06 Trautmann (JB/D) 1/5/01 3:26 pm Page 268

Anthropological Theory

Copyright © SAGE Publications

London, Thousand Oaks, CA
and New Delhi
Vol 1(2): 268–287
[1463-4996(200106)1:2;268–287;017340]

The whole history of

kinship terminology in
three chapters
Before Morgan, Morgan, and after Morgan
Thomas R. Trautmann
University of Michigan

Abstract
The article questions the current consensus that kinship terminologies evolve from
something like the Dravidian to something like the English terminology, examining it
over three time periods. Before Morgan the study of kinship terminology was
embedded within a comparative study of core vocabularies to determine historic
relations among nations (e.g. Leibniz). Morgan’s breakthrough was to disembed the
terms of kinship from the vocabulary list and conceptualize them as a set. His vision
of their evolution had two phases. Before the revolutionary expansion of ethnological
time in the mid-19th century, he developed an evolutionary view of the Indo-
European kinship terminology that was very acute but tied to a short chronology for
world history that the time revolution shortly exploded; after the time revolution he
conceived the Iroquois and the English (as types of the Classificatory and the
Descriptive) terminologies as an evolutionary series caused by successive reformations
of the marriage rule. After Morgan, Dravidian and its structural neighbors have come
to play the role of evolutionary starting-point. The article concludes with reasons to
be skeptical of the current consensus and ways to move forward.

Key Words
Dravidian • evolution • history of anthropology • kinship • kinship terminology •
Leibniz • Morgan • time • time revolution

The invention of kinship was virtually the invention of anthropology itself; and kinship
terminology was at the heart of that inventing. Kinship and kinship terminology as
anthropological objects of study once were privileged sites of theorizing and have never
entirely disappeared, although they had declined greatly following the skeptical essays
of Needham (1971) and Schneider (1972, 1984), which called their coherence into
question. The new kinship studies of the 1980s and 1990s have made up a lot of the

268

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 269

TRAUTMANN The whole history of kinship terminology

lost ground. But it is my belief that anthropology’s brainchild will only thrive when it
grows beyond anthropology and makes effective alliances with other disciplines. The
study of kinship, and of kinship terminologies, has uncovered an order of facts that is
of the first significance for the understanding of human social life. In uncovering these
facts, anthropology, and anthropology alone, has made a contribution of permanent
value. But anthropology’s breakthrough understanding of kinship, as that which includes
the family but also something else that lies beyond and among families, has a value for
other disciplines that is still to be realized.
My own interest in kinship and kinship terminology was from the perspective of the
deep history of India. The Dravidian kinship terminology of South India and Sri Lanka
is a classic type whose geographical distribution correlates approximately with that of
the Dravidian language family – setting aside, for the moment, the many instances of this
type of kinship terminology in Oceania and the Americas. What makes the Dravidian
terminology so pleasing is the clarity with which it associates a rule of marriage – that of
cross-cousin marriage – with the semantic organization of the terms themselves.
The basic organizing principles of Dravidian systems are two. The first is shared with
Iroquois systems, and was noted by Morgan when he said of the Iroquois that ‘the father’s
brother is equally a father’, and the mother’s sister a mother; that is, the father-word is
applied to the father’s brother, and the mother-word to the mother’s sister. This is true
of Dravidian languages, except that the father’s brother is called ‘big’ or ‘little’ father, the
mother’s sister ‘big’ or ‘little’ mother, according to their age relative to the father or
mother. A consequence is that the child of such fathers and mothers are ego’s brothers
and sisters and are unmarriageable; moreover the children of these same-sex sibling
pairs are the sons and daughters of both, and are siblings to one another. Following
Lounsbury, we call this the principle of same-sex sibling merger (Lounsbury, 1964b). If
we imagined English transformed into Dravidian by the application of this rule, we
could say that the thickened categories of father, mother, sister, brother, son and daughter
that result constitute the set of parallel kin that overrides and replaces the distinction in
English of lineal and collateral kin.
The second principle of Dravidian systems is not shared with Iroquois: it is the prin-
ciple of cross-cousin marriage. By it, though a brother-sister pair may not marry, their
children should marry. The consequences of this are many. In combination with the rule
of same-sex sibling merger, we may imagine English transformed into Dravidian with the
formation of a category of cross kin, consisting of the uncles and aunts (minus those who
have become fathers and mothers through same-sex sibling merger), cousins (minus those
who have become siblings through the same principle), and nephews and nieces (minus

Figure 1. The whole history of kinship terminology diagram.

269

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 270

ANTHROPOLOGICAL THEORY 1(2)

those who have become sons and daughters through the same principle). But the cross-
kin category is thickened by addition of the affines: father- and mother-in law become
uncle and aunt; spouses and spouse’s siblings become cousins; children’s spouses become
nephews and nieces.
The pleasing simplicity of this terminological system – its evident logical integrity and
the difference of that logic from that of English kinship terminology – makes it an ideal
starting point for discussion of kinship terminology, and there has been an abundance
of ethnographic reports on terminologies of this kind from around the world. But it is
of special strategic value for studying the deep history of Indian civilization, for the many
ethnographic instances across South India and Sri Lanka can be shown to be so many
variations on a single theme; and the anthropological record of the present can be joined
up with evidence from ancient lawbooks, chronicles and inscriptions to show that the
pattern of Dravidian kinship is traceable for the better part of two thousand years and is
remarkably durable and resistant to change. Joining the anthropological study of kinship
terminology with a rich historical record leads us to think that the structures of kinship
terminology may be very slow to change and resistant to effects of changed political,
economic or social circumstance, or to the calculated interests of individual actors.
Kinship terminologies, these findings suggest, are not, after all, sensitive indicators of
changes in other aspects of social organization or modes of production or forms of politi-
cal association. For that very reason they are especially useful as traces of distant origins,
like languages themselves. Though South Indians are in every way participants in a
general Indian cultural pattern, including many aspects of family organization and lan-
guage, their language and kinship terminology nevertheless remain discoverably differ-
ent from those of North India and of those belonging to the Munda language family.
The lesson of history is that kinship terminology is very conservative and resistant to the
effects of other levels.
It was L.H. Morgan who invented the study of kinship terminologies, or what he
called ‘nomenclatures of relationship’, and in doing so served to create the field of
kinship by framing it as an object of study containing various different, but coherent
and logical, systems of relationships, the differences among which constituted a
problem worthy of close study. Kinship terminology was the site of the anthropologi-
cal breakthrough, destabilizing the idea of the family as an effect of nature and having
a fixed character, as Maurice Godelier (1995) has said. If kinship and, a fortiori, kinship
terminology have suffered from thinking small, it is worth revisiting Morgan’s concep-
tion to re-examine it in a larger, three-century context of before and after. That is what
I wish to do in this paper. The central issue will be to comment on an unpublished text
of Morgan from the first draft of the Systems of Consanguinity and Affinity (Morgan,
1870). This will be the middle part of the three natural divisions into which the history
of kinship terminology (and indeed anthropology) falls: Before Morgan, Morgan, and
After Morgan. Fast forwarding through history at this terrific rate will have the effect
of caricature, the virtue of which will be to draw into relief the gross features of the
topic by blurring the detail.
My purpose in doing so will be to subject a current consensus to critical scrutiny.
There is a remarkable uniformity of tendency among theorists in the 20th century to
assume that the beginning point for the evolution of kinship terminologies was a system
something like the Dravidian, and that the overall directionality of change in terminologies

