988:414). data will only be overcome if they are seriously ad- 15. Damuth's (1985) avatars are groups of organisms of a single dressed. species in a single geographically restricted community and therefore encapsulate three deme concepts: ecodeme, topodeme, and genodeme (see Gilmour and Heslop-Harrison 1954}. r6. Eldredge (1989:140) preferred to award primacy to reproduction Replicators and Interactors, Entities (" 'moremaking' of entities of like kind") rather than replication, and Attributes arguing that "both interaction and reproduction must occur within a single category of entity at a given level [in the bio-organisational At a deeper level, the issues raised by Cavalli-Sforza et hierarchy] for selection truly analogous to natural selection to be said to occur at levels higher (or lower) than the organismic level." al. reflect one of the most fundamental controversies in By this definition, entities more inclusive than single organisms evolutionary biology: the identification of units of selec- can be subjected to selection s.l. but not natural selection s.s. 12 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

more) potentially shared attributes. The only constraint phenotype (Dawkins 19820, b). If so, languages are on an attribute is that it be less inclusive than the entity merely attributes of humans and can in theory be used to that it describes. This characteristic has two important infer human relationships, either independently or by com- implications. First, it offers the scientist considerable parison with patterns exhibited by other types of attribute. flexibility of approach within the organisational hierar- Although the possession of speech is reflected ana- chy of entities; for example, the same attribute can be tomically and confers strong selective advantage, the recorded for an individual organism, a deme, or a species. possession of a specific language is much more prone to Second, an attribute can be defined only relative to a conscious manipulation. Human languages represent particular entity. A systematist may use morphological conceptual systems that can be selected (in their en- features as attributes to characterise species-level en- tirety or in part) and adapted by particular groups of hu- tities, but an ecologist may use the resulting species as mans. The selective advantage in possessing language attributes of more inclusive entities such as com- per se is generally much greater than for other concep- munities. Thus, an attribute is not conceptually distinct tual systems such as religious or political allegiances, from an entity; rather, it is a temporary methodological but the adoption of a particular language is determined status awarded to entities that are used to investigate by cultural pressures similar to those that determine the relationships of other, more inclusive, entities.17 adoption of a particular religion or political stance. Although they are complex and difficult to convey, we Thus, it is more difficult to regard a human language as a believe that these philosophical concepts provide vital phenotypic character than to so regard, for example, insights into whether evolutionary patterns and pro- pair-bonding in birds, which may be at least partly inher- cesses in human populations are homologous or merely ited. We doubt that a Frenchman is born with a genet- analogous and therefore whether or not they are poten- ically induced predilection for the French language (cf. tially reconcilable. Wang 1987). The phenogram presented by Cavalli-Sforza et al. os- Also, as is noted by Cavalli-Sforza and Feldman (1981), tensibly depicts relationships between lineages of hu- Wiener (1987:219), and Wang (1987), language transmis- man populations, using overall similarities derived from sion has a vertical (parental, phylogenetic) component, the quantification of selected genes. The probable prime but this may be subordinate to oblique (non-parental replicators (individual genes) are used to assess the his- teacher) and horizontal (peer-group) components. torical relationships of aggregates of the probable prime Because languages do not reside in the genome, they interactors (individual humans). The status assigned to are freed from the severe constraints that restrict the the linguistic phyla is much more ambiguous. Cavalli- occurrence of genetic innovations and the speed of their Sforza et al. present human populations and linguistic subsequent dissemination through biological popula- phyla as separate lists of entities that can be directly tions. Most important, attributes such as language are compared; our discussion of the correspondence of popu- masked from direct selection and selected for indirectly lation clusters and linguistic phyla makes the same tacit by factors such as the social status of their proponents assumption. However, in order to calculate a consis- (Wiener 1987:219, 225). As a particular language is not tency index for their relationship, we treat the languages subject to direct selection, it need not be adaptively as dependent attributes of the populations, potentially superior to other languages to increase in frequency capable of resolving their phylogenetic relationships. (witness the present relative frequencies of an eclectic, Languages undoubtedly form lineages, persisting in- ad hoc historical language, English, and a deliberately definitely through time as a result of replication. A lan- manufactured "universal" [and ostensibly ahistorical] guage also possesses the characteristics of a replicator language, Esperanto). Indeed, it need not compete di- (structure, longevity, fecundity, and fidelity), and its rectly with other languages (in the way that alternative structure persists through successive replications. How- alleles compete for occupancy of a genetic locus), since a ever, unlike a gene, a language is not pre-programmed for single human can assimilate several languages. Conse- replication, and its role during replication is passive; it is quently, we believe that the success of a language, is propagated through the conscious will of its human primarily determined not by the genealogy of its advo- host. Prior to the advent of writing, languages had no cates but by their social influence, assisted by positive independent existence. A language is also dependent feedback that tends to sustain the expansion of a lan- upon its host for its expression and so cannot function as guage whose frequency is already increasing. an independent interactor with its environment to cause We conclude that human languages "evolve" only in differential replication. Languages exhibit differential the most general sense. Many of the superficially attrac- replication (i.e., undergo selection) only as a result of tive similarities between patterns of genetic and linguis- interactions between humans, either as individuals or tic inheritance reflect analogous rather then homolo- as societies (i.e., speech communities). Such cultural gous processes. replicators have been termed "memes" by Dawkins (1976:206), who has promoted the controversial argument that they are features of an extended concept of Conclusions

17. This conclusion renders rather less profound Hull's (1988:522) Satisfactory reconciliation of human phylogeny and lin- argument that "conceptual entities are replicators, not traits" (i.e., guistic history requires several methodological and con- not attributes). ceptual advances: BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 13

1. Increased discourse between linguists and biologists 115-56]). Any benefits gained from simplification of lin- is needed. Distinguishing homologous and analogous guistic patterns with increasing time depth would, how- processes in linguistic and biological evolution would ever, be negated by constraints imposed by lack of docu- allow standardisation of terms and concepts where ap- mentation and inferior time resolution. propriate. 8. The demographic methods that we advocate for de- 2. Objective techniques for quantifying languages are limiting languages and human races are essentially desirable in order to provide data for algorithmic analy- agglomerative, using individual humans as basic oper- sis. Admittedly, some linguists doubt that such tech- ational taxonomic units. They therefore require ac- niques can be successfully applied. quisition of substantial bodies of compatible data and 3. Human populations and languages should be de- preclude a rapid solution to the global problems posed by limited by large-scale demographic census and subse- Cavalli-Sforza et al. We suggest that more intense, geo- quent multivariate ordination. This should reveal any graphically localised sampling confined to regions of positively correlated depressions in frequencies of char- greatest potential offers a much better opportunity to acter states that would reflect meaningful boundaries of reconcile some population and linguistic divergences. entities and thereby minimise the effects of the prob- Promising candidates include the Beringian region lems associated with partitioning continuous variation (Krauss 1988), New Guinea (Wurm 1982, Stoneking, (e.g., Prentice 1986). Bhatia, and Wilson 1986), Australia (Dixon 1980), and 4. The resulting entities could then be characterised China (Yu et al. 1988). Experience and conceptual prog- using large quantities of fully compatible data collected ress achieved during investigations would increase the throughout their geographic and ethnographic ranges. probability of success in even more challenging regions Extant hominids are sufficiently abundant and wide- such as Europe, central Asia, and Melanesia. Never- spread to allow large-scale demographic sampling. Al- theless, satisfactory global resolutions remain a very dis- though the paucity of fossil hominids restricts pa- tant goal. laeoanthropological studies to small-scale typological characterisation, recent technical advances have allowed DNA studies of fossil vertebrates (e.g., Paabo 1989, Paabo, Higuchi, and Wilson 1989), thereby provid- Comments ing a more direct method of tracing the evolution of human populations. 