LINGUIST List 11.286

Fri Feb 11 2000

Review: Dixon: The rise and fall of languages

Editor for this issue: Andrew Carnie <carnielinguistlist.org>

Directory

1. Angus B. Grieve-Smith, Review of Dixon 1997

Dixon, R.M.W. 1997. The rise and fall of languages. Cambridge:
Cambridge University Press. 152 pages, 32 pounds 50 hardcover, 10
pounds 95 paperback.

Angus Grieve-Smith, University of New Mexico

	Drawing on his experience with Australian and South American
languages, R.M.W. Dixon proposes combining the areal and family-tree
macro-level models of language change into a single punctuated
equilibrium model, based on similar models in evolutionary biology and
geology. This is a vast improvement over the family-tree model alone,
as Dixon shows with several examples, but it still glosses over
important situations that may fall midway between the two.

	Dixon begins this short book with a discussion of linguistic
areas and a critique of the family tree model. He is particularly
critical of the excesses of family-tree approaches such as Nostratic
hypotheses and quantitative theories like glottochronology. He then
gives a summary of the ways that languages can change, and integrates
those changes into a punctuated equilibrium model. The model, as he
describes it, allows for a language area to be in one of two states.
States of equilibrium come about in areas where a number of languages
coexist in prolonged contact at roughly equal levels of prestige, with
none of the groups aiming to conquer any of the others, such as
probably existed in much of Africa, Australia and the Americas before
European colonization. States of punctuation are precipitated by
large-scale population expansions, such as the Polynesian migration or
the Roman conquest, and result in language change of the type
traditionally described with family trees.

	The next chapter goes on to discuss proto-languages in that
light. The processes that result in language families are
punctuations, he argues, so most if not all proto-languages must have
emerged from a state of equilibrium. In fact, following Meillet 1967,
he argues that we have no evidence that language families like
Indo-European originated in a single language rather than a number of
neighboring languages. He even conjectures that Indo-European and
Uralic languages might be similar because they arose from two
languages that were part of the same general area under a period of
equilibrium.

	Dixon continues with a discussion of the major sociolinguistic
events of the past five hundred years of European colonialism, and
ends by summarizing the implications of his argument for contemporary
issues. Since the beginning of European colonization around 1500,
ethnic groups around the world have been either exterminated, or
forced by torture or economic circumstances to abandon their languages
in favor of prestige European languages. Just over the past hundred
and fifty years, languages have been disappearing at an alarming rate.
Dixon argues that linguists need access to data from the widest
possible range of languages in order to form an accurate model of the
nature of language. He argues that a Basic Linguistic Theory of
"fundamental theoretical concepts" exists, essentially the consensus
of all the world's linguists. Before a linguist does any advanced
theoretical work, he firmly suggests that they should use their Basic
Theory to analyze one previously unstudied language, thus enriching
the pool of data that linguists have to draw on and giving them
practical insights into another language that probably has significant
typological differences from their native language.

	The book is short, concise and readable, almost a 150-page
essay. It is primarily aimed at linguists outside of the field of
contact linguistics, with the goal of encouraging them to look at
contact phenomena. Most of the ideas are laid out in the
Introduction, and supported in detail in subsequent chapters. Dixon's
tone borders on arrogant at times: he dismisses one Nostraticist claim
in a footnote where he calls it "palpable poppycock," and has a
section titled "What every linguist should do." While this is unkind
to the Nostraticists and mildly offensive to every linguist but Dixon,
these attacks ultimately come off as more amusing than objectionable.

	Dixon's book is itself one of the best pieces of evidence in
support of his admonitions to linguists to study endangered languages. The
punctuated equilibrium hypothesis emerges from the Australian data Dixon
has been working with for years, data he would not have had access to if
the European settlers had exterminated aboriginal Australians the way that
they did aboriginal Tasmanians. He would not have drawn the same
conclusions had he confined himself to the members of the language
families that are more popular with linguists. And it is very possible
that current and future language extinctions may deprive us of further
valuable generalizations. But it is not clear from my interaction with
other linguists that Dixon's "Basic Theory" really exists, or that it can
record everything we need to record. There may exist important aspects of
language that we may not know to record until it is too late.

	While the punctuated-equilibrium model is a valuable
contribution to historical linguistics, it draws heavily on the
established field of language contact and areal studies, where
linguists have long been presenting evidence that contact phenomena
play a larger role in language change than is often thought. Dixon's
presentation also glosses over some important aspects of language
contact. He presents the case of Indo-European as a classic
punctuation stage; after all, it was the basis for the family tree
model. But the Indo-European languages have been in contact
situations all along, with each other and with other families such as
Uralic and Basque. Arguments have been made for substrate or
adstratum influence in many of the Indo-European splits: a Dravidian
substrate for Indic, Uralic substrates for East Slavic, and a Gaulish
substrate and Frankish adstratum influence for French. It is more
useful to see the punctuation-equilibrium contrast as a continuum
rather than a dichotomy, with extreme equilibrium represented by the
Australian languages and extreme punctuation represented by a case
like Polynesian, where there were no other languages for the expanding
population to come into contact with.

	 Grammaticization theory (Hopper and Traugott 1993), among
others, argues that at a certain level of analysis it is not productive to
draw a line between synchronic and diachronic approaches to languages: in
order to understand the present structure of a language, we must know
something about its past. Theories of language contact, including
Dixon's, add another piece of the puzzle, demonstrating that a language's
past almost always involves several other languages. We can eventually
expect an integrated linguistics that models the intricate weavings of
forms from one language to another as an important part of the structures
of individual languages and the social uses of those forms.

References

Hopper, Paul J., and Elizabeth Closs Traugott. Grammaticalization. 
Cambridge: Cambridge University.

Meillet, Antoine. 1908 [1967]. The Indo-European dialects. Translated
by Samuel Rosenberg. Tuscaloosa: University of Alabama.

Angus B. Grieve-Smith is a PhD student at the University of New
Mexico. His major projects are a computer signed-language synthesis
application, a pilot English-to-ASL machine translation project and a
corpus-based study of topicalization in conversational French. Other
interests include language contact and variation, phonology and
functional syntax.