270

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 271

TRAUTMANN The whole history of kinship terminology

is toward something like the kinship terminology in English. At the beginning of the
new century it is worth examining that assumption skeptically, to probe its strengths
and weaknesses, and ascertain whether it will be serviceable for the future. My strategy
for gaining purchase on the problem will be to throw the present consensus in relief by
going back to the manuscript of L.H. Morgan just mentioned, written in about 1865
but never published in full, on the evolution of kinship terminology, which expounds
an alternative to the current consensus. I will then consider the grounds on which the
current consensus stands, and reasons for skepticism.
In narrativizing the history of anthropology, one has, at the outset, to make a choice
about whether to stress continuity or discontinuity. To a degree the choice of a narra-
tive strategy is arbitrary. But I think that there are a number of good reasons why it is
useful to think of anthropology as we know it originating all at once, as the result of a
Big Bang, in around 1860. This Big Bang was what I have called the Revolution in
Ethnological Time; the revolution, that is, by which the short, Biblical chronology for
human history suddenly gave way to a very much lengthened chronology. In England
this collapse of the short chronology was especially associated with the excavation of
human artifacts in association with extinct fossil species at Brixham Cave by William
Pengelly and Hugh Falconer, and announced dramatically by Charles Lyell at the British
Association for the Advancement of Science in 1859, the year as well of the publication
of Darwin’s Origin of Species (Gruber, 1965; Trautmann, 1992).
The crisis that was provoked by the sudden immense lengthening of human history
had several effects. The most dramatic was that the Bible and the Greek and Latin clas-
sics, the oldest written records known to Europeans, in short, no longer gave a picture
of humanity’s ‘primitive’ or original state; and the vast new territory of ‘prehistory’ (the
very word was invented in this period), which came suddenly into being, required filling
by prehistoric archaeology. The most important result, perhaps, was the constituting of
‘primitive man’ as a concept. Hitherto, ‘savages’ were thought of as feral humans who
had lost the arts of civilized life with which God had fitted them in Eden (Hodgen,
1936); now the contemporary savage was seen not through the lens of a theory of degen-
eration but as a prolongation of the primitive state into the present. ‘Primitive man’ was
born, the first child of the time revolution (Trautmann, 1992).
The time revolution divided the scholarly activity of the generation of Morgan,
Darwin and Marx in two. In the case of Morgan, I have recovered the greater part of a
first draft of the Systems from among the Morgan Papers at the University of Rochester,
and have shown that the first draft is under the reign of the short chronology, while the
published version has been revised to accommodate the interpretation to the suddenly
lengthened time frame for human history. Thus Morgan’s study of kinship terminology
begins before the Big Bang, and is revised in the light of the Big Bang. The middle term
of our three-chapter history of the study of kinship terminology, therefore, will be sub-
divided into Morgan A and Morgan B, before and after the Big Bang.

BEFORE MORGAN
Briefly, as to the ‘before’ part of the picture. Morgan did not create kinship terminology
as an anthropological object out of the blue, but from within a long-standing tradition
of linguistic ethnology, the project of which was to determine the historical relations
among nations by determining the relations among their languages. This is a very

271

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 272

ANTHROPOLOGICAL THEORY 1(2)

old European project; but for our purposes it will suffice to pick up the story at the
beginning of the 18th century, during which we see the standardized vocabulary list
emerging as the simple-seeming method by which history was to be derived from com-
parative study of languages. Nowhere was this more important than in America, where
ancient written sources of history were lacking, since it was believed that language com-
parison could uncover historical relationships among Indian nations that had no ancient
written sources from which to recover them. Thomas Jefferson, no less, was the formu-
lator of such a standardized questionnaire, which, in the hands of his protégé and suc-
cessor as president of the American Philosophical Society, Stephen du Ponceau, laid the
foundations of Americanist historical linguistics (Jefferson, c. 1782; Du Ponceau, 1838;
see Trautmann, 1987: 80–7). At about the same time, Sir William Jones in Calcutta was
adumbrating the idea of the Indo-European language family on the basis of such lists,
and Catherine of Russia was inaugurating a plan for the study of the world’s languages,
beginning with those of her own vast empire, through an elaborate vocabulary list sent
to her officials in every region, and to distant correspondents who included George
Washington (Jones, 1788; Pallas, 1786–89).
The discovery of new and unexpected families of languages thoroughly revolution-
ized the deep history of the world. The standard vocabulary lists that were to uncover
historical relations among nations were taken around the world on the tide of European
imperial expansion, and they created an utterly new ethnological map of the world and
a new universal history. Many new, unexpected and counter-intuitive groupings were
discovered, such as the Indo-European language family, which united two widely sepa-
rated blocs of peoples, in Europe on the one hand, and in Iran and India on the other,
or the far-flung Malayo-Polynesian group ranging from Madagascar to Hawaii.
The standardized vocabulary that was the principal tool of this revolution in ethno-
logical knowledge looks to be a simple commonsense listing; but in fact it rests on specific
assumptions and is a highly theorized structure based on ideas that underlie much 18th-
century discussion of language, often implicitly, sometimes breaking into expression.
Within the short, Biblical time frame for human history – and the even shorter time span
of about 4000 years since the Confusion of Tongues at Babel – some original or stock
languages have survived more or less unchanged. Mixed languages have formed through
borrowing; but borrowing occurs among the words of science or art that a people need
when they have passed the primitive stage. At the heart of a language, then, is a core
vocabulary of the words that every language must have from the start; words expressing
those primitive notions and naming those things which are at the minimum threshold
of human life and speech. The vocabulary list focuses on that primitive core.
Leibniz published a simple list in 1718 that appears to have been a very influential
model, perhaps the model for subsequent lists (Leibniz, 1718). Many, perhaps most ver-
sions of the standardized vocabulary list that are so heavily in use in the 18th and 19th
centuries go back to this list. It is a very interesting list in which kinship terms figure
prominently, along with numbers, parts of the body, words for foodstuffs and utilitarian
objects, heavenly bodies, weather, animals and actions.
The list and the project it encodes set up a number of oppositions: primitive v. recent;
simple v. refined; native v. foreign, borrowed. A great deal of interpretation is built into
this tool, whose purpose is to find the indigenous core of a language in order to find its
nearest of kin.

272

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 273

TRAUTMANN The whole history of kinship terminology

This project, which was well under way by the 18th century, continues in the 21st
century; and an impressive number of the 18th-century findings remain current among
those who are supposed to know about these things, the historical linguists. The
mapping of the world by comparison of word-lists has been remarkably successful. Even
though one can no longer believe that the core vocabulary captures the state of a lan-
guage at the beginning of human history, or that history began only about 6000 years
back, it appears that there is enough truth in the notion that a core vocabulary is deeply
conservative to account for this success. Thus, to take an example from Indo-European,
the following Sanskrit words for kinship terms are surprisingly close to their English
counterparts: pitr. (father), mātr. (mother), śvasā (sister), bhrātr. (brother), duhitr. (daugh-
ter); though the word putra (son) is quite different, the more archaic word sunu is a
recognizable cognate of English son.
The other aspect of this European project of understanding the relations among
nations through the relations among languages is not so obvious from the vocabulary
list itself. That is the underlying ethnological framework within which are interpreted
the similarities among languages revealed by the vocabulary list. This framework is the
Tree of Nations, or as I call it, because its model is found in the Book of Genesis attrib-
uted to Moses, the Mosaic Ethnology (Trautmann, 1998: ch. 2). The Mosaic Ethnol-
ogy is a branching tree connecting the different nations through descent from their
founding patriarchs, the descendants of Noah and his three sons after the universal flood,
after whom, in Genesis, they are named – the Hebrews from the patriarch Eber, the
Greeks from Javan (i.e. Ionians), and so forth. This Tree of Nations has a shape that
Evans-Pritchard taught us to call a segmentary lineage; what the Mosaic Ethnology imag-
ines is a world made of nations who are related, not as Self and Other, but closer or more
distant (unilineal) relations of kinship forming a single extended segmentary lineage
uniting the human family.
This was the ethnological framework within which new ethnological knowledge was
located and interpreted, not only by Europeans, but by all the Peoples of the Book, so
to say, Jewish, Christian and Muslim, for several thousand years. And, in a sense, it con-
tinues today in the ‘cladistic’ maps of relationships figuring in current works of historical
linguistics, though its Biblical model is forgotten, or lingers only marginally in names
like Semitic and Hamitic (from Sem or Shem and Ham, sons of Noah).