5. The laudable interdisciplinary approach of Cavalli- DAVID F. ARMSTRONG Sforza et al. to the integration of genetic and linguistic Gallaudet University, Washington, D.C. 20002, U.S.A. data should be expanded to include other types of rele- 24 VIII 89 vant information. For example, although previous attempts at morphological analysis of humans have Bateman et al. provide a thorough critique of the com- proved controversial and often been discredited (Gould bined linguistic/phylogenetic approach to analyzing hu- 1981), there is no scientific reason for abandoning the man diversification proposed by Cavalli-Sforza et al. collection of such data. Their critique could, however, be extended a step further 6. Phylogenetic interpretation is facilitated by the ap- to a full exploration of the question when and under plication of inherently historical (i.e., cladistic) al- what conditions the application of tree diagrams to hu- gorithms. They generate phylogenies that can be related man subspecific variation is justified. Dichotomous tree to specific character transitions and are not dependent diagrams may not be the most appropriate models for on assumptions of constant evolutionary rates. All types describing either the genetic or the linguistic histories of of data can be summarised in a single phylogeny, or sev- large proportions of the world's current and former popu- eral phylogenies can be compared a posteriori. However, lations. it may be impossible to determine true congruence be- Cladograms are most appropriate to the description of tween such phylogenies, as the boundaries of different populations that are known to be diverging or that have types of entity (e.g., human genodemes versus languages) diverged genetically, that is, at the species level and are unlikely to coincide precisely and the degree of inter- above (e.g., Rensch 1959). It is arguable whether they dependence between organisms and cultural traits such may also be appropriately applied at the subspecific level as language and behaviour requires clarification. to populations that experience extremely limited gene 7. Such phylogenies are likely to be "noisy," reflecting flow over long periods of time (see n. 4). Humans are the complicating factors such as hybridisation, the less con- most mobile and adaptable of the mammals, and it strained evolution and competitive displacement among is unlikely that any pair of modern human populations cultural traits, and the fact that contemporaneity of hu- is or was headed toward intergroup infertility or specia- man gamodemes and languages cannot be assumed. It tion. could be argued that relatively recent languages are ex- For example, according to the scheme of Cavalli- ceptionally complex and therefore provide a poor anal- Sforza et al., the population of modern Ethiopia repre- ogy for the simpler patterns that presumably charac- sents an anomaly—genetically African but speaking a terised the early evolution of language (particularly if language that belongs to a linguistic group (Afro-Asiatic) dominated by iconic [representational] rather than sym- whose speakers are primarily "Caucasoid." The tree dia- bolic [arbitrary] grammatical processes [Swadesh 1971: gram showing a phylogenetic split between African and 14 CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

other peoples forces Cavalli-Sforza et al. to describe be more appropriate theoretical structures that could be Ethiopians as a primarily African population with "ge- constructed to describe the genetic and linguistic netic admixture." The historical model this implies is diversification of human populations, for example, a re- that, after the split, invading Eurasians came back to ticular structure stretched over a sphere (globe) and Africa bringing their genes and languages. There is an- growing outward through time. Cavalli-Sforza et al. other possible historical model that the tree diagram ^988:6005), referring to Wolpoff, Wu, and Thome cannot account for—that there never was a split suf- (1984), recognize the possibility that continuous net- ficiently complete to justify representation of a work models could be applied to these data but dismiss phylogenetic branch and that, since the expansion of it. Cavalli-Sforza has been producing human phy- modern humans out of Africa, Ethiopians and other logenetic trees, based on merged sets of diverse data, for inhabitants of Northern Africa have been clinally inter- many years, and his efforts continue to be valuable. We mediate between Sub-Saharan Africans and Eurasians. should welcome attempts to rejoin phylogenetic and lin- This model would not preclude the back-migration from guistic data in nonracist formulations, although this ap- Eurasia to Africa—it would accommodate it easily. proach has been in disrepute for much of this century. Ethiopians, on the African side of the Red Sea, are, not However, despite increases in the quantity and quality surprisingly, genetically more like Africans from further of data and the sophistication of methods, questions south on the continent, and Arabs, on the Eurasian side, raised previously (Barnicot 1964:204-5) have not been are more like other Eurasians. In fact, the question is fully answered. open whether major populations of Homo sapiens sapiens have ever been separated to the extent and over the periods of time that would be required to justify the DONN BAYARD representation of their current diversity or history Department of Anthropology, University of Otago, through the use of tree diagrams, whether phenetic or Box 56, Dunedin, New Zealand. 11 vm 89 cladistic. This, along with recognition of the need for vigilance against racist interpretations, was the funda- My research interests are limited to the linguistic and mental observation that underlay the anthropological at- archaeological prehistory of Southeast Asia and Oceania tack on the race concept that was mounted a quarter- over the past 7,000 years or so and the sociolinguistics of century ago (e.g., Montagu 1964). contemporary New Zealand. Hence I cannot comment The question is not whether trees can be fitted statis- on Bateman et al.'s detailed genetic evidence and such tically to the data but whether they represent the most concepts as "topodemes," but my own experience cer- appropriate models. Bateman et al. point out that anasto- tainly leads me to agree with their refutation of Cavalli- moses of linguistic stocks call into question the applica- Sforza et al.'s grand scheme. tion of tree diagrams to linguistic diversification. (The 1. I have been suspicious of phenograms as "God's same can equally be said of genetically defined popula- truth" since cranial analyses from my excavations at tions.) The origin of English provides a familiar example Non Nok Tha, Thailand, yielded completely different of the problem. It is common to place modern English at affiliative dendrograms depending on whether metrical the end of a branch of the Germanic group of Indo- or non-metrical variables were employed (Pietrusewsky European languages, because of the known history of its in Bayard and Solheim n.d.). As I commented on a simi- speakers and elements of its core vocabulary and sound lar study of Southeast Asian and Oceanic cranial mor- system. It could be argued, however, on the basis of lex- phology, "It is clear that small variations in sample size ical and syntactic features, that it is really Italic and not or variable selection can produce vastly different clus- Germanic, closely related to modern French. The truth, ters and hierarchies" (Bayard 1987:115). of course, is that it is neither—it is descended from 2. As Bateman et al. point out, only a tiny minority of an amalgam of Anglo-Saxon and Norman French and the world's linguists would accept the validity of en- has evolved subsequently for a millennium relatively tities like Nostratic, Eurasiatic, or even Austric. If lin- independently of other Germanic languages. Gould guists cannot agree on the affiliations of a well-defined (1989:20), in his review of Cavalli-Sforza et al., recog- family like Austronesian with Tai-Kadai or Austro- nizes this point but does not discuss its implications for asiatic (or even Japanese? [Reid 1988:32]), it is difficult to the representation of linguistic evolution by constantly take such larger classificatory entities seriously. branching tree diagrams. The tendency to place modern 3. Given their criticisms, Bateman et al. seem English and French at the ends of two ramified branches strangely hopeful about the rationale for such a scheme having a common ancestor as remote as Proto-Indo- in general but call for the gathering of substantial bodies European grossly distorts the actual relationship of data from geographically localised populations fol- between the two languages and their speakers, and lowed by "multivariate ordination." Sociolinguists the situation of English is not unique in human his- have been doing this for some 25 years now. The results tory. indicate the extreme complexity of linguistic interac- Bateman et al. recognize and discuss the implications tion and transmission among "extant hominids" (why of these and other problems inherent in the use of tree not "people"?) and make it clear that any tidy division diagrams, but more discussion of a philosophical nature of language transmission into vertical, oblique, and is needed, extending beyond the adequacy of statistical horizontal "components" is a gross oversimplifica- measures of genetic distance and "treeness." There may tion. BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic Histoiy \ 15

In short, I am a bit puzzled by the "torchbearer" con- ing of language has undoubtedly advanced, it has left clusion to Bateman et al.'s paper; surely a brief rebuttal linguistics relatively ill-equipped to grapple with varia- would have sufficed. Perhaps I am being overly negative tion and thus with vital issues in the reconstruction of or blinkered, but I would suggest that Bateman et al. phylogeny. douse the torch and concentrate on more realistic and Research methods in behavioral science that directly attainable research goals, say, the affiliation of Japanese address variation are rarely taught in the linguistics cur- or the Austro-Tai hypothesis. As they conclude, any riculum, and statistics courses are conspicuous by their global solutions remain a very distant goal indeed. absence. Even in research that includes variation, as in sociolinguistics, statistics are often absent or poorly understood (see Wardhaugh 1986). Textbooks on statis- BEN G. BLOUNT tics in linguistics have, however, begun to appear (Butler Department of Anthropology and Linguistics, 1985, Woods, Fletcher, and Hughes 1986). Linguists in- University of Georgia, Athens, Ga. 30602, U.S.A. terested in reconciling human phylogeny and the history 16 viii 89 of language should not only converse with biologists but expand their research methodology to include behav- Bateman et al. make a compelling case that reconcilia- ioral science. This will prepare them to carry out the tion of human phylogeny and the history of language is subsequent recommendations, which require quantifica- premature. As they point out, the lack of a "linguistic