Genes and Tongues

Author(s): Richard T. O'Grady, Ives Goddard, Richard M. Bateman, William A. DiMichele,

V. A. Funk, W. John Kress, Rich Mooi and Peter F. Cannell
Source: Science, New Series, Vol. 243, No. 4899 (Mar. 31, 1989), p. 1651
Published by: American Association for the Advancement of Science
Stable URL: https://www.jstor.org/stable/1703461
Accessed: 23-02-2019 05:44 UTC

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1703461?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

American Association for the Advancement of Science is collaborating with JSTOR to

digitize, preserve and extend access to Science

This content downloaded from 192.248.32.231 on Sat, 23 Feb 2019 05:44:22 UTC
All use subject to https://about.jstor.org/terms
 nation of original data sources reveals many tions must be attributed to the independent
very small samples, sometimes of single indi- origin of a language in more than one race,
viduals. Differences in the data preclude or to the replacement of one language by
identification of the zones of positively cor-another.
related depressions in allele frequencies nec- 8) We conclude by asking why one would
Detection of Plastic Explosives essary to delimit races and assess intrarace expect correspondence between phylogenet-
variation. ic and linguistic relationships. Why should
M. Mitchell Waldrop, in his News & 3) The linguistic classification uses the 17 languages be expected to arise and persist in
Comment article "FAA fights back on plas- "families or phyla" recognized by Ruhlen a lineage like physical traits? The history of
tic explosives" (13 Jan., p. 165), describes (5). Poorly supported phyla such as Amer- language is known to have included exten-
the Federal Aviation Administration's new ind, Na-Dene, Altaic, and Austric are given sive shifting and extinction without con-
thermal neutron activation detector for plas- the same weight by Cavalli-Sforza et al. as comitant events in the races involved. Corre-
tic explosives in airline passengers' baggage. well-supported families such as Indo-Euro- spondence between linguistic phyla and ge-
He says initial testing demonstrated "more pean and Dravidian. netic clusters does not necessarily indicate
than 95% detection probability, with less 4) The correspondence between the ge- similar origins; it may arise from the inde-
than 5% false alarms." netic and linguistic data is not particularly pendent parallel effects of geographical con-
If one makes the very generous assump- "remarkable." Cavalli-Sforza et al. state that tiguity or separation. Any attempt to recon-
tions that there may be 100 bombs among "every linguistic phylum corresponds to struct global human history must deal with
the 10 billion bags that pass through U.S. only one of the six major genetic clusters evidence that linguistic relationships reflect
airports annually, then there will be about defined by the [race] tree" (1, p. 6005), but a much later period in human history than
500 million false alarms each year. One subsequently note six exceptions. Elsewhere, the genetic relationships among human
wonders whether the FAA has plans in place appearance of correspondence is increased populations. The wide reporting of the
to deal with the accompanying loss of vigi- by delimiting six races on the basis of linguis-
study by Cavalli-Sforza et al. does, however,
lance the human operators of these detectors tic affinity (1, p. 6003). Potential problems, show that it is possible to achieve a remark-
are bound to experience when they learn such as Munda, the form of Austroasiatic able impact just by asking an important
that the detectors signal the existence of 5 spoken in India, are not considered. Seven question.
million or more bombs for each bomb dis- of the linguistic phyla are each uniquely RIcHARD T. O'GRAi)Y
covered. associated with a single race on the genetic- National Museum of Natural History,
JAMES R. SPEER data tree and are thus in agreement with any Department of Invertebrate Zoology,
Department of Psychology,tree. Five of the remaining nine phyla sup- Smithsonian Institution,
Stephen F. Austin State University, port the tree presented: they occur only in Washington, DC 20560
Nacogdoches, TX 75962-3046 races grouped in a major genetic cluster. The IvEs GODDARD
remaining four phyla do not support the Linguistic Editor,
tree: they occur in races belonging to non- Handbook of North American Indians, and
contiguous clusters. Thus, of the phyla ame- Department of Anthropology,
Genes and Tongues nable to a test of agreement at the level of National Museum of Natural History
the major genetic clusters, the fit is only RIcHARD M. BATEMAN
We would like to comment on Roger 56% (5/9). WILLIAM A. DIMICHELE
Lewin's Research News article (28 Oct. 5) Only Greenberg's (6) controversial Department of Paleobiology,
1988, p. 514) about a paper by Luigi Luca Amerind phylum (7) corresponds with the National Museum of Natural History
Cavalli-Sforza et al. (1). The paper examines grouping of individual races within a major V. A. FUNK
human history with genetic, archeological, cluster. Thus, of the nine testable linguistic W. JOHN KREss
and linguistic data and has attracted other phyla, only one (1 1%) does not conflict with Department of Botany,
reviews in the scientific press (2, 3)-appar- the genetic-data tree in some way. National Museum of Natural History
ently with minimal scrutiny by systematists 6) Neither of the linguistic "superphyla," RICH MOoi
and linguists. Its main conclusion, that there Nostratic and Eurasiatic, precisely corre- Department of Invertebrate Zoology,
is "considerable parallelism between genetic sponds with the genetic-data tree. Cavalli- National Museum of Natural History
and linguistic evolution" (1, p. 6002), is not Sforza et al. state that Nostratic "includes six PETER F. CANNELL
adequately supported empirically and meth- phyla that all belong to the Northeurasian Department of Vertebrate Zoology,
odologically (4). major cluster" (1, p. 6005). However, their National Museum of Natural History
1) The tree of relationships among 42 figure 1 associates one Nostratic phylum
"aboriginal" human races, constructed from with a non-Northeurasian race and seven
REFERENCES AND NOTES
120 allele frequencies, is a phenogram for non-Nostratic phyla with seven Northeura-
which Cavalli-Sforza et al. concede impor- sian races. The proposed addition of Amer- 1. L. L. Cavalli-Sforza, A. Piazza, P. Menozzi, J.
Mountain, Proc. Natl. Acad. Sci. U.S.A. 85, 6002
tant statistical tests are not reported. The ind to Nostratic unites two groups erected (1988).
data-matrix contains an unacceptable num- by mutually exclusive techniques. 2. J. M. Diamond, Nature 336, 622 (1988).
ber of missing values. A phenogram is an 7) If, as the authors intended, the genetic- 3. S. J. Gould, Nat. Hist., 20 (1989).
4. I. Goddard et al., in preparation.
ahistorical clustering diagram of overall sim-data tree is treated as a phylogeny, we can 5. M. Ruhlen, A Guide to the World's Languages, vol. 1,
ilarity and cannot be used directly to infer map the linguistic data onto the tree and Classification (Stanford Univ. Press, Stanford, CA,
1987).
historical relationships. Despite assumptions measure the goodness of fit (8). This reveals
6. J. H. Greenberg, Language in the Americas (Stanford
to the contrary (1, 3), phenetic similarity is that only 48% of the race-language associa- Univ. Press, Stanford, CA, 1987).
not necessarily indicative of kinship. tions support a conclusion of development 7. I. Goddard, Curr. Anthropol. 28, 656, ibid. 29, 435
(1987); L. Campbell, Language 64, 591 (1988).
2) The races were delimited by arbitrary and retention of a language within a racial
8. With the Consistency Index, excluding nonhorno-
pooling within geographic regions. Exami- lineage. The remaining 52% of the associa- plasious autapomorphies.

31 MARCH i989 LETTERS 165i

This content downloaded from 192.248.32.231 on Sat, 23 Feb 2019 05:44:22 UTC
All use subject to https://about.jstor.org/terms