MORGAN
Through this widely-dispersed project of linguistic ethnology, kinship terms were col-
lected from many parts of the world for a long time before Morgan, but they had no
special existence as a set until Morgan abstracted them from the larger vocabulary list in
which they were embedded and treated them as a set. The move by which kinship terms
were constituted as a set was rather complex. Adopting the realist, pre-Saussurian dis-
course of words and things that prevailed at the time, the vocabulary list had been created
to elicit the words that name a standardized list of things identified by their Latin,
English or French names. Comparison of languages, then, was a matter of comparing
names, holding things constant through the standardization of the questionnaire.
Morgan did not simply isolate the kinship terms from the larger set of the vocabu-
lary list. He wanted to argue, rather, that the terminology of kinship constituted not a
vocabulary or lexicon but a set of a different kind: a semantic set, a set of meanings or

273

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 274

ANTHROPOLOGICAL THEORY 1(2)

classes of things. That is, he recognized that the underlying things of kinship are them-
selves culturally organized into differing logics, and are not a fixed set of real things exist-
ing everywhere in the world and to be held constant. Comparison in Morgan became a
matter of examining the patterning of names in order to find distinctive configurations
of culturally constituted things.
The breakthrough move came in two steps, boiling down a complex story to its main
elements. The first was the recognition that for the Iroquois ‘the father’s brother is
equally a father’, or, as we might say, the word for father was extended to the father’s
brother (Morgan, 1851: 85–7). This ‘classificatory’ logic of same-sex sibling merger
operated throughout the Iroquois terminology, creating large classes of relatives that
classed or merged relatives held distinct in English. The simple vocabulary list would
not evoke facts of that order, and a special ‘schedule’ or questionnaire was required.
The second step was the discovery that Ojibwa had kinship terms with very similar
‘classificatory’ mergers of kin, similar enough to lead Morgan to conclude that he had
shown through this ‘new instrument for ethnology’ what linguistic ethnology could not,
that the Iroquois and the Ojibwa were historically related – even though their languages
were so different that they could not be shown to be linguistically related. Kinship would
be a more powerful tool for history than philology was, Morgan believed, because it
directed its analysis to an archaic and deeply conservative core of language (Feeley-
Harnik, 1999; Trautmann, 1987: 92–8).
Morgan’s breakthrough led to his masterwork, the Systems of Consanguinity and Affin-
ity of the Human Family (1870), in which all the kinship terminologies of the world
known to Morgan are put in one of two families, the classificatory, including Iroquois,
and the descriptive, including English. The project of the work was to deliver a proof,
through comparison of kinship terminologies, of the unity and Asiatic origin of the
American Indians. His proof consisted in showing that all kinship terminologies of the
world known to him, and collected into three massive tables in the Systems, could be
assigned to one of two families, the classificatory and the descriptive. Classificatory
systems, in which relations of lineal and collateral were classified or merged (such as the
father’s brother with the father), are logically integrated but different from descriptive
systems, in which the lineal and collateral lines are held distinct, and not merged. They
map onto two large contiguous world regions: descriptive for Europe and the Near East,
classificatory in Asia, Oceania and the Americas. Q.E.D. Much the most valuable aspect
of the book, however, is not the proof itself but the creation of kinship terminology as
an anthropological object, and the fact that Morgan devised much, perhaps most, of the
tools and basic categories used in subsequent kinship work.
In the draft of the Systems, prior to the time at which the time revolution reached
Morgan, he interpreted the descriptive/classificatory binary as the difference between
kinship systems that are natural v. those that are artificial. For Morgan the descriptive
system ‘followed the nature of descents’ by observing the boundary between lineal and
collateral kin; it followed the tutelage of nature. The classificatory system, by contrast,
was artificial. It was a stupendous work of art, Morgan thought, in that it departed from
the tutelage of nature by forming vast classes of kin that overrode the lineal/collateral
boundary; and the design, once formed, was perpetuated by its own logical integration
(Trautmann, 1987: ch. 6).
The time revolution, which reached Morgan as he was completing his masterwork,

274

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 275

TRAUTMANN The whole history of kinship terminology

allowed him to reconceptualize the relation between descriptive and classificatory

systems not as two parallel, contemporary formations but as successive stages in a long
series of evolutionary stages from primitive promiscuity to modern Euroamerican
monogamy, as we shall shortly see. Let us first examine Morgan A, before the Big Bang,
before going on to examine the reconceptualization in Morgan B, after the Big Bang, in
which kinship comes to play a role in the consolidation of the idea of modernity.
What follows is a summary of one of the most fascinating of the unpublished texts
in the Morgan Papers, a suppressed chapter of the first version of the Systems, on the
‘growth of nomenclatures of relationship’, that is, a kind of natural history of kinship
terminology. This is Morgan A.

MORGAN A
The text in question is Morgan’s manuscript on the growth of kinship terminologies or,
as he calls them, ‘nomenclatures of relationship’, in the Morgan Papers (Morgan, c.1865).
This text was to have been Chapter Six, Part I of his great summa on kinship, the Systems
of Consanguinity and Affinity of the Human Family (Morgan, 1870). Much of the text of
the first version Morgan wrote and submitted to the Smithsonian has been preserved in
the Morgan Papers at the University of Rochester, and comparison of the first version
with the published version shows a number of significant changes. Changes were made
to accommodate the not unreasonable demand of the Smithsonian to shorten the text
(the published version is over 600 pages) and to take account of a much lengthened con-
ception of human history, having to do with the revolution in ethnological time: the
breaking open of the short, Biblical time frame for human history, according to which
(in the Ussherite system, one of many variants of Bible-based world chronology) the
Creation occurred about 6000 years ago (at 4004 BC) and the ethnological variety of
the world even more recently, about 4000 years ago, following the unpleasantness at the
Tower of Babel (details in Trautmann, 1992). It is important in what follows to keep it
in mind that this first-version text of Morgan is a pre-time-revolution text, whose
horizon for human history is still supplied by the short Biblical chronology, and to note
how the time revolution upset the assumptions of the argument he made in it.
Let us examine, then, Morgan’s pre-time-revolution idea about the evolution of
terminologies. The draft chapter referred to focuses on the pattern of variation in the
kinship terminologies of the descriptive system, especially the Indo-European language
family and, given its assumptions, the argument is very acute. It asks how a system
originally purely descriptive could naturally grow and the probable causes of its develop-
ment. The causes Morgan finds to be of a logical and cognitive nature.
The analysis turns on the fact that his comparative table revealed that the words for
lineal kin are more or less stable in the various languages of the Indo-European family.
There are exceptions, of course; we have already seen that the ordinary Sanskrit word
for son (putra) is not cognate with English son, although a more archaic cognate (sunu)
is known, and we could easily add that Latin words for son and daughter (filius, filia)
and Greek for brother (adelphos) do not fit the pattern. But overall there is a consider-
able incidence of evident cognates among Indo-European words for the lineal kin,
F,M,Z,B,S,D. To this list Morgan adds the words for husband and wife, considering
these the primary kin, and taking them to belong to the primitive core of the language.
For kinship terms beyond this central core, Morgan finds little of the consistency in