tion. It will also increase understanding of biological clock" renders even the superimposition of language transmission (per the introductory quotation). phylogeny on population (genodeme) phylogeny highly Dialogue between biologists and linguists is not, how- problematic. There are no clearly established units of ever, unidirectional. Bateman et al. acknowledge that language that can be used to measure change, at least not the success of a language (survival and spread) is due to any appreciable time depth. As they further note, lan- more to the social influence of its advocates than to its guages, unlike genes, are not constrained to a reproduc- genealogy. Sociolinguists would agree, but they would tive cycle or preprogrammed for replication. Biological add that one of the most robust findings of their inquiry evolution and language evolution are thus substantially is that language is a property of social groups and that different in these terms. the success of a language must be viewed in those terms. One aspect of the difference merits further discussion. Language is an expression of group membership and Since languages are not directly constrained by the varies as an individual invokes membership in different genome, they can undergo radical and extensive change groups. In a sense, the engine that drives language main- in a shorter time span than the genodeme. Assuming tenance and change and thus produces variation is the that pressure for language change is not a rare phenome- way in which groups identify aspects of language and non, a high degree of variation within and diversity enforce them normatively. Language is thus an attribute across languages (although with mitigating factors such not only of individuals but of groups. Language variation as borrowing and convergence) may be expected. Recon- across gender and across ethnic, class, occupational, and struction of language protof orms thus will be more accu- religious groups is enforced and reinforced to the extent rate to the extent that variation is documented. that individuals commit themselves to the norms of Linguists have long been sensitive to the need for his- those groups. The operational taxonomic units in rec- torical documentation, but within a particular perspec- ommendation 8 will have to be individual humans but tive. The prototypical form of language reconstruction, as members of groups. as Bateman et al. note, is a description of structural Concern about language as an attribute of groups change from time 1 to time 2. Historical information would require that the temporal progression of initial allows for more, and shallower, time points and thus an gene divergence —> racial divergence —> linguistic diver- enriched record of the increments of change. Variation gence include social divergence prior to linguistic di- has been viewed by historical linguists principally as vergence. The concept of the differentiated social group either finer gradations on a phyletic branch or points of (sociodeme) as the transitional unit in language change origin for new branches; beyond this its importance has is consistent with Bateman et al.'s recommendation of been neglected. It has not been viewed in the way that cladistics as historical algorithms (6), since changes will biologists typically treat it, i.e., as an integral part of the not be at a constant rate, and it is also consistent with evolutionary process, in which selection acts on the ex- the key recommendation (3) that human populations tant variation in a genodeme. The differences in the and languages be delimited by large-scale demographic ways in which biologists and linguists typically view census and subsequent multivariate ordination to per- variation cut to the heart of Bateman et al.'s conclud- mit the detection of meaningful boundaries of entities. ing prescriptions. Linguists typically work with languages as ideal types—not unmindful of the existence of variation but CATHERINE A. CALLAGHAN recognizing that the notions of ideal speakers, languages, Department of Linguistics, Ohio State University, and communities remove the clutter of social and his- Columbus, Ohio 43210, U.S.A. 9 vm 89 torical factors from language. Major theoretical positions in linguistics from Bloomfield to Martinet to Bateman et al. are to be congratulated for their cogent Chomsky have adopted this stance. While understand- multidisciplinary response to the attempts by Cavalli- 16 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

Sforza et al. (1988, 1989) to correlate possible population similarities between the daughter languages will have a phylogeny with the deep genetic groupings proposed by nongenetic explanation. The daughter languages will Ruhlen (1987) and Greenberg (1987). They have per- have become unrelated (or "disrelated"), and it will be formed a special service by delineating the weaknesses, impossible to recover their common history. errors, and inconsistencies underlying such amalgama- tions as "Amerind." In this regard, some readers may be unaware of the L. L. CAVALLI-SFORZA, A. PIAZZA, P. MENOZZI, vast differences among Amerindian languages. My own AND T. MOUNTAIN research has focused on the Utian (Miwok-Costanoan) Department of Genetics, Stanford University, Stanford, languages of Central California. I would estimate that Calif. 9430s, U.S.A. (Cavalli-Sforza, Mountain)/ Southern Sierra Miwok and Mutsun Costanoan are Dipartimento di Genetica, Universita di Torino, roughly as far apart as early Latin and Sanskrit. Through Torino, Italy (Piazza)/ Istituto di Ecologia, Universita the discovery of regular and recurrent sound correspon- di Parma, Parma, Italy (Menozzi). 31 vm 89 dences involving stems and affixes, I have been able to prove Utian to be a genetic unity, and I have recon- In this article and in their earlier letter to Science structed a fair portion of the protolanguage. When we (O'Grady et al. 1989), these critics express their reserva- compare Proto-Utian with Yokuts (another "California tions concerning our database, methods of analysis, and Penutian" family of languages), genetic kinship is al- interpretations. In the following comments we counter ready controversial despite some striking similarities. their concerns and point out that their suggestions are The nearby Porno languages are utterly unlike Utian, unrealistic and unnecessary. Indeed, many of their criti- and this without our even having left Central California. cisms surprise us; for example, they have accused us of Greenberg disdains comparison based on regular using gene frequencies based on "very small samples, sound correspondences because they will disappear after sometimes of single individuals" (O'Grady et al. 1989). a given period. Yet they allow us to identify Southern Presumably they are unfamiliar with the data, methods, Sierra Miwok kil-a- 'liver' and Mutsun Costanoan sire and models of human population genetics. In fact, apart 'liver' as perfect cognates even though they do not share from one linguist (whose negative comments we know a single phoneme. Reconstruction allows us to weed out will be answered by Greenberg), they are biologists spurious resemblances such as Sioux ni- and Cheyenne whose expertise seems to be in taxonomy and the sys- ne- 'second person pronominal prefix'. Greenberg may tematics of various zoological and botanical species. well be right that erosion will ultimately remove the Concepts and methods for the study of interspecific dif- evidence of regular sound correspondences, but erosion ferences are not necessarily suitable for that of will even more quickly invalidate the vaguely perceived intraspecific ones, especially in a young species such as similarities of mass comparison. modern humans. Because space is limited, we focus on One of the principal contributions of this article is a the criticisms we consider most relevant. systematic discussion of the dissimilarities between the Data. Bateman et al. state that our data "do not inspire phylogeny of organisms and the evolution of languages. confidence." Human population geneticists are familiar We know that cats are distantly related to catfish be- with the large body of data we use. Tabulations from cause both groups possess backbones and the other three books (Mourant, Kopec, and Domaniewska- structural characters of vertebrates. Languages appar- Sobczak 1976, Steinberg and Cook 1981, Tills, Kopec, ently lack such nonuniversal markers of remote genetic and Tills 1983) and 2,815 papers, in part overlapping relationship. (As an aside, I must admit that the Utian with those abstracted in the books above, were com- languages are characterized by metathesis as a mor- puterized. From these data on ca. 3,000 "populations" (a phological process, imperative verbal inflection, and statistical and genetic term) identified by the authors of suppletive imperative verbal forms, none of which is a the original publications, geographic gene maps and universal attribute of language.) The traits that define hundreds of distributions of gene frequencies and sample linguistic subgroupings are subject to erosion through sizes were generated. These will be presented in a book leveling, replacement, and phonological change. Lan- now in the final stages of preparation. For multivariate guages may also borrow extensively from unrelated lan- analysis, the 3,000 populations were reduced to 42 by guages, while exchange of genes between distantly re- eliminating populations and genes too poorly repre- lated species is rare (although gene exchange is common sented and by condensing populations that share an between the human populations cited by Cavalli-Sforza ethnic (or, in a few indicated cases, a linguistic) name et al.). and geographic area. We may never obtain the "perfect" Given the susceptibility of the nonuniversal features body of data that Bateman et al. call for, but the existing of language to erosion through sound change and other collection of data on blood groups, protein and enzyme factors, we must question whether it makes any sense to polymorphisms, etc., which is extremely rich, should postulate genetic relationship at very remote time certainly be analyzed thoroughly. depths. After several thousand years, it may well be that Definition of populations. We do not confuse "to- all or most of the nonuniversal attributes of the parent podemes," "glottodemes," "genodemes," and "gamo- language will have disappeared from all or all but one of demes." The last two concepts are impractical to use the daughter languages. At this point, any nonuniversal with human populations; the definition of a Mendelian BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic Histoiy \ 17

