Cross-linguistic categories - what are they?

Tom Givon tgivon at uoregon.edu
Sun Mar 14 19:45:34 UTC 2010


Thanks, Jack. In doing the historical bit, not to mention the primate 
fun-&-games, I actually withheld my main points. To wit:

(a) CATEGORIES--Platonic/Chomskian (either/or absolute/logical) vs. 
biological/cognitive Prototypes. Ernst Mayr has a beautiful discussion 
of this in biology, completely compatible with Elinor Rosch. And Esa did 
a good job of raising this issue, tho he used "clustering" instead of 
"prototypes", and forgot to cite some people who talked about it before 
2001. [My "Syntax: A Functional-Typological Intro. (1984), "Context as 
Other Minds" (1989), "Bio-Linguistics (2002) and "Context as other 
minds" (2005) all have chapters on this issue; but the discussion 
started much earlier; see C. Craig (ed. 1986) "Categorization & Noun 
Classification"].

(b) UNIVERSALS: We should not let our rejection of Chomskian/Platonic 
universals blind us to REAL universals, those that are not either/or, 
but allow a range of flexibility, multi-factored complexity and 
context-mediated diversity. As in biology again, empirically-responsible 
linguistic universals do not involve statements such as "all languages 
have Feature X". This is a Chomskian parody of universals, a mere straw 
man. Rather, real universals define both universality AND diversity, 
thus the range of possible--considerable but never 
unconstrained--variation. This is totally in line with Ernst Mayr (1974) 
"closed vs. open program" idea. And as in biology, the diversity in some 
areas of grammar is much more constrained than in others.

(c) UNIVERRAL & LINGUISTIC FIELD WORK: Having worked on at least 30 
languages over the past 45 years, some of them in considerable depth 
(1-year plus), I've always found universals the to be the best 
guideposts for discovering the particulars. Of course, the universals I 
have in mind are not Platonic/Chomskian but rather Greenbergian/Mayrian. 
So maybe Matt Dryer (1997; or after his recent foray to New Guinea) 
would like to teach us about the great, distortive evils of universals 
in doing field-work. And maybe Martin Haspelmath (2009) would like to 
teach us about the great, biasing evils of having an--explanatory, 
widely-connected--unabashedly theoretical perspective. So far, I have 
found "universal" categories such as noun/verb or subject/object, and 
the complex theory behind them, to be indispensable in my own 
descriptive filed work. I am still waiting to hear serious argument why 
I've been in grave error. (I hear plenty of those in the Church, where 
dogma prevails).

(d) COMPARISON WITHOUT UNIVERSALS: This is a logical issue, of the type 
that Esa Itkonen is well-known for raising. It is sheer logical nonsense 
to compare A to B to...Z unless you have some standard terms for the 
comparison. Of course, you can play semantic games and call your 
universals terms "desceriptive", "heuristic", "contingent", "inductive", 
"pragmatic" or "ephemeral". For all I care, you may call them 
"abracadabra" or "apple pie". But they are still a logical pre-requisite 
for any meaningful comparison in ANY domain. Nothing special here about 
linguistics. Just read Aristotle's "Metaphysics".

Peace, TG

=========


John Du Bois wrote:
> Thanks, Talmy. Good words on the range of ideas needed for typology. 
> ("Semantics" alone is not going to do it.)
>
> And I share your interest in the parallels with biological 
> theory/evolution.
>
> Jack