275

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 276

ANTHROPOLOGICAL THEORY 1(2)

the patterning of words from one Indo-European language to another. His leading
example is the words for nephews, nieces and grandchildren. Cognates of Greek nepos
are found in all the Indo-European languages of his table, but their meaning varies
between nephew and grandchild, even in English. Thus in his will, Shakespeare
bequeathed a hundred pounds to his niece Elizabeth Hall, who was, however, his daugh-
ter’s daughter, showing that the word meant granddaughter as well as sibling’s daughter
in earlier forms of English. This is confirmed in a table of the degrees of kinship and
consanguinity in the important compendium of English law, Coke upon Littleton, in
which one finds the lineal nephew and niece (i.e. grandchildren) together with the
collateral nephew and niece (Coke, 1817; Trautmann, 1987: 136). This difference
between the relatively stable lineal terms and the relatively unstable collateral and more
distant terms Morgan attempts to explain by a conjectural history of how kinship terms
developed from the very beginning, and in doing so he implies a belief that the Indo-
European origin is near the origin of human history, thought to be but a few thousand
years ago.
Morgan’s idea of how kinship terminologies develop is that a core vocabulary con-
sisting of primary kin terms is formed first. After their invention the more distant
relationships could for a time be readily described by the juxtaposition of terms, such as
‘mother’s brother’. Progress beyond this stage would be the result of generalization and
classification or merger.
To indicate collateral relationships there were three alternatives: to invent new terms;
to continue to generate descriptive terms through the juxtaposition of existing terms;
‘or, by a false generalization . . . to classifiy the collateral with the primary relatives, as
uncles with the father, aunts with the mother, & cousins with the brother’ as in the clas-
sificatory system. Morgan considers the middle method, that of pure description (i.e.
the generation of phrasal terms), to be too inconvenient, so there would be a tendency
for the invention of special terms for the near collateral kin; but a terminology ample
enough to express all collateral relationships would be even more inconvenient, so we
should expect new terms to be limited to the nearest degree of collateral kin – the terms,
that is, for uncle, aunt, cousin, nephew and niece. He continues, ‘each new term would
be apt to be used in a general sense, until it became restricted in its application, by the
introduction of additional terms’, so that kinship terminology would develop from the
center (ego) outward, from the primary terms to the secondary, and from lineal to
collateral kin, each additional term to be devised lessening the semantic scope of the
existing terms.
After the core terms had been invented, the next stage would probably be the inven-
tion of a general term for nephew, grandson and perhaps cousin. In Indo-European
the cognates of nepos, from their universality, must have existed ‘in the primitive
speech’ but in a wide sense. Since (as will be shown) it is older than terms for grand-
father and grandmother, uncle and aunt, it must have existed without a reciprocal in
the form of a special term. The semantic scope and variation of the evidence of this
word in the different Indo-European languages leads to the conclusion that this word
was the first to have been introduced into the Indo-European kinship terminology ‘to
indicate relationships beyond those expressed by the primary terms, and that at first it
was used in a more comprehensive sense than at present’ (Morgan, c.1865: 78). The
nepos-word, then, had a wide range of meaning that was narrowed down differently in

276

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 277

TRAUTMANN The whole history of kinship terminology

the different Indo-European languages, as more words were invented to designate the
outer band of non-primary kin.
‘With the progress of discrimination by which the several degrees of relationship were
made distinct, and by the invention or introduction of new terms,’ nepos and its cog-
nates were restricted in usage and new terms occupied part of its original semantic space.
In Sanskrit, Latin, Spanish and Portuguese it was restricted to grandson. In English,
Flemish, Norman, French and German it became restricted to the son of a brother or
sister, and, having no original term for grandson, these languages fell back upon descrip-
tive phrases for it, except for German, which developed a new term (Enkel). In Anglo-
Saxon, Italian and Dutch the nepos-word was used generally for nephew and grandson,
and in the last for cousin as well.
Terms for grandfather and grandmother, and for uncle and aunt, would next be
required. Here the variation is even wider, and patterns form and conflict. It is at least
certain for Latin that avus, avia (grandfather, grandmother) were formed earlier than the
term for maternal uncle, avunculus, which is made by adding a diminutive: ‘little grand-
father’ (and the source for English uncle). The two relations were perhaps formerly one,
and reciprocal (Morgan calls them correlative) to nepos. Further, the relationships of
uncle and aunt ‘in the concrete’ do not appear to have been discriminated in the primi-
tive Indo-European language since they are so various from one Indo-European language
to another. Many of the Indo-European terms for these relationships are derivatives of
words for father and mother such as the set pitr. vya (Skt.), patruus (Lat.), patros (Gk.) for
father’s brother, and mātulā (Skt.) matertera (Lat.) and mētrōs (Gk.) for mother’s brother’s
wife, and the lack of terms that have cognates through all the branches of the family (or
in other words the heterogeneity of the lexicon) indicates strongly that descriptive terms
were used for these relationships at the period of the separation of the languages from one
another. For the remaining uncles and aunts, the different branches of the Indo-European
family exhibit a great variety of terms, showing that the terms came into being at the period
of separation of the Indo-European languages, not before. Once the relationships of uncle
and aunt were fully discriminated by means of special terms, the separation of the relation-
ship of nephew and grandson by means of special terms or descriptive phrases would soon
follow, and these would each be furnished with a proper correlative (reciprocal).
The term for cousin would be the last invented, it being the most remote collateral
relation discriminated by means of a special term. The English word cousin is based on
a generalization of four distinct relationships, the children of a father’s brother and sister,
and those of a mother’s brother and sister. The absence of a common term among the
Indo-European languages, and the presence of several borrowed terms for this relation-
ship, shows that the original method for designating the relationship of cousin was
descriptive (e.g. father’s brother’s son, etc.). English cousin and its cognates are from
Latin consobrinus, -a, coming from con plus soror, sister, meaning the children of two or
more sisters, which in English is extended to cousins of all four kinds.
Morgan summarizes his view of the process by which the terminologies of the Indo-
European family became very diverse in vocabulary, while remaining quite similar in
semantic structure, in a passage that gives the rationale of this most important aspect of
his thought, for it is the point at which the analysis of kinship terminologies takes on a
life of its own separate from that of philology, and becomes a new science of its own. It
is worth quoting the passage in full:

277

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 278

ANTHROPOLOGICAL THEORY 1(2)

The great diversity of modes which now exist in the several Aryan [i.e. Indo-Euro-
pean] languages for the same objects, and many of them the most common, must
find its chief explanation in the theory of development from root forms after the
separation of those languages became complete. We can hardly suppose that [the
words for cousin] consobrinus and vetter or amita and muhme existed contempora-
neously in the Latin and German while the two languages were yet associated as
dialects of a common speech; and that the Latin perpetuated one, the German the
other, while several of the remaining original Aryan dialects lost both. A dialect
cannot arise without the isolation of the community in which it originates. When
the separation from the mother tongue has become so complete as to produce a well
defined dialect, from that moment an amount of divergence has arisen sufficient to
establish, in effect, an independent language with its full inheritance of root forms,
grammatical organism, and ethnic life. It lives from thenceforth, in the brains of the
people. Any outgrowth of words, or forms, become a distinct creation of the nascent
language; and if such new words were borrowed by other dialects they must pass
through the ordinary process of naturalization. The real cause of the diversity of
words which finally becomes so apparent, must be sought in the falling out of use of
old terms, which may, or may not, have had synonyms, and the substitution of newly
coined words in their places; and the original formation of a large class of words to
keep pace with the advance of experience and knowledge. (Morgan, c.1865: 54)