population, the evidence of mating patterns, and the imum parsimony to gene frequencies. We too distrust nearly continuous nature of spatial variation (Cavalli- the uncritical assumption of constant evolutionary Sforza and Bodmer 1971) make it impossible to establish rates. We have shown, however, that at least for the nonarbitrary boundaries. Bateman et al. wrongly deny major fissions genetic distance is proportional to the importance to the correlation between geographic and time since separation. Our method of tree reconstruc- genetic distance (e.g., Morton 1982, Cavalli-Sforza 1984). tion therefore provides a cladistic answer. Methods for Barbujani, Oden, and Sokal (1989) have recently devel- species phylogenesis are not always directly applicable oped a strategy to search for boundaries of gene frequen- to intraspecific data; for example, our critics will find it cies; its value for building trees remains to be tested. hard to apply the search for synapomorphies to gene fre- Missing data. Even though we aimed to maximize the quencies. They also confuse population trees with gene number of genes per population, 23.7% were missing trees such as those built for the mutational history of from our final data set of 42 populations and 120 genes. mitochondrial DNA. Methods and conclusions applica- Bateman et al. state that we violate a "general theoreti- ble to the former do not necessarily apply to the latter cal principle" for which they refer to Temple (1982), but (Nei 1987). Finally, we are accused of overlooking an- he merely suggests striking out all incomplete rows and thropometric data; we have omitted them because, un- columns until one reaches an acceptable percentage of like genes, they are overly sensitive to selective factors gaps, "say, 5%." This is a suggestion, not a general theo- such as climate (Matessi, Gluckman, and Cavalli-Sforza retical principle. By following it with our data set one 1979, Piazza, Menozzi, and Cavalli-Sforza 1981). Neutral would lose too much information. Instead, we have genetic markers are best for reconstructing evolutionary treated the problem innovatively with the bootstrap (Ef- history (Cavalli-Sforza and Edwards 1967). ron 1982, Felsenstein 1985). By omitting genes randomly Congruence of genetic and linguistic trees. Three with replacement one estimates the extent of statistical things struck us in the comparison of the genetic tree variation due to gene sampling, including that due to and the linguistic classification: (1) Almost all the lin- missing data. The bootstrap fills a very significant gap in guistic phyla (= families) correspond to low-level ge- the analysis of trees,- we have used the approach to eval- netic clusters or single populations, indicating [a) that uate the compactness of clusters and, with it, the valid- the evolution of phyla was relatively late and [b] that ity of tree nodes. The method of comparing the fre- successive language replacements were insufficient to quency of clusters observed in bootstraps with that blur the general pattern. (2) The major exceptions to this expected if there were no clustering is derived from stan- rule can be satisfactorily explained (see below). (3) The dard application of the binomial distribution. few proposed linguistic superfamilies show correspon- Partitions. Bateman et al. consider a partition into six dence with higher-level genetic clusters. clusters "methodologically incorrect." We named ten Consistency index. Bateman et al. compare the ge- clusters simply for expository purposes. They recom- netic and linguistic trees with the consistency index, mend that partitions be based on transverse dissection of ordinarily used for measuring optimality. They obtain a the tree at specific genetic distances. This rule assumes value of 0.48, which they declare low, and state that unrealistic accuracy of genetic distances and constant "over half of the linguistic data do not fit." The follow- evolutionary rates. Moreover, we have repeatedly said ing errors contribute to this evaluation: (1) They choose that we are not interested in taxometric classifications to keep the consistency index (as well as their measure or their racial implications (see also Cavalli-Sforza and of congruence) low, e.g., by rejecting "autoapomorphies" Edwards 1964, 1967; Edwards and Cavalli-Sforza 1964). (Brooks, O'Grady, and Wiley 1986), because "phyla They further criticize "a posteriori" pooling of five Euro- confined to a single population are trivially consistent pean populations which show very little genetic differ- with any tree." A trait present or absent in only one ence, but their inference regarding inter- and intrapopu- population may be uninformative for tree reconstruc- lation variation is invalid. They do not consider random tion, but in this application they reject strong evidence sampling variation (Weir and Cockerham 1984), the ef- in favor of congruence of linguistic and genetic evolu- fects of the degree of subdivision on variation of gene tion. The cases they reject are actually many popu- frequencies (Cavalli-Sforza and Feldman 1989), or the lations included under a single label. For example, difference between individual and population variation "Australian" includes many tribes, all of which speak (Nei and Roychoudhury 1974, Lewontin 1972). Their Australian languages. When these non-"autoapo- comparison with mitochondrial-DNA data is invalid for morphies" are included, the six "glottodemes" (popula- these reasons and because of the different system of in- tions defined by linguistic affinity) still being omitted, heritance. the index rises too. $6. Bateman et al. also keep the value Tree reconstruction. Bateman et al. rely on most- low by considering one language replacement as two parsimonious methods of tree reconstruction only. For events. After correction for this, the index increases fur- metric traits such as gene frequencies, simulations of a ther to 0.74 (the maximum being 1.0). (2) Even the 0.48 simple evolutionary model (Astolfi, Kidd, and Cavalli- value they calculate indicates that the two trees are Sforza 1981) have shown that a most-parsimonious highly consistent. Archie (1989) has shown that the con- method does not fare as well as maximum-likelihood or sistency index is negatively correlated with the number average-linkage methods. No clear genetic model with of taxa (populations). For a sample of 28 maximum- testable predictions supports the application of max- parsimony trees based on real data, the index decreases 18 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990 from ca. 0.8 for 10 taxa to ca. 0.2 for 40 taxa. The latter 1988 (organized by L. L. Cavalli-Sforza, A. Piazza, P. case corresponds most closely to ours (42 taxa). As 0.48 Ramat, and W. S-Y. Wang). is much higher than 0.2, Bateman et al. have inadver- The central question is why there should be any con- tently shown that there is indeed remarkable agreement gruence between genetic and linguistic evolution. The between the genetic and the linguistic tree. (3) They remain reason is that the two evolutions follow in princi- ject a significance test of the index with the specious ple the same history, namely, sequence of fissions. Two argument that "the tree . . . represents a sample size of populations that have separated begin a process of differ- one." One could, however, generate random permuta- entiation of both genes and languages. These processes tions of the languages and calculate the consistency in- need not have exactly constant evolutionary rates, but dex for each permutation. In this way one could obtain rough proportionality with time is a reasonable expecta- the sampling distribution of the index, given the null tion for both. They should therefore be qualitatively hypothesis of zero consistency, and proceed to a signifi- congruent, except for later events such as gene flow or cance test of the observed index. language replacement. These may blur the genetic and Nostratic and Eurasiatic superphyla. The Nostratic the linguistic picture, but our conclusion is that they do and Eurasiatic superphyla are recent classifications by not obscure it entirely. independent researchers. Therefore it is not surprising Bateman et al. believe that future research at the re- that they differ. They are likely to be modified and ex- gional, rather than global, level may tell us if there is any tended in the future, and it is reasonable to assume that correlation between linguistic and genetic evolutions. fully congruent hierarchies will eventually arise. We Such research already exists (see, e.g., Spuhler [1979] on stated that there is remarkable agreement between the North Amerinds, Greenberg, Turner, and Zegura [1986] genetic tree and the union of the two sets formed by on the Americas, and Sokal [1988], which lists other Nostratic and Eurasiatic, to which later work has added examples, on Europe). Two very detailed unpublished Amerind languages. The counterevaluations of our crit- investigations also give highly significant correlations. ics do not address this point. One, on Sardinia (Piazza et al. 1989), shows a correlation Rotations of tree nodes. Nodes were not rotated "to of 0.80 between genetic and linguistic data; the other achieve maximum apparent congruence." The aim was (Barrantes et al. 1989) gives a correlation of 0.74 for instead just the opposite, to emphasize discrepancies Chibcha-speaking groups of Panama and Costa Rica. such as the association of Ethiopians with Caucasoid Our analysis demonstrates that the two evolutions are speakers of Afroasiatic languages. When this was not remarkably congruent at the global level and generates technically feasible, e.g., for Tibetans, we had to high- results of greater time depth than is possible with re- light the exception with a star or discuss it in the text. gional studies. Discrepancies between the two trees. We listed and discussed all the discrepancies between the two trees, postulating that they were due to language replacement, JOSEPH H. GREENBERG a historically well-known phenomenon. In some in- Department of Anthropology, Stanford University, stances it may be difficult to distinguish this explana- Stanford, Calif. 