============
>
> Tom Givon wrote:
>>
>>
>>
>> Dear FUNK people,
>>
>> I have been debating, and vacillating, whether to join in this 
>> discussion. First because Esa Itkonen had already raised the very 
>> same issues ca. a year ago, and we all had a spirited go at it. But 
>> mostly because this discussion reminds me of a book I have just 
>> finished reading, Dave Geary's "Male Female: The Evolution of Human 
>> Sex Differences". This may sound less-than-obvious to many of you, 
>> but the history of linguistics for the past 100 years is eerily 
>> reminiscent of the evolution of male coalitions in Homo sapiens, with 
>> inter-coalition warfare and intra-coalition power-struggles--and 
>> inevitable splits of new coalitions (see DeWaal's "Chimpanzee 
>> Politics", 1982).
>> Let's see: Bloomfield challenged his teacher Hermann Paul's 
>> broad-scoped functional-typological coalition and split it, in the 
>> process radically narrowing down the domain of investigation to 
>> structure on its own and "inductive generalizations". Then Chomsky 
>> challenged Bloomfied's structuralist coalition and split it, creating 
>> a variant of structuralism--but with apriori Platonic universals 
>> propped upon rather frail empirical legs. Soon, 
>> Ross-Lakoff-Postal-McCawley challenged Chomsky's coalition and pulled 
>> out the Gen Sem faction, challenging autonomous syntax but adding 
>> only semantics to the relevant explanartory mix.
>> Then a bunch of us in the 1970s banded together into a new coalition, 
>> enlarging the relevant domain of structure's connectivity to 
>> communicative function (discourse), cross-language typology 
>> (diversity), diachrony (emergence), neuro-cognition (the processor), 
>> ontogeny (child language) and phylogeny (evolution). We were deeply 
>> interested in universals, but only if they were explanatory and 
>> empirically-grounded. It never occurred to us that "theory" was a 
>> dirty word, only that "theory" was not synonymous with "formalism". 
>> Our notion of "theory" demanded both empirical foundations and 
>> explanatory connectivity. But we were also deeply interested in 
>> structure--how could a linguist not be?--provided it were studied and 
>> explained in its relevant broad connectivity. This was, 
>> transparently, a return to the wide-scoped agenda of Hermann Paul, 
>> even if we hadn't read his work. So now a new coalition of alpha 
>> males are splitting off and, in a reprise of Bloomfield's maneuver, 
>> are narrowing the domain once again. This history is, leastwise to 
>> me, profoundly depressing. Try as I might, I find it hard to tell 
>> scientific substance from coalitional warfare.
>> One of the worst features of Chomsky's legacy to linguistics is his 
>> extreme reductionism, his insistence that you are either an 
>> empiricist or a rationalist, an inductivist or a deductivist, a 
>> universalist or variationist, a theorist or a data-monger, a 
>> formalist or a functionalist--with no room in the middle. As Esa and 
>> many others have pointed out, these are false dichotomies in the 
>> methodology of science, where an intensive cyclic interaction takes 
>> place among multiple strategies. But Chomsky's reductionist gambit 
>> also mis-represents mature theory-building, where universality and 
>> variability are inseparable, being both the products of development 
>> (‛emergence') in diachrony, acquisition or evolution.
>> Perhaps some day we will remember to remember that neither the 
>> structure-function ‛semiotic' relation nor the conundrum of 
>> universality-diversity, nor the central role of development in 
>> mediating the inter-dependence between these seeming extremes, are 
>> specific to linguistics. They have all been noted long ago in biology 
>> and evolution. For a biologically-based phenomenon such as 
>> language/culture to cleave to one extreme against the other, and to 
>> periodically resurrect these tired old false dichotomies, is not only 
>> counter-productive. It is also, perversely enough, an unintended 
>> validation of Chomsky's disruptive agenda.
>>
>> Peace, TG
>>
>> ================
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> Esa Itkonen wrote:
>>> Dear Funknetters: A debate has been going on in recent years 
>>> concerning the nature of cross-linguistic categories. Are they 
>>> universally valid psychologically real entities or (nothing but) 
>>> useful fictions? I think there is room for a third alternative, 
>>> namely one that exemplifies - to use a very refined expression - 
>>> "coherentist inductivism" Ã la Nicholas Rescher. This "third 
>>> alternative" is expounded more fully on my homepage, in the article 
>>> 'Concerning the role of induction in typological linguistics', which 
>>> is at the moment the latest entry in the list "available as full 
>>> texts". This is the written version of a talk that will be given at 
>>> the 4th Language - Culture - Mind conference, to be held in Turku 
>>> (Finland), June 21-23, 2010.
>>> Esa
>>>
>>> Homepage: http://users.utu.fi/eitkonen
>>
>



More information about the Funknet mailing list