To summarize: around a largely unchanged primitive core vocabulary of primary kin

terms, the Indo-European languages developed terms for more distant collateral and
lineal relationships, earlier inventions such as the nepos word first denoting larger seman-
tic spaces which shrank as new terms were invented to make finer discriminations.
The interpretation is interesting and acute, but it is by no means the only interpre-
tation possible. In particular, Morgan might have made the argument that in the older
Indo-European languages, as in Iroquois and Dravidian, the ‘father’s brother is equally
a father’, that is, the fact that the word for father’s brother is evidently a derivative of the
father word could have been seen as a kind of classification or merger, as could other of
these phenomena: avus with avunculus; mātulā/matertera/mētrōs with the word for mother;
and the expansion of consobrinus to all the cousins. In short, he could have denatural-
ized Indo-European and seen that it is as artificial as the Iroquois or the Dravidian.

MORGAN B
The effect of the time revolution, which intruded itself on Morgan’s consciousness as he
was revising his manuscript for publication, was to incapacitate his theory of the origin
and growth of kinship terminology based on Indo-European; for no matter how acute
his analysis, the beginning of Indo-European could not go back in time more than a few
thousand years, and certainly not to the newly remote beginning of human history. Thus
the text in question never saw print – or rather, it survived only in severely reduced form,
as a pair of long footnotes (Morgan, 1870: 35 fn, 37 fn).
We have seen that Morgan A read the descriptive and classificatory systems as simul-
taneous and parallel but differing developments at the beginning of human history.
Morgan B offered a completely new interpretation, which served to put the two systems
in chronological series, making the classificatory system ancestral to the descriptive. The

278

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 279

TRAUTMANN The whole history of kinship terminology

means by which he did so was to devise a theoretical series of marriage rules and reforms
stretching from the zero of primitive promiscuity (recycled from the classics) to modern
European and Euroamerican monogamy, with matriarchal and patriarchal stages among
the stages in between. As I have shown elsewhere, the whole interpretive scheme sprang
from a suggestion of a friend (the Rev. Joshua Hall McIlvaine; see Trautmann, 1987:
158–72) and was offered tentatively. Morgan B has a decided feel of being added on to
a book that was originally conceived upon the assumption of the short, Biblical chronol-
ogy for human history.
Morgan’s origin myth for the natural development of kinship terminology is that it
grows outward like a crystal from a central core of primary terms, consisting of names
for naturally given things immediate to consciousness and needing no explanation. The
account is a good one for the patterning of similarity and variation among kinship terms
of the Indo-European languages, assuming that the Indo-European comparisons lead
one back to the originary condition of mankind at the start of its history, to ‘man’s primi-
tive state’, which of course they do not.
Two things separate origin myths devised by anthropologists of the 20th century from
this text: the central place the Dravidian system comes to occupy, and the assumption
of a much longer chronology – so much longer that the beginnings of the Indo-European
languages come to be thought to be much too recent to carry us back to the beginnings
of language and of humanity’s primitive state. The time revolution, as I have said, over-
took Morgan soon after he wrote this text, and it prompted him to adopt an entirely
new interpretation of the relation between the descriptive and the classificatory systems;
dropping the idea that the first was natural and the second artificial, he now interpreted
the second as a stage ancestral to the first. This became the ‘practice of time’ that struc-
tured the published version of the Systems and of his great synthesis of social evolution,
Ancient Society. By this move the contemporary savage became a primary source
(Morgan, 1877) for knowledge of man’s primitive state, which now was moved so far
back as to be out of reach of the oldest written sources in Latin and Greek and in the
Bible. The savage became the primitive, in the new sense of being contemporary, but
culturally of an earlier era. What Johannes Fabian calls ‘the denial of coevalness’ (Fabian,
1983) was the immediate fruit of the time revolution and the source of coherence for
the new science of anthropology that was born from this Big Bang.

AFTER MORGAN
The new practice of time and its new creation, ‘primitive man’, was a new interpretative
frame that had an empty slot into which the Dravidian system quickly dropped and has
since remained. That slot occupied a place at or near the zero of development, as a stage
some time in the distant past, and most distant structurally from the kinship system of
Europe. Within that frame, Dravidian appeared to be within or in the vicinity of
humanity’s primitive state. This has become so generally believed as to be more or less
unquestioned. I suggest that the time has come to question it.
Morgan himself prepared the way for this development by identifying Dravidian with
Iroquois and opposing them as examples of the classificatory system to the descriptive
system of Europe and the Near East. The identity of Dravidian and Iroquois was long
accepted, till Lounsbury published a classic article drawing attention to differences in
the ways Iroquois and Dravidian extend the cross/parallel distinction (Lounsbury,

279

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 280

ANTHROPOLOGICAL THEORY 1(2)

1964a). Analysis shows, I argue, that this difference stems from the fact that while
Iroquois and Dravidian have in common the principle of same-sex sibling merger,
Iroquois lacks the second principle of Dravidian classifications, cross-cousin marriage,
so that Iroquois has a separate terminology for affines (i.e. they are not merged with cross
kin) (Trautmann, 1981: 47–62; Viveiros de Castro, 1998: 352–64). The relation of
Dravidian and Iroquois to one another was an especially fruitful question for theoreti-
cal development at the 1993 Paris round table on Dravidian held in the Maison Suger
and published in the Transformations of Kinship volume, in papers by Trautmann and
Barnes, Tjon Sie Fat, Ives, Hornberg, Parkin, Kryukov, Allen, Vivieros de Castro and
Godelier (in Godelier et al., 1998).
Morgan failed to recognize the role of cross-cousin marriage in the Dravidian system,
and he was particularly obtuse about it, since his suppliers of information on the
Dravidian system made clear the rule and its connection with the terminology. It is a
fateful coincidence, as I have argued, that Morgan himself married, as it happens, his
mother’s brother’s daughter, a cross cousin in systems with such a rule (Trautmann,
1987: 243–5); which, I believe, accounts for his blind spot on this subject when dealing
with the Dravidian, wrongly identified with the Iroquois.
However that may be, the logic of the Dravidian system quickly became apparent to
other analysts, for whom it became a classic site for kinship theory-formation. Thus
Dravidian figures prominently in an important early article of E.B. Tylor on kinship ter-
minologies, and centrally in W.H.R. Rivers’ kinship book, which effectively renewed and
revived the work of Morgan (Tylor, 1889; Rivers, 1914). Lévi-Strauss’ Elementary Struc-
tures of Kinship (1949, 1969) is shot through with instances of Dravidian or similar type.
The same may be said of Needham and many recent theorists of the big picture of
kinship terminology, especially N.J. Allen and Michael Kryukov (Allen, 1998; Kryukov,
1998) in the Transformations of Kinship volume from the Maison Suger round table on
Dravidian (Godelier et al., 1998).
Lévi-Strauss’ great work on kinship, dedicated to Morgan, has rather little of kinship
terminologies in it, to be sure, and is directed rather toward rules of marriage. More-
over, Lévi-Strauss by no means encourages historical or evolutionary readings of his
structuralist analysis of regimes of marriage, and is skeptical of history in several ways,
identifying it as the surface of things, while structure is identified with the depths. Never-
theless the analysis lies well within the practices of time that established anthropology
in the era of Morgan. Thus the elementary systems of direct, indirect and delayed
exchange in Lévi-Strauss, based on a positive rule of marriage, all of them directly
exemplified by variants of the Dravidian system, are opposed to complex systems, having
only a negative marriage rule, exemplified by modern European systems, with Crow-
Omaha systems occupying the transition from elementary to complex. The argument is
both structural and historical or evolutionary. One can scarcely not read it so – it is
nearly impossible to take it as a pure structuralism without evolutionary significance.
And the whole is set in motion by a preliminary visit to ground zero, the beginning of
the incest tabu, which is taken to be the beginning of culture as such; a scientific origin
myth, in short – Morgan brought up to date. The second masterpiece of kinship studies
is in this important respect perfectly aligned with the terms of the first.
N.J. Allen’s tetradic theory (Allen, 1986, 1989, 1998) is an exceptionally elegant
example of more recent theoretical work that illustrates the overall directionality I have