94305, U.S.A. 22 vm 89 tion from the complementary hypothesis of gene replacement. The anomalies represented by Ethiopians My comments are concerned with the rejection by Bate- and Lapps are far from being inconveniences: both popu- man et al. of much of the linguistic classification found lations are mixtures of two genetic clusters, belonging in Cavalli-Sforza et al. (1988). The agreement with the linguistically to one and genetically more to the other. genetic data will not be within the scope of these com- The hypothesis that when two populations speaking dif- ments, although others have found it impressive. ferent languages mix only one language will survive is Bateman et al. perceive such difficulties in distin- certainly plausible. The most likely scenario for the pro- guishing historically significant resemblances from acci- cess is gene flow from a neighboring population. This dental convergences that it becomes difficult to see how has been observed historically (e.g., for Black Americans) linguistic classification can get started. "Perceived lin- and can easily generate the observed proportions of ge- guistic similarities (whether they reflect phonology, netic admixture in realistic periods of time. Admixture grammar, semantics, or syntax) have a highly prob- evaluations are standard (Cavalli-Sforza and Bodmer lematic relation to historical connections." In this con- 1971), and the method used is described by Wijsman text terms such as "phonology" are far too vague. They (1984). disregard a fundamental property of language, namely, Points of agreement with Bateman et al. We have ex- the in-principle arbitrariness of the relation of sound to pressed ideas similar to those of Bateman et al. on the meaning in specific forms, whether lexical or grammat- Basques (Cavalli-Sforza 1988, Piazza et al. 1988). We join ical. Moreover, languages have large numbers of such several other participants in this discussion in pleading signs which are essentially independent of each other. It for advanced quantitative research in evolutionary lin- makes sense to distinguish, as typological and without guistics. We agree with these critics on the importance necessary historical significance, resemblances in sound of reciprocal education of biologists and linguists. Two only (e.g., tonal systems in Southeast Asia and in Africa) of us have promoted workshops specifically for this pur- and in meaning only (e.g., gender in Chinook and in pose, at Stanford in 1984 (organized by L. L. Cavalli- French) from resemblances involving sound and mean- Sforza, M. Feldman, and W. S-Y. Wang) and at Turin in ing simultaneously (e.g., German gut, English "good" BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 19

and Hebrew -t and Arabic -t, both marks of the feminine of the term "phylum" would be a methodological ad- gender). It was the elimination of typological criteria vance, since it merely adds a rhetorical notion of re- that was the single most important factor in the suc- moteness and suggests that a different and more myste- cess of my African classification. The vast majority of rious methodology is involved. convergences are typological. Those involving sound- Finally, the statement "These difficulties may not be meaning similarities must be diligently sought. Far from insuperable, particularly if initial studies sample a re- being handicapped in relation to biological classifica- gional rather than a global catchment" is but one more tion, linguistics has great advantages because of the basi- instance of the strange belief that the narrower one's cally arbitrary relation between form and function and factual basis, the more successful one is likely to be. The the number of essentially independent form-meaning enormously widespread n first person and m second per- sets. son in the Americas as against m first person and t sec- The concrete example that Bateman et al. adduce is ond person in Europe and Northern Asia is powerful second person singular Sioux ni, Cheyenne ne, and evidence which requires an explanation. Bateman et al. Ojibwa ke, in which the Algonquian languages Chey- have neither the capability for making such observations enne and Ojibwa differ while Cheyenne agrees with nor the type of theory which can account for them. Sioux. Such instances fool no one. In Greenberg (1987a) we find Algonquian k, second person, and I did not con- nect the Sioux and Cheyenne forms. How did they de- KENNETH JACOBS tect it? They used the comparative method, but this im- Departement d'Anthropologie, Universite de Montreal, plies the prior classification of the Algonquian languages C.P. 6128, succursale A, Montreal, Que., Canada H3C as a group within which k is obviously normal and n 3/7. 4 vm 89 deviant. According to them, linguistics cannot classify until it has distinguished true resemblances from con- This article, neither fish nor fowl, leaves me somewhat vergences, but, as we have just seen, this can only be confused as to its purpose. The authors attempt both to done on the basis of a classification which has already critique an article by Cavalli-Sforza et al. and to present been carried out. We are caught in a vicious circle. their own methodological scenario for the establishment Just how frequent are accidental form-meaning con- of patterns of congruence between linguistic and other vergences in language? A whiff of reality dispels this aspects of human evolution ("human" being a term in- supposed difficulty and reveals a truly Gargantuan gap adequately defined). They have clearly underscored an between their theory and reality. In Greenberg (1988:24) important issue: to what extent material-cultural (ar- a table of nine basic words in 24 European languages is chaeological), linguistic, and human morphological data presented. Before one has gotten to the third word the can be employed to reconstruct the genealogy (phy- basic division into Indo-European, Finno-Ugric, and logeny seems optimistic) of recent human populations. Basque appears, and by the fourth word one has the ac- That the data set used by Cavalli-Sforza et al. is ill- cepted subgroupings of Indo-European. All the similari- chosen or the statistical analyses inappropriate may be ties immediately perceivable are incorporated in stan- true; I have not the training to judge. dard Indo-European etymological dictionaries, and some One aspect that troubles me, however, is the introduc- that are less obvious are to be found. This is the result of tion into the text of neologisms such as "glottodeme," almost two centuries of investigation by a substantial "topodeme," "entity," and so on, without, to my mind, portion of the linguistic community. One cannot reject adequate definition. Especially in a journal that reaches all these obvious similarities as illusions without de- out to all kinds of anthropologists, a simple reference to stroying comparative linguistics. the article in which the term was coined seems, at best, From the qualification "s.l." [sensu lato) after "Na- insufficient. A footnote defining terms might have aided Dene" it appears that the authors must still doubt the comprehension. affiliation of Haida despite Greenberg (1987a.chap. 6; cf. As a biological anthropologist I am especially troubled also 19876.651-52). Levine (1979), a supposed refutation by the irregular and ambivalent use of the terms "race" of Sapir, has been cited by Goddard (1986:191) as the and "racial." At times, these terms appear in quotes, kind of critique of Sapir's work that is needed. As I have implying a certain disdain for the concept that there are shown, even after applying Levine's largely unreason- biological races. At times, "racial entities" are accepted able criteria 17 etymologies remain linking Haida, Tlin- as real and scientifically definable. This latter point of git, and Athabaskan and 14 linking Haida and Athabas- view is reinforced when we read that "several tech- kan. The same criteria applied to three branches of Indo- niques allow objective delimitation of genetic races." European leave only 6 by a liberal application of Levine's While perhaps being a useful means for the authors to rules, probably none if they are applied strictly. This divide their own data base, "race as such," as Harrison shows that, if anything, Na-Dene as proposed by Sapir is (1988:326) has commented, "explains practically noth- a more tightly knit group than Indo-European and that ing." Biological anthropologists having increasingly the methods approved by Goddard would refute even given up the racial paradigm, a fuller explanation of how Indo-European. It also shows that the family-phylum we can set the standards for an "objective" linguistic distinction the authors wish to maintain, by which Indo- cladogram would have been welcome. European is a family while Na-Dene is a phylum, is a The authors are to be commended for their efforts. No property not of languages but of linguists. The abolition entente cordiale should be expected, however, since the 20 | CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

modes of transmission of material and linguistic culture, be to compare data sets from two groups of populations much less those of the genome, are only cursorily and at two different but contiguous times using phenetic confusedly discussed. Without more detailed discussion, methods and to repeat the comparison for each past time the very tempting and important questions they broach span. This is possible to some extent for morphological will remain unanswered. Still, they have made a first, if and linguistic data and may soon be so for DNA data. halting, step. Linguistic inheritance is different from genetic inheritance in some ways, to be sure, but linguistic data still seem also useful for reconstructing the microevolution- YUfl MIZOGUCHI ary processes and migrations of human populations dur- Department of Anthropology, National Science ing historic times provided that we know the character- Museum, 3-23-1 Hyakunin-cho, Shinjuki-ku, Tokyo istics of the linguistic attributes. Once a series of such 169, Japan. 