280

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 281

TRAUTMANN The whole history of kinship terminology

been speaking of. Unlike Lévi-Strauss’ theory of marriage, Allen’s tetradic theory is
directly concerned with kinship terminology, specifically with discovering the simplest
imaginable terminology. This is found to be a terminology with only four terms, whence
the name ‘tetradic’. The four terms are divided along two axes, which we may translate
roughly as generation A v. generation B, and cross v. parallel (or eight, taking account
of gender, which is left out of the analysis). This tetradic system is the theoretical start-
ing-point of all further developments, and it is very strikingly similar to the Dravidian
system, having the characteristic Dravidian equations such as MB=FZH=SpF; the
equations (Morgan’s ‘classification’) of different kinds of consanguines and affines into
a few large classes gives the tetradic model its great compactness and elegance.
Allen derives all other types of kinship terminology from the tetradic starting-point
by breaking the equations of kin. For example, the Dravidian equation MB=FZH=SpF
obtains in the tetradic system, but is broken open in Iroquois, and in English for that
matter, by the existence of a separate set of affinal terms. The tetradic theory aims to
derive all possible kinship terminologies from the tetradic model, positing it as the point
of origin and showing how all other terminologies can be derived by the loss of the many
equations making up the tetradic mother of them all. The result can be read in a weak
form as a structuralist analysis or in a strong form as an evolutionary or developmental
sequence, taking the tetradic model as the starting-point.
Reading the sequence as an evolutionary one, how do we know that the directional-
ity of change in the model corresponds to what happens in history? Allen offers a most
interesting argument: it is easy to imagine the tetradic equations being undone, but it
is hard to imagine a society without them deliberately making the tetradic equations.
This argument transforms a structural typology of kinship terminologies into an evolu-
tionary sequence of stages, running, roughly, from systems of Dravidian and similar type
to something more like the English. Even without it, the tetradic theory would remain
interesting as a structural analysis, though here again, it is hard not to read it in the
stronger, evolutionary way, and it can be asked whether the structural analysis is not
already implicitly an historical one.
Inadequate though it is, this brief characterization of the study of kinship and of
kinship terminologies after Morgan in our three-chapter history of the study of kinship
terminologies must suffice, and I leave it to readers to supply the complications for them-
selves; for it is necessary, now, to consider what the future may hold, and why the way
we go from here may require a shift of direction.

WHAT NEXT?
As we have already suggested, the practices of time that undergird the consensus after
Morgan are under challenge, notably by Johannes Fabian’s book. The figure of ‘primi-
tive man’ that anthropology constructed, and that constructed anthropology, in the wake
of the time revolution made contemporary savages into ancestors of the ‘modern,
Western’, lenses into the past that saw more deeply than the oldest writings would allow,
deeply beyond the literature of the ancient Greeks, Romans, Hebrews. A sudden post-
modernist skepticism about the story of the rise from ‘man’s primitive state’ to ‘civiliz-
ation’, that is, about the story of modernity itself, which anthropology did so much to
create through kinship studies among other things, can itself take stronger and weaker
forms, and one need not entirely reject the notion that complexity of economies and

281

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 282

ANTHROPOLOGICAL THEORY 1(2)

polities is acquired only incrementally through history to be skeptical about the current
consensus concerning the evolutionary location of Dravidian and to suspect that its role
in the story of modernity is overdetermined from the start, that it dropped into a pre-
viously fashioned slot in the interpretative machinery.
Reason for skepticism no. 1: the non-correlation of kinship terminologies with econ-
omic or political forms or stages. Kinship terminologies, like languages, seem not to cor-
relate with stages of economic or political development. The non-correlation of language
with economic or political stages of development has been a central tenant of linguistic
anthropology from Sapir (1921); the non-correlation of kinship terminology with such
stages has taken longer to become established, but seems equally certain. After years of
trying to insert kinship and language into the story of modernity by main force, it seems
necessary to recognize that the evidence, after all, has enough strength to resist our best
efforts to press it into preformed molds.
This seems especially true of Dravidian. In its classic terrain, India and Sri Lanka, the
Dravidian system is found associated with economies and polities of the widest variety,
from forest peoples practicing shifting agriculture or hunting and gathering to the large
states and empires built upon economies of wet rice cultivation in the great river deltas;
and we can go to hunting bands of Ojibwa-speakers thinly distributed across vast terri-
tories in North America, or to many societies of Amazonia, to extend the range of econ-
omic and political associations of societies with kinship terminology of Dravidian type.
The first reason for skepticism, then, must be that the current consensus seems to restore
to Dravidian by other means the indexical status in the story of modernity that all
kinship terminologies have, at length, been released from.
Reason for skepticism no. 2: The mechanism Allen proposes to give directionality to
change in kinship terminologies is just not sustained empirically. The argument that
movement away from equations is imaginable while the reverse is not seems a good one,
but many counter-examples could be given. We can even find them in the text of Morgan
we have been considering. Morgan, as we have seen, shows how the nepos words of Indo-
European languages once were based on the equation SbS=ChS, nephew plus grandson,
which later broke down, the cognates becoming specialized as the word for one relation-
ship but not the other. This agrees with the posited directionality of change. But the
English word cousin, Morgan says, is probably from Latin con + sobrinus, which had
indicated only the children of sisters before being broadened to cover all cousins, imply-
ing a movement toward greater equations or, as Morgan puts it, classification. Other of
his Indo-European examples are the extensions of the father word to the father’s brother,
and the mother word to the mother’s brother’s wife, in Sanskrit, Latin and Greek, at
least as Morgan sees it.
There is a more general consideration why we should not buy into the idea that equa-
tions are a one-way street leading away from the tetradic model. Kinship terminologies
are strongly areal in character; that is, world regions seem to have a limited repertoire of
kinship terminological types within them, and a high degree of recurrence. This is cer-
tainly true of Dravidian South India, whereas there are other world regions from which
Dravidian is practically unknown, such as Europe, the Near East and Africa. That being
so, any group that happens to move into the Dravidian south of India is likely to take
on the characteristics of the dominant type over time. We can hardly doubt that this has
happened, for example, in the cases of Arab merchant groups settled on the Kerala coast

282

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 283

TRAUTMANN The whole history of kinship terminology

(Gough, 1961), or the many Brahmin communities of South India that are Dravidian
in kinship but are thought to be ultimately of North Indian origin (Trautmann, 1981:
302–15), or the Sinhalese of Sri Lanka, whose language derives from North India but
whose kinship system is Dravidian (Yalman, 1962, 1967). Counter-examples tending to
show that the direction of change can go toward equations must be abundant, but we
will be satisfied to have given a few. If there is a net tendency of kinship terminologies
to change in a single direction it must be a very weak one.
A thought experiment: suppose a group having a kinship terminology of Eskimo type
should become highly developed economically and politically, and should conquer and
settle far distant places, bringing the Eskimo type wherever it went. Would there not be
some likelihood that the Eskimo type would propagate itself wherever colonies were
established, and would have an influence on other kinship terminologies where its rule
was established? But of course this thought experiment has already been performed by
history, for the British have a system of Eskimo type; and it leads, in Morgan and after
Morgan, to a story of kinship in which the Eskimo type becomes the apex of civilization
and the objective of historical directionality.
A different thought experiment: suppose that the Chola empire of 12th-century
South India, having expanded southward into Sri Lanka, northward into Central India
and eastward toward Malaysia and Indonesia, had been able to continue its expansion
until it had constituted itself a large empire extending across continents. Would not the
Dravidian system have followed wherever its people settled, and have had an influence
on other systems wherever its rule established itself? Would not some Dravidian Morgan
have explained the association of Dravidian kinship with civilization by a story of kinship
the reverse of the modern consensus? Isn’t the current consensus influenced, conditioned
by, the course of world history? And don’t such theories reimport what has been rejected,
the correlation of kinship terminological types with stages in the scale of civilization?
My sense of what is most needed for the study of kinship terminologies to have a
future is a good dose of deep history. This is best achieved through broad, detailed
surveys of the world’s regions similar to the kind I provided for Dravidian in South Asia
some years back. While studies of particular societies will continue to bring forth new
types of kinship terminology not known before – for in kinship studies the age of dis-
covery is not over – what is more likely to be fruitful at this stage is examination of the
patterning of contiguous systems worldwide, and a consideration of the causes of
unevenness of spread. Such area surveys need to be explicitly historical in character,
which is not to say evolutionary, but concerned with seeing the in-region variation as a
field of movement within a time frame of the last couple of thousand years.
It is this time frame that is most weakly developed in anthropology today. At its birth,
anthropology was to have provided the new universal history for the vastly expanded
chronology revealed by the revolution in ethnological time. To a degree, anthropology
has been delivering on that promise over the years, but the coverage is uneven. Deep,
deep history is addressed by evolutionary anthropology; and cultural anthropology, com-
mitted to the phenomenology of contemporary life, increasingly sees its field in larger
historical contexts of colonialism and post-colonialism. But the deep history that lies
between, say, the end of the last ice age and the beginning of the Victorian era, is not
thickly populated by anthropologists, especially cultural anthropologists, with notable
exceptions such as Jack Goody, Marshall Sahlins, Eric Wolf, and Sidney Mintz. For all

283

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 284

ANTHROPOLOGICAL THEORY 1(2)

the contemporary commitment of cultural anthropology to history, the deeper past is

greatly neglected.
So what? In the case of kinship terminologies, I would argue that studies confined to
the ethnographic present and lacking access to the deep history of this most interesting
of anthropological objects can entirely miss its most interesting feature: the durability,
conservatism and language-like lack of correlation with political and economic regimes.
Clinging to the ethnographic surface of history and without access to this truth, anthro-
pologists have imagined, to the contrary, that kinship terminologies change rapidly with
changes of residence rules and other aspects of social structure, as Murdock (1949)
believed, or through the aggregated manipulations of interested individuals according to
practice theories inspired by Bourdieu (1977). Both views I believe are profoundly
flawed, and can only have arisen because of a more or less complete lack of access to
actual historical data on kinship terminologies. Studies of kinship terminologies in East
Asia (Kryukov, 1998) and in India (Trautmann, 1981) show what may be learned of
kinship terminologies by bringing the historical record into relation to the ethnogra-
phies of the present.
Thanks to Morgan and all the anthropological collection of terminologies that fol-
lowed in his wake, the data needed to accomplish this project are at hand, and need only
to be assembled and surveyed in a manner he would have appreciated, with an open-
ness to alternative explanations and a skeptical view of the consensus that has been built
up after Morgan. The examination of his suppressed text on the growth of kinship ter-
minologies is worth revisiting, not indeed that we can restore that earlier interpretation
as he formulated it – the time revolution has swept it away forever – but to encourage
us to reimagine a history of kinship terminologies in which there is not a single sequence
from A to Z, but many parallel developments coexisting from the start and surviving as
traces of a deep past.

Note
The title is inspired by that of a comprehensive history of sailing from the time of Noah’s
ark, attributed to John Locke: The Whole History of Navigation from its Original to This
Time (1704), in The Works of John Locke, new edn, 10 vols., London: printed for Thomas
Tegg, 1823, vol. 10, pp. 357–512. It was prefixed to Churchill’s Collection of Voyages.
Kinship terminology, of course, begins before Noah’s ark and the invention of naviga-
tion, but the formal study of it begins with Morgan in the 19th century.

References
Allen, N.J. (1986) ‘Tetradic Theory: An Approach to Kinship’, Journal of the
Anthropological Society of Oxford 17: 87–109.
Allen, N.J. (1989) ‘The Evolution of Kinship Terminologies’, Lingua 77: 173–85.
Allen, N.J. (1998) ‘The Prehistory of Dravidian-type Terminologies’, in Maurice
Godelier, Thomas R. Trautmann and Franklin Tjon Sie Fat (eds) Transformations of
Kinship, pp. 314–31. Washington, DC: Smithsonian Institution Press.
Bourdieu, Pierre (1977) Outline of a Theory of Practice. Cambridge: Cambridge
University Press.
Coke, Edward (1817) The First Part of the Institutes of the Laws of England. Or, A
Commentary upon Littleton, 17th edn, 2 vols. London: W. Clarke and Sons.