17 vm 89 two-group comparisons has been carried out, the cladistic procedure will of course also be available. Most of the arguments and conclusions of Bateman et al. Incidentally, the overall correlation between a biolog- seem reasonable. Among the problems they point out, ical tree and a linguistic classification may be estimated however, that of developing objective techniques for and statistically tested by Mantel's matrix permutation quantifying language change has already been partly re- inference procedure (Mantel 1967, Dietz 1983, Dow and solved by mathematical linguistic methods such as Os- Cheverud 1985) if the distance matrices based on such walt's shift test (Yasumoto 1978). Increased discourse data are phenetically estimated. Furthermore, canonical between linguists and biologists will be more easily correlations between biological and linguistic attributes achieved in the near future, but, as the authors them- may be estimated on the basis of ecological correlations selves realize, it is difficult to apply specifically phy- obtained by the use of mean values or some other statis- logenetic algorithms such as cladistic ones to data, tic on each sample, and their significance can be tested whether genetic or linguistic, on modem populations by the above-mentioned bootstrap method. only because of difficulties in distinguishing derived In any case, Bateman et al.'s paper is very stimulating. characters from primitive ones. To trace the evolution- I share their hope that a more accurate human phy- ary process of appearance, retention, disappearance, or logenetic tree can be developed through the integration reappearance of any character, fossil or past evidence is of genetic (both morphological and biochemical) and lin- needed. Although DNA investigations of fossil samples guistic data. have recently become possible to some degree, it will take a long time to accumulate such data on a scale comparable to that of morphological data on fossils. MILTON NUNEZ Problems concerning the delimitation of entities or pop- Bureau of Museums, SF-22100 Mariehamn, Aland ulations, the definition of characters or attributes, and Islands, Finland. 27 vm 89 large-scale sampling are much more fundamental and in practice very difficult to resolve. At an early stage of inves- Bateman et al. have succeeded in presenting a clear, tigation, therefore, preliminary analyses such as those per- thorough discussion of the problems faced in attempting formed by Cavalli-Sforza et al. (1988), even if based on to reconcile the evolutionary history of humans with insufficient data, should be allowed as a stepping stone. that of their languages. Unfortunately, the omission of At present, if changes in attributes such as DNA in key information by Cavalli-Sforza et al. (1988) hinders the course of evolution are regarded as stochastic, the not only the appraisal of their results but also the ade- most likely of possible phylogenetic trees can be deter- quate evaluation of Bateman et al.'s serious objections. I mined by Felsenstein's maximum-likelihood method refer to questions such as whether one can allow 23.7% without any assumption that the rate of evolution is gaps in a data-matrix, even for the noble purpose of max- constant in all the organisms studied (Hasegawa 1988), imizing attributes. The objections of Bateman et al. to and the statistical significance of the maximum loga- such unorthodox procedures certainly seem justified. rithmic likelihood can be tested by Efron's (1982) boot- The same applies to the "population means" used by strap method (Hasegawa 1988). But linguistic and mor- Cavalli-Sforza et al.; these critics are probably right in phological attributes and even some genetic ones are not. predicting greater variation within single populations simple stochastic variables. The construction of a phy- than between different "population means." I suspect logenetic tree on the basis of these attributes therefore this to be the case at least with the Lapps. requires a different procedure. Whether the procedure It should be pointed out, however, that Cavalli-Sforza used is cladistic or phenetic, the distinction of primitive et al. (1988) appear to place great confidence in the abil- and derived characters and their susceptibility to envi- ity of the "bootstrap" method to test the validity of their ronmental factors is very important in interpreting the data. I am not qualified to judge,- nor do Bateman et al. results. The effect of environmental factors may be esti- comment on this new and seemingly useful statistical mated to some degree for some characters, but, as men- tool. The core of the argument lies in whether rigour tioned above, the derivation of characters cannot be de- requirements may be relaxed if results are subsequently termined without historical evidence. Therefore the best submitted to "bootstrap" testing. Cavalli-Sforza et al. procedure for reconstructing phylogenetic lines seems to think so,- Bateman et al. do not. BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 21

The use of recent, albeit controversial, linguistic phyla amount of specialized terminology, as a concession to by Cavalli-Sforza et al. (1988) appears justified in that the general reader, would have been appreciated. the lumping approach of these classifications better fits Figure 1 of Cavalli-Sforza et al. (reproduced as figure 1 the genetic tree. Correspondence with the tree would here) shows a coincidence of tree diagrams based on ge- not have significantly suffered by using the more ac- netic and linguistic factors. Whether the degree is re- cepted classifications, at least in the case of certain con- markable depends on expectations. Some parallels and troversial phyla (e.g., Altaic, Amerind, Na-Dene). Their some divergence are to be expected, for reasons dis- use has basically a cosmetic effect on fig. 1. We must cussed in Bateman et al., who also point out many of the also bear in mind that the controversial phyla and super- inadequacies in the statistical treatment underlying the phyla were derived independently by others and, conse- genetic tree and the controversial nature of the linguistic quently, their correspondence with the genetic tree can- classification at the superphylum level. I agree with not be dismissed without explanation. Furthermore, their criticisms but, rather than speaking in general widely accepted phyla result from traditional methods of methodological terms, will illustrate the development of historical linguistics, which fall out of range beyond a my opinion on the subject through an examination of few thousand years B.P. Despite claims of shallow time- concrete results at the more local level. depth by Bateman et al., "mass comparison" methods The first anomaly to come to my notice is the one lack the limitations of traditional historical linguistics linking "Uralic (Ling.)" closely with Mongol and only and may therefore be better equipped to work at greater after six nodes with the Caucasoid group. Uralic is a time-depths. Undoubtedly a better alternative, endorsed common linguistic term for a stock as ancient and well- by Bateman et al., would be the quantitative approach established as Indo-European (see Anttila 1972). Cavalli- suggested by Diamond (1988). Sforza et al. have taken a sample of "Samoyed and other Bateman et al. have shown that the correlation be- Uralic language speakers living near the Ural moun- tween human and linguistic evolution may not be as tains" and labeled it "Uralic (Ling.)." They apparently straightforward as the conclusions of Cavalli-Sforza et believe that their sample is representative of Uralic as a al. would suggest. Nevertheless, I cannot help feeling whole, as they later state, "Lapps associate linguistically that they come down too hard on the pioneer efforts of with speakers of Uralic but genetically with Cauca- Cavalli-Sforza et al. and that perhaps we should with- soids." The anomaly is not of the Lapps, who are simply hold judgement until more is known about their Caucasoid along with millions of other Caucasoid methods and data. We are all aware of the differences in speakers of the western Uralic languages, but of the two the evolutionary processes affecting living organisms branches of Uralic: the divergent, small (in population) and languages, and it is precisely this that raises doubts eastern (Samoyed) linguistic branch apparently fitting about their possible correspondence. On the other hand, better in the Northeast Asia genetic group and the over- for 50-100 millennia language-bearing human groups whelmingly more populous western (Finno-Ugric) have been chronically interacting with their neighbours. branch being Caucasoid. At the superphylum level the During periods of separation both their genes and their evidence is fairly strong that Uralic and Indo-European languages have evolved differentially in isolation; during are related (Pedersen 1962 [1931], Oswalt n.d.), support- contacts the exchange of genetic and linguistic elements ing the conclusion that it is the eastern branch of Uralic has taken place to varying degrees and in varying ways that has shifted its genetic affiliation. along with other aspects of culture. In other words, de- Cavalli-Sforza et al. say that by their bootstrap test spite the fact that they are subjected to somewhat differ- Berbers leave the Caucasoid cluster 20% of the time and ent processes and rates of transmission, adoption, and tend to join the African cluster. A shift 20% of the time preservation, both genetic and linguistic elements tend sounds substantial to me but would perhaps be accept- to be passed on and received during the same interaction able if Berber were simply insecurely placed between periods. It is perhaps traces of these very processes that two very recently formed branches. Berber, however, is we see in Cavalli-Sforza et al.'s results. shifting between the two branches of the first bifurca- A more reliable picture may be obtained by following tion of all mankind. With such results, one can wonder Bateman et al.'s recommendations, but I personally about the appropriateness of a tree diagram for display- would like to see archaeology playing a more important ing the data. Would not some technique for presenting role. Culture, including language, is often transmitted clinal relationships be more appropriate? Cavalli-Sforza together with genes, and therefore archaeological data et al. argue against one such model of an almost continu- may serve as a useful control. ous population network (that of Wolpoff, Wu, and Thome 1984), but I still do not understand their resis- tance to utilizing something like multidimensional scal- ROBERT L. OSWALT ing. If the calculations are too formidable, that is just California Indian Language Center, 99 Purdue Ave., another reason to work out the methods on the local or Kensington, Calif. 94708, U.S.A. 29 vm 89 regional level before going global. In the intriguing region through which the Americas These comments are on both the critique (Bateman et were peopled, Cavalli-Sforza et al. find the Eskimo clos- al.) and the critiqued (Cavalli-Sforza et al. 1988). I found est to the Chukchi and closer to the peoples of Northeast the critique the tougher reading; reduction of the Asia (including the Japanese and Korean) than to North- 22 | CURRENT ANTHROPOLOGY Volume 31, Number 1, February 1990