284

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 285

TRAUTMANN The whole history of kinship terminology

Du Ponceau, P.S. (1838) Mémoire Sur le Systéme Grammatical des Langues de Quelques
Nations Indiennes de l’Amérique du Nord. Paris.
Fabian, Johannes (1983) Time and the Other: How Anthropology Makes its Object. New
York: Columbia University Press.
Feeley-Harnik, Gillian (1999) ‘ “Communities of Blood”: The Natural History of
Kinship in Nineteenth-Century America’, Comparative Studies in Society and History
41: 215–62.
Godelier, Maurice (1995) ‘Is Social Anthropology Indissolubly Linked to the West, its
Birthplace?’, International Social Science Journal 47: 141–58.
Godelier, Maurice, Thomas R. Trautmann and Franklin Tjon Sie Fat, eds (1998)
Transformations of Kinship. (Papers of the Round Table on Dravidian Kinship,
Maison Suger, Paris, 3–5 June 1993.) Washington, DC: Smithsonian Institution
Press.
Gough, Kathleen (1961) ‘Mappilla: North Kerala’, in David M. Schneider and
Kathleen Gough (eds) Matrilineal Kinship, pp. 415–42. Berkeley, Los Angeles,
London: University of California Press.
Gruber, Jacob W. (1965) ‘Brixham Cave and the Antiquity of Man’, in Melford E.
Spiro (ed.) Context and Meaning in Cultural Anthropology, pp. 373–402. New York:
The Free Press; London: Collier-Macmillan.
Hodgen, Margaret (1936) The Doctrine of Survivals: A Chapter in the History of
Scientific Method in the Study of Man. London: Allenson & Co. Ltd.
Hornborg, Alf (1998) ‘Serial Redundancy in Amazonian Social Structure: Is There a
Method for Poststructuralist Comparison?’, in Maurice Godelier, Thomas R.
Trautmann and Franklin Tjon Sie Fat (eds) Transformations of Kinship, pp. 8–186.
Washington, DC: Smithsonian Institution Press.
Ives, John W. (1998) ‘Developmental Processes in the Pre-Contact History of
Athapaskan, Algonquian, and Numic Kin Systems’, in Maurice Godelier, Thomas
R. Trautmann, and Franklin Tjon Sie Fat (eds) Transformations of Kinship, pp.
94–139. Washington, DC: Smithsonian Institution Press.
Jefferson, Thomas (c.1782) Notes on the State of Virginia. N.p., n. pub. (First Edition of
the Notes on Virginia, found in the Bodleian Library, Oxford University).
Jones, Sir William (1788) ‘The Third Anniversary Discourse: On the Hindus’, Asiatic
Researches 1: 415–31.
Kryukov, M.V. (1998) ‘The Synchro-Diachronic Method and the Multidirectionality
of Kinship Transformations’, in Maurice Godelier, Thomas R. Trautmann and
Franklin Tjon Sie Fat (eds) Transformations of Kinship, pp. 294–313. Washington,
DC: Smithsonian Institution Press.
Leibniz, Gottfried Wilhelm von (1718) Desiderata Circa Linguas Populorum, ad Dn.
Podesta, Interpretem Caesareum Transmissa. Otium Hanoveranum, ed. Joachimis
Fridericus Fellerus. Leipsig: J.C. Marini. (Reprinted in G.W. Leibnitii opera omnia
ed. L. Dutens, Geneva, vol. 6, part 2 (Collectanea etymologica), pp. 228–31.)
Lévi-Strauss, Claude (1949) Les Structures Élémetaires de la Parenté. Paris: Presses
Universitaires de France.
Lévi-Strauss, Claude (1969) The Elementary Structures of Kinship (Les Structures
Élémetaires de la Parenté). Rev. ed. James Harle Bell and John Richard von Sturmer
(tr.) and Rodney Needham (ed.). London: Eyre & Spottiswoode.

285

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 286

ANTHROPOLOGICAL THEORY 1(2)

Lounsbury, Floyd G. (1964a) ‘The Structural Analysis of Kinship Semantics’, in

Horace G. Lunt (ed.) Proceedings of the Ninth International Congress of Linguists, pp.
1073–93. The Hague: Mouton.
Lounsbury, Floyd G. (1964b) ‘A Formal Account of the Crow and Omaha-type
Kinship Terminologies’, in W.H. Goodenough (ed.) Explorations in Cultural
Anthropology, pp. 351–93. New York: McGraw-Hill.
Morgan, Lewis Henry (1851) League of the Ho-d-no-sau-nee, or Iroquois. Rochester,
NY: Sage and Brother.
Morgan, Lewis Henry (c.1865) Nomenclatures of Relationship. Morgan Papers, Rush
Rees Library, University of Rochester (manuscript).
Morgan, Lewis Henry (1870) Systems of Consanguinity and Affinity of the Human
Family. Smithsonian Contributions to Knowledge, vol. 17. Washington, DC:
Smithsonian Institution.
Morgan, Lewis Henry (1877) Ancient Society, or Researches in the Lines of Human
Progress from Savagery through Barbarism to Civilization. Reprint with a foreword by
Elisabeth Tooker 1985. Tucson: University of Arizona Press.
Murdock, George Peter (1949) Social Structure. New York: The Free Press; London:
Collier-Macmillan.
Needham, Rodney (1971) ‘Remarks on the Analysis of Kinship and Marriage’, in R.
Needham (ed.) Rethinking Kinship and Marriage. (ASA Monographs, 11), pp.
xiii–cxvii. London: Tavistock.
Pallas, P.S. (1786–89) Linguarum Totius Orbis Vocabularia Comparativa, 2 vols. St.
Petersburg: Carl Schnoor.
Parkin, Robert (1998) ‘Dravidian and Iroquois in South Asia’, in Maurice Godelier,
Thomas R. Trautmann and Franklin Tjon Sie Fat (eds) Transformations of Kinship,
pp. 252–70. Washington, DC: Smithsonian Institution Press.
Rivers, W.H.R. (1914) Kinship and Social Organisation. London: Constable & Co.
Sapir, Edward (1921) Language: An Introduction to the Study of Speech. New York:
Harcourt, Brace and Company.
Schneider, David M. (1972) ‘What is Kinship all About?’, in Priscilla Reining (ed.)
Kinship Studies in the Morgan Centennial Year, pp. 32–63. Washington, DC: The
Anthropological Society of Washington.
Schneider, David M. (1984) A Critique of the Study of Kinship. Ann Arbor: University
of Michigan Press.
Tjon Sie Fat, Franklin E. (1998) ‘On the Formal Analysis of “Dravidian,” “Iroquois,”
and “Generational” Varieties as Nearly Associative Combinations’, in Maurice
Godelier, Thomas R. Trautmann and Franklin Tjon Sie Fat (eds) Transformations of
Kinship, pp. 59–93. Washington, DC: Smithsonian Institution Press.
Trautmann, Thomas R. (1981) Dravidian Kinship. (Cambridge Studies in Social
Anthropology, vol. 36.) Cambridge: Cambridge University Press.
Trautmann, Thomas R. (1987) Lewis Henry Morgan and the Invention of Kinship.
Berkeley, Los Angeles, London: University of California Press.
Trautmann, Thomas R. (1992) ‘The Revolution in Ethnological Time’ (the Marett
Memorial Lecture 1991), Man n.s. 27: 379–97.
Trautmann, Thomas R. (1998) Aryans and British India. Berkeley, Los Angeles,
London: University of California Press.

286

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015

06 Trautmann (JB/D) 1/5/01 3:26 pm Page 287

TRAUTMANN The whole history of kinship terminology

Trautmann, Thomas R. and R.H. Barnes (1998) ‘ “Dravidian’”, “Iroquois,” and

“Crow-Omaha” in North American Perspective’, in Maurice Godelier, Thomas R.
Trautmann, and Franklin Tjon Sie Fat (eds) Transformations of Kinship, pp. 27–58.
Washington, DC: Smithsonian Institution Press.
Tylor, E.B. (1889) ‘On a Method of Investigating the Development of Institutions:
Applied to Laws of Marriage and Descent’, Journal of the Anthropological Institute
18: 245–72.
Viveiros de Castro, Eduardo (1998) ‘Dravidian and Related Kinship Systems’, in
Maurice Godelier, Thomas R. Trautmann and Franklin Tjon Sie Fat (eds)
Transformations of Kinship, pp. 332–85. Washington, DC: Smithsonian Institution
Press.
Yalman, Nur (1962) ‘The Structure of the Sinhalese Kindred: A Re-examination of the
Dravidian Terminology’, American Anthropologist 64: 548–75.
Yalman, Nur (1967) Under the Bo Tree. Studies in Caste, Kinship and Marriage in the
Interior of Ceylon. Berkeley and Los Angeles: University of California Press.

THOMAS R. TRAUTMANN is author of Dravidian Kinship (Cambridge, 1981), Lewis Henry Morgan and
the Invention of Kinship (California, 1987) and is co-editor (with Maurice Godelier and Franklin Tjon Sie Fat)
of Transformations of Kinship (Smithsonian, 1998). He also writes on ancient India and Orientalism, notably
in his book Aryans and British India (California, 1997) and other current work, focusing on language and
race. Address: Department of History, University of Michigan, Ann Arbor, MI 48109, USA. [email:
ttraut@umich.edu]

287

Downloaded from ant.sagepub.com at GEORGIAN COURT UNIV on March 3, 2015