west Amerind or any Amerind. This is potentially a BAYARD, DONN, AND w. G. SOLHEIM ii. n.d. Archaeological ex- truly important piece of evidence unfortunately dimin- cavations at Non Nok Tha, northeastern Thailand, 1966-1968. ished by their statement that the cluster is "not very Vol. 2. Honolulu: University Press of Hawaii. In press, [DB] BENDER, M. L. 1969. Chance CVC correspondences in unrelated tight" and by doubts raised by other results. The precise languages. Language 45:519-31. composition of the sample is critical here: Is the "North- BERGSLAND, K., AND H. VOGT. 1962. On the validity of glotto- west Amerind (Na-Dene Ling.)" population made up chronology. CURRENT ANTHROPOLOGY 3:115-58. mainly or entirely of Indians far removed from the Es- BROOKS, D. R., R. T. O'GRADY, AND E. O. WILEY. 1986. A mea- kimo (Navajo or Haida, perhaps), or is it a nicely spaced sure of the information content of phylogenetic trees and its use as an optimality criterion. Systematic Zoology 35:571-81. sample? To me, the presentation would be more re- BUTLER, CHRISTOPHER. 1985. Statistics in linguistics. Oxford vealing if instead of such large-scale pooling samples had and New York: Basil Blackwell. [BGB] been taken of smaller divisions of the Eskimo-Aleut and CAMPBELL, A. 1983. "Transposons and their evolutionary of the divergent languages grouped under Na-Dene. significance," in Evolution of genes and proteins. Edited by M. Nei and R. K. Koehn, pp. 258-79. Sunderland, Mass.: Sinauer Although it might not be apparent from the recent Associates. dates on the references to linguistic classifications, dis- CAMPBELL, D. T. 1979- A tribal model of the social system vehi- cussion of the relationships between genetic and linguis- cle carrying scientific knowledge. Knowledge: Creation, Diffu- tic groupings has a long history, even as do many of the sion, Utilization 1:181-201. CAMPBELL, L. 1988. Review of: Language in the Americas, by present proposals for far-reaching linguistic affiliations, T. H. Greenberg (Stanford: Stanford University Press, 1987). Lan- especially those in the Eurasian landmass. William guage 64:591-615. Dwight Whitney, in a series of lectures in 1864 and CANN, R. L., M. STONEKING, AND A. C WILSON. I987. MltO- 1865 (Whitney 1872: esp. Lectures 8-11), elegantly and chondrial DNA and human evolution. Nature 325:31-36. clearly discussed the topics of close and distant linguis- CAVALLi-SEORZA, L. L. 1984. "Isolation by distance," in Human population genetics: The Pittsburgh Symposium. Edited by A. tic classification, the relationship between race and lan- Chakravarti, pp. 229-48. New York: Van Nostrand. [LLC, AP, guage, and the relative merits of "lumpers" and "split- PM, JM] ters" (not his terms). We are accumulating more and . 1988. The Basque population and ancient migrations in better data and improving statistical techniques, yet a Europe. Munibe (Antropologia-Arqueologia), suppl. 6, pp. 129- 37. [LLC, AP, PM, JM] great deal of what was said over 100 years ago could be CAVALLI-SFORZA, L. L., AND W. F. BODMER. I97I. The genetics usefully consulted now. of human populations. San Francisco: Freeman, [LLC, AP, PM, JM] [Bateman et al. will respond in the next issue.—Editor.] CAVALLI-SFORZA, L. L., AND A. W. F. EDWARDS. I964. "Analy- sis of human evolution," in Genetics today (Proceedings of the nth International Congress of Genetics), vol. 2, pp. 923-33. New York: Pergamon Press, [LLC, AP, PM, JM] . 1967. Phylogenetic analysis: Models and estimation procedures. American Journal of Human Genetics 19: .223-57/ Evolution 21:550-70. [LLC, AP, PM, JM] References Cited CAVALLI-SFORZA, L. L., AND M. W. FELDMAN. I98I. Cultural transmission and evolution. Princeton: Princeton University ANTTILA, RAIMO. 1972. An introduction to historical and com- Press.) parative linguistics. New York: Macmillan. [RLO] . 1989. Spatial subdivision of population and estimates of ARCHIE, j. w. 1989. Homoplasy excess ratios: New indices for genetic variation. Theoretical Population Biology. In press. measuring levels of homoplasy in phylogenetic studies and a [LLC, AP, PM, JM] critique of the consistency index. Systematic Zoology. In press. CAVALLI-SFORZA, L. L., A. PIAZZA, P. MENOZZI, AND J. [LLC, AP, PM, JM] MOUNTAIN. 1988. Reconstruction of human evolution: Bring- ASTOLFI, P., K. K. KIDD, AND L. L. CAVALLI-SFORZA. I98I. ing together genetic, archaeological, and linguistic data. Pro- A comparison of methods for reconstructing evolutionary ceedings of die National Academy of Science of the U.S.A. trees. Systematic Zoology 30:156-69. [LLC, AP, PM, JM] 85:6002-6. BARBUJANI, G., N. L. ODEN, AND R. R. SOKAL. 1989. Detecting . 1989. Genetic and linguistic evolution. Science regions of abrupt change in maps of biological variables. Sys- 244:1128-29. tematic Zoology. In press, [LLC, AP, PM, JM] DAMUTH, J. 1985. Selection among "species": A formulation in BARNICOT, N. A. 1964. "Taxonomy and variation in modern terms of natural functional units. Evolution 39:1132-46. man," in The concept of race. Edited by A. Montagu, pp. 180- DAVIDSON, 1., AND w. NOBLE. 1989. The archaeology of percep- 227. New York: Macmillan. [DFA] tion: Traces of depiction and language, CURRENT ANTHROPOL- BARRANTES, R., P. E. SMOUSE, H. W. MOHRENWEISER, H. OGY 30:125-55. GERSHOWITZ, J. AZOFIFA, T. D. ARIAS, AND J. V. NEEL. DAWKINS, R. 1976. The selfish gene. New York: Oxford Univer- 1989. Microevolution in Central America: Genetic character- sity Press. ization of the Chibcha-speaking groups of Costa Rica and . 1982a. "Replicators and vehicles," in Current problems in Panama, and a consensus taxonomy based on genetic and sociobiology. Edited by King's College Sociobiology Group, pp. linguistic affinity. American Journal of Human Genetics. In 45-64. Cambridge: Cambridge University Press. press, [LLC, AP, PM, JM] . 1982b. The extended phenotype. San Francisco: Free- BATEMAN, R. M., AND 1. DENHOLM. 1989a. Morphometricpro- cedure, taxonomic objectivity, and marsh-orchid systematics. DE QUEIROZ, K., AND M. j. DONOGHUE. 1988. Phylogenetic sys- Watsonia 17:449-55. tematics and the species problem. Cladistics 4:317-38. . 1989b. The complementary roles of organisms, popula- DIAMOND, J. M. 1988. Genes and the tower of Babel. Nature tions, and species in "demographic" systematics (abstract). 336:622-23. American Journal of Botany, suppl., 76:226. DIETZ, E. j. 1983. Permutation tests for association between two BAYARD, DONN. 1987. Review of: The prehistory of the Indo- distance matrices. Systematic Zoology 32:21-26. [TM] Malaysian archipelago, by Peter Bellwood (Sydney: Academic DIXON, R. M. w. 1980. The languages of Australia. Cambridge: Press, 1985). Journal of the Polynesian Society 96:114-18. [DB] Cambridge University Press. BATEMAN ET AL. Reconciling Human Phylogeny and Linguistic History | 23

DOERFER, G. 1985. Mongolo-Twigusica. Wiesbaden: Otto Har- HASEGAWA, M. 1988. Molecular phylogenetics and the origin of rassowitz. man (in Japanese with English summary). Biophysics 28:119-24. DOW, M. M., AND j. M. CHEVERUD. 1985. Comparison of dis- HENNIG, w. 1966. Phylogenetic systematics. Urbana: University tance matrices in studies of population structure and genetic of Illinois Press. microdifferentiation: Quadratic assignment. American Journal HOENIGSWALD, H. M. i960. Language change and linguistic re- of Physical Anthropology 68:367-73. [YM] construction. Chicago: University of Chicago Press. DUNN, c, AND B. s. EVERITT. 1982. An introduction to mathe- . 1973. Studies in formal historical linguistics. Dordrecht: matical taxonomy. Cambridge: Cambridge University Press. Reidel. EDWARDS, A. w. F., AND L. L. CAVALLi-SFORZA. 1964. "Recon- 1987. "Language family trees: Topological and metrical," struction of evolutionary trees," in Phenetic andphylogenetic in Biological metaphor anfl cladistic classification. Edited by classification. Edited by V. E. Heywood and J. McNeill, pp. 67- H. M. Hoenigswald and L. F. Weiner, pp. 257-67. Philadelphia: 76. London: Systematics Association, [LLC, AP, PM, JM] University of Pennsylvania Press. EFRON, B. 1982. The jacknife, bootstrap, and other resampling HULL, D. L. 1988. Science as a process. Chicago: University of plans. Philadelphia: Society for Industrial and Applied Mathe- Chicago Press. matics, [LLC, AP, PM, JM, YM] JASANOFF, j. H. 1988. Review of: Archaeology and language, by ELDREDGE, N. 1989. Macroevolutionary dynamics: Species, Colin Renfrew (New York: Cambridge University Press, 1987]. niches, and adaptive peaks. New York: McGraw-Hill. Language 64:800-802. FARRIS, J. S. 1981. "Distance data in phylogenetic analysis, "in KAISER, M., AND v. SHEVOROSHKIN. 1988. Nostratic. Annual Advances in cladistics, vol. I. Edited by V. A. Funk and D. R. Review of Anthropology 39:309-29. Brooks, pp. 3-23. New York: New York Botanic Garden. KIMURA, M. 1983. The neutral theory of evolution. Cambridge: . 1983. "The logical basis of phylogenetic analysis," in Ad- Cambridge University Press. vances in cladistics, vol. 2. Edited by N. I. Platnick and V. A. KLUGE, A. G., AND j. s. FARRIS. 1969. Quantitativephyletics and Funk, pp. 7-36. New York: Columbia University Press. the evolution of anurans. Systematic Zoology 18:1-32. FELSENSTEIN, j. 1985. Confidence limits on phylogenies: An ap- KRAUSS, M. E. 1988. "Many tongues—ancient tales," in Cross- proach using the bootstrap. Evolution 39:783-91. [LLC, AP, roads of continents. Edited by W. W. Fitzhugh and A. Crowell, PM, JM] pp. 145-50. Washington, DC: Smithsonian Institution Press. GILMOUR, j. s. L., AND j. w. GREGOR. 1939. Demes: A suggested LAITMAN, J. T. 1985. "Evolution of the upper respiratory tract," new terminology. Nature 144:333. in Hominid evolution: Past, present, and future. Edited by P. V. GILMOUR, j. s. L., AND J. HESLOP-HARRISON. 1954. Thedeme Tobias, pp. 281-86. New York: Alan R. Liss. terminology and the units of microevolutionary change. Genet- LAMBERG-KARLOVSKY, c. c. 1988. Indo-Europeans: A Near- ica 27:147-61. Eastern perspective. Quarterly Review of Archaeology 9:1, 9- GIMBUTAS, M.N 1988. Indo-Europeans: Guests or hosts? Quar- 10. terly Review of Archaeology 9:1-3. LEVINTON, j. 1988. Genetics, paleontology, and macroevolution. GINGERICH, p. 0. 1986. Temporal scaling of molecular evolution Cambridge: Cambridge University Press. in primates and other mammals. Molecular Biology and Evolu- LEWIN, R. 1988a. Trees from genes and tongues. Science 242:514. tion 3:205-21. . 1988b. American Indian language dispute. Science GODDARD, 1. 1975. "Algonquian, Wiyot, and Yurok: Proving a 242:1632-33. distant genetic relationship," in Linguistics and anthropology: LEWONTIN, R. c. 1972. "The apportionment of human diversity," In honor of C. F. Voegelin. Edited by M. D. Kinkade, K. L. Hale, in Evolutionary biology, vol. 6. Edited by T. H. Dobzhansky, and O. Werner, pp. 249-92. Lisse: Peter de Ridder Press. M. K. Hecht, and W. C. Steere, pp. 381-98. New York: . 1986. "Sapir's comparative method," in New perspectives Appleton-Century-Crofts. [LLC, AP, PM, JM] in language, culture, and personality. Edited by W. Cowan, MC GHEE, R. 1984. "Thule prehistory of Canada," in Handbook of M. K. Foster, and K. Koerner. Amsterdam/Philadelphia: John North American Indians, vol. 5, Arctic. Edited by D. Damas, Benjamins, [JHG] pp. 369-76. Washington, DC: Smithsonian Institution. 1987. Review of: Language in the Americas, by J. H. MANTEL, N. 1967. The detection of disease clustering and a gen- Greenberg (Stanford: Stanford University Press, 1987). CURRENT eralized regression approach. Cancer Research 27:209-20. [YM] ANTHROPOLOGY 28:656-57; 29:435. MARGULIS, L. 1982. Early hfe. Boston: Science Books Interna- 1988. "Pre-Cheyenne *y," in In honor of Mary Haas: From tional. the Haas Festival conference on Native American linguistics. MATESSI, R., P. GLUCKMAN, AND L. L. CAVALLI-SFORZA. I979. Edited by W. Shipley, pp. 345-60. Berlin: Mouton de Gruyter. Climate and the evolution of skull matrices in man. American GOULD, s. j. 1981. The mismeasure of man. New York: Norton. Journal of Physical Anthropology. 50:549-64. [LLC, AP, PM, JM] . 1989. Grimm's greatest tale. Natural History, no. 2, pp. MELLARS, p. 1988. The origin and dispersal of modern humans. 20-28. CURRENT ANTHROPOLOGY 29:186-88. COWER, j. c. 1984. "Multivariate analysis, ordination, multiple MILLER, w. R. 1986. "Numic languages," in Handbook of North scaling, and allied topics," in Handbook of apphcable mathe- American Indians, vol. n, Great Basin. Edited by W. L. matics, vol. 6, Statistics. Edited by E. H. Lloyd, pp. 727-81. D'Azevedo, pp. 98-106. Washington, DC: Smithsonian In- New York: Wiley. stitution. GREENBERG, j. H. 1963. The languages of Africa. International MONTAGU, A. 1964. The concept of race. New York: Macmillan Journal of American Linguistics 29:1-177. [DFA] . 1987a. Language in the Americas. Stanford: Stanford Uni- MORTON, N. E. 1982. Outline of genetic epidemiology. Basel: versity Press. Karger. [LLC, AP, PM, JM] 1987b. Author's precis [CA-fr book review of Language in MOURANT, A. E., A. C. KOPEC, AND K. DOMANIEWSKA- the Americas], CURRENT ANTHROPOLOGY 28:647-67. [JHG] SOBCZAK. 1976. The distribution of the human blood groups 1988. Language and archaeology. Quarterly Review of Ar- and other polymorphisms. London: Oxford University Press. chaeology 9:1, 8-9. [LLC, AP, PM, JM] GREENBERG, J. H., C. G. TURNER 11, AND S. L. ZEGURA. 1986. NEI, M. 1987. Molecular evolutionary genetics. New York: Co- The settlement of the Americas: A comparison of the linguis- lumbia University Press, [LLC, AP, PM, JM] tic, dental, and genetic evidence, CURRENT ANTHROPOLOGY NEI, M., AND A. K. RO YCHOUDHURY. 1974. Genie variation 27:477-97. [LLC, AP, PM, JM] within and between the three major races of man, Caucasoids, GUDSCHINSKY, s. 1964. "The ABC's of lexicostatistics (glotto- Negroids, and Mongoloids. American Journal of Human Genet- chronology]," in Language in culture and society. Edited by ics 26:421-43. [LLC, AP, PM, JM] D. Hymes, pp. 612-23. New York: Harper and Row. NELSON, G. A., AND N. I. PLATNICK. 1981. Systematics and HARRISON, G. A., et al. 1988. 3d edition. Human biology. London biogeography: Cladistics and vicariance. New York: Columbia and New York: Oxford University Press. University Press. 24 I CURRENT ANTHROPOLOGY Volume 31, Number i, February 1990

O'GRADY, R. T., I. GODDARD, R. M. BATEMAN, W. A. human immunoglobulin allotypes. Oxford: Oxford University DIMICHELE, V. A. PUNK, W. J. KRESS, R. MOOI, AND P. F. Press, [LLC, AP, PM, JM] CANNELL. 1989. Genes and tongues. Science 243:1651. STONEKING, M., K. BHATIA, AND A. c. WILSON. 1986. "Mito- OSWALT, ROBERT L. n.d. "A method for assessing distant linguis- chondrial DNA variation in Eastern Highlanders of Papua New tic relationships," in Sprang from some common source. Edited Guinea," in Genetic variation and its maintenance in tropical by Sydney Lamb et al. Stanford: Stanford University Press. In populations. Edited by C. F. Roberts and G. Destafano, pp. 87- press, [RLO] 100. Cambridge: Cambridge University Press. PAABO, s. 1989. Ancient DNA: Extraction, characterization, SWADESH, M. 1971. The origin and diversification of language. molecular cloning, and enzymatic amplification. Proceedings of Chicago: Aldine Atherton. the National Academy of Science U.S.A. 86:1939-43. TAKAHATA, N., AND M. NEI. 1985. Gene genealogy and variance PAABO, s., R. G. HIGUCHI, AND A. c. WILSON. 1989. An- of interpopulational nucleotide differences. Genetics 110:325- cient DNA and the polymerase chain reaction: The emerging 44- field of molecular biology. Journal of Biological Chemistry TEMPLE, j. T. 1982. Ordination of paleontological data. Miscel- 264:1-4. laneous Papers of the Geological Society of London 14:224-36. PEDERSEN, HOLGER. 1962(1931). The discovery of language: THOMASON, s. G., AND T. KAUFMAN. 1988. Language contact, Linguistic science in the nineteenth century. Translated by creolization, and genetic linguistics. Berkeley: University of John W. Spargo. Bloomington and London: Indiana University California Press. Press, [RLO] TILLS, 0., A. c. KOPEC, AND R. E. TILLS. 1983. The distribution PIAZZA, A., N. CAPPELLO, E. OLIVETTI, AND S. RENDINE. I989. of the human blood groups and other polymorphisms. Oxford: The Basques in Europe: A genetic analysis. Munibe (An- Oxford University Press, [LLC, AP, PM, JM] tropologia-Arqueologica), suppl. 6, pp. 168-76. [LLC, AP, PM, JM] UNGER, J. M. n.d. "Japanese and what other Altaic languages?" in PIAZZA, A., R. GRIFFO, N. CAPPELLO, M. GRASSINI, E. Papers from the workshop on linguistic change and reconstruc- OLIVETTI, s. RENDINE, AND G. ZEI. 1988. "Genetic structure tion methodology. Edited by P. A. Baldi. Berlin: Mouton de and lexical differentiation in Sardinia," in Language change and Gruyter. biological evolution. MS. [LLC, AP, PM, JM] WANG, w. s.-Y. 1987. "Representing language relationships," PIAZZA, A., P. MENOZZI, AND L. L. CAVALLI-SFORZA. I98I. in Biological metaphor and cladistic classification. Edited Synthetic gene frequency maps of man and selective effects of by H. M. Hoenigswald and L. F. Wiener, pp. 243-56. Philadel- climate. Proceedings of the National Academy of Science phia: University of Pennsylvania Press. U.S.A. 78:2638-42. [LLC, AP, PM, JM] WARDHAUGH, RONALD. 1986. An introduction to sociolinguis- PLATNICK, N. i., AND H. D. CAMERON. 1977. Cladistic methods tics. Oxford and New York: Basil Blackwell. [BGB] in textual, linguistic, and phylogenetic analysis. Systematic WEIR, B. s., AND c. c. COCKERHAM. 1984. Estimatingf-statis- Zoology 26:380-85. tics for the analysis of population structure. Evolution 38:1358- PRENTICE, H. c. 1986. "Continuous variation and classification," 70. [LLC, AP, PM, JM] in Infraspecific classification of wild and cultivated plants. WHEELER, Q. D., AND K. c. NIXON. n.d. Another way of looking Edited by B. T. Styles, pp. 21-32. Oxford: Systematics Associa- at the species problem: A reply to de Queiroz and Donoghue. tion and Oxford University Press. Cladistics. In press. RANKIN, R. L. 1988. "Quapaw: Genetic and areal affiliations," in WHITMAN, J. n.d. "A rule of medial *-r- loss in Pre-Old Japa- In honour of Mary Haas: From the Haas Festival conference on nese," in Papers from the workshop on linguistic change and Native American linguistics. Edited by W. Shipley, pp. 629-50. reconstruction methodology. Edited by P. A. Baldi. Berlin: Berlin: Mouton de Gruyter. Mouton de Gruyter. REID, LAWRENCE. 1988. Benedict's Austro-Tai hypothesis—an WHITNEY, WILIAM D WIGHT. 1872. Language and the study of evaluation. Asian Perspectives 26(11:19-34. [DB1 language: Twelve lectures on the principles of linguistic sci- RENFREW, c. 1987. Archaeology and language: Thepuzzleof ence. New York: Scribner. [RLO] Indo-European origins. New York: Cambridge University Press. WIENER, L. F. 1987. "Of phonetics and genetics: A comparison of . 1988. Author's precis [CA-fr book review of Archaeology classification in linguistic and organic systems," in Biological and language], CURRENT ANTHROPOLOGY 29:437-41. metaphor and cladistic classification. Edited by H. M. Hoenigs- RENSCH, B. 1959. Evolution above the species level. New York: wald and L. F. Wiener, pp. 217-26. Philadelphia: University of Columbia University Press, [DFA] Pennsylvania Press. RICHARDS, 1. A. 1936. The philosophy of rhetoric. New York: Ox- WIJSMAN, E. M. 1984. Techniques for estimating genetic ad- ford University Press. mixture and applications to the problem of the origin of the Ice- RUHLEN, M. 1987. A guide to the world's languages. Vol. 1. Stan- landers and the Ashkenazi Jews. Human Genetics 67:441-48. ford: Stanford University Press. [LLC, AP, PM, JM] RUVOLO, M. 1987. "Reconstructing genetic and linguistic trees: WILEY, E. o. 1981. Phylogenetics. New York: Wiley. Phenetic and cladistic approaches," in Biological metaphor and WOLPOFF, M. H., W. WU, AND A. G. THORNE. I984. "Modem cladistic classification. Edited by H. M. Hoenigswald and L. F. Homo sapiens origins: A general theory of hominid evolution Wiener, pp. 193-216. Philadelphia: University of Pennsylvania involving the fossil evidence from East Asia," in The origins of Press. modern humans: A world survey of the fossil evidence. Edited SHEVOROSHKIN, v. 1988. Methods in inter-phyletic comparison. by F. H. Smith and F. Spencer, pp. 411-83. New York: Liss. Wiesbaden. [DFA, RLO] SNEATH, p. H. A., AND R. R. SOKAL. 1973. Numerical taxonomy. WOODBURY, A. c. 1984. "Eskimo and Aleut languages," in San Francisco: Freeman. Handbook of North American Indians, vol. 5, Arctic. Edited by SOKAL, R. P. 1988. Genetic, geographic, and linguistic distances in D. Damas, pp. 49-63. Washington, D.C.: Smithsonian Institu- Europe. Proceedings of the National Academy of Science tion. U.S.A. 85:1722-26. [LLC, AP, PM, JM] WOODS, ANTHONY, PAUL FLETCHER, AND ARTHUR HUGHES. SPUHLER, j. N. 1979. "Genetic distances, trees, and maps of 1986. Statistics in language studies. Cambridge: Cambridge North American Indians," in The first Americans: Origins, University Press, [BGB] affinities, and adaptations. Edited by W. S. Laughlin and A. B. WURM, s. A. Papuan languages of Oceania. Tubingen: Narr. Harper, pp. 135-83. New York: Gustav Fischer, [LLC, AP, PM, JM] YASUMOTO, B. 1978. Nihongo no seiritsu (The formation of Japa- . 1988. Evolution of mitochondrial DNA in monkeys, apes, nese]. Tokyo: Kodansha. [YM] and humans. Yearbook of Physical Anthropology 31:15-48. YU, M. S., X. F. QIU, J. L. XUE, Z. D. LIU, J. Z. TAN, H. J. LI, 0. X. STAGE, c. A. 1975. Hybridization and the flora of the British LIU, L. LI, W. Z. YU, X. Z. TANG, AND M. R. ZHOU. I988. Isles. London: Academic Press. Mitochondrial chondrial DNA polymorphism in Chinese. Sci- STEINBERG, A. G., AND c. E. COOK. 1981. The distribution of the entia Sinitica B 31:860-72.