Reflections on Universal
Grammar and the importance—or otherwise—of Language Diversity.
[This is the revised text of an earlier paper on Language Universals and Language Relativity: it appears in a Festschrift for my colleague David Rycroft, to mark his retirement from full-time teaching at Konan]
Download paper here
Download paper here
3 comments:
These striking structural similarities cannot come from a shared historical ancestor, nor from areal contact: the only remaining possibility—or so Chomskyan linguists conclude—is that they derive from a ‘biological blue-print’ for language which, like the menus in a restaurant, admit of a finite set of grammatical alternatives.
I'd like to take issue with your meta-menu of three selections. If your only choice, apart from shared ancestry and contact, is a biological blue-print, don't you presume too much about what can feasibly be encoded in the genome? Surely a fourth choice would be convergent cultural evolution — i.e. structural similarities due to the fact that languages are culturally evolving entities subject to selective pressures that may be similar across the globe (such as cooperative principles, communicative demands and social considerations of the Brown/Levinson 1978 and Sperber/Wilson 1987 type).
I always find it hard to square the enthusiasm of biolinguists about biological evolution with their entirely unmotivated avoidance of cultural evolution. Surely both these things must play roles in one's account of how complex systems like languages come to be shaped. Surely not everything can be ascribed to a biological blueprint (that is simply not how biological evolution works). And surely there is massive order beyond what is ascribed to the biological blueprint — order that is describable and that scientists of language would do well to try to explain.
To me, this combination of different evolutionary tracks is precisely what makes language so interesting. Shouldn't scientists of language — whether Chomskyans or not — be open to all the different ways in which linguistic diversity may be limited and constrained, instead of betting all their money on the genome or an admittedly possibly empty FLN?
I welcome your comment. Notice, however, the original text asserts "—or so Chomskyan linguists conclude—", not that this is a necessary conclusion: it is surely possible that there are other menu options. I am not trying to evade responsibility here; I simply don't know the right answer. But I'm at least as suspicious of the explanatory potential of "convergent cultural evolution" as of FLN. (I must admit that reference to "cultural" explanations of anything provokes an allergic reaction, especially when they purport to explain non-semantic, and parameterized features of language, such as the headedness of phrasal constituents).
As for the rhetorical question of whether scientists of language [should] be open to all the different ways in which linguistic diversity may be [...] constrained, I'm not sure that the answer is as clear-cut as you suggest: as is frequently rehearsed in debates about creationism, science proceeds as much by the apriori *setting aside* of certain types of explanation as by the embrace of all possible alternatives. If one applied your approach to evolutionary biology, one would constantly need to consider the possibility that this or that biological grouping—say marsupials—were introduced ad hoc by aliens, or that science should defer to religious opinion in developing hypotheses. I'm not going to get into that kind of debate, but the point should be clear.
I may not be the biggest fan of FLN, but there is at least a coherence and rigor to the approach that one can respect. I think I know what it would mean to disconfirm it. Culture is—to put it in the most charitable terms—a whole 'nother can of worms.
Thanks for your response Nigel. Your comment seems to suggest that some of the resistance against what I call "convergent cultural evolution" may derive from a knee-jerk reaction against the term "culture". To be clear, I said cultural evolution and not "culture" as a generic concept (indeed a can of worms). And when I say cultural evolution, I am not thinking of the kind of question-begging account à la Everett 2005, which to me assumes way too much about how cultural values could causally influence linguistic structure.
I'm thinking rather of well worked out proposals that take into account the socially transmitted nature of language, for instance Bill Croft's evolutionary linguistics, Simon Kirby's empirical-experimental work on artificial grammar learning in an iterated learning paradigm, and Nick Enfield's work on linguistic epidemiology and transmission biases as factors that may guide and constrain the structure of linguistic systems.
These are not "cultural explanations". They are just empirical approaches to language that try to take seriously the fact that a great deal of linguistic knowledge is acquired by means of social transmission, and that seek to investigate the extent to which that fact has implications for how linguistic systems are structured. To me, these approaches, when taken in combination with solid empirical work on the cognitive foundations of our linguistics abilities, are good examples of the virtue of being open to all the different ways in which linguistic diversity may be constrained.
Certainly these approaches are doing something radically different than invoking aliens to explain biological diversity. That analogy comes out of the blue and does no justice at all to the empiricist, process-based, cognitively grounded, psychologically and socially realistic nature of the work I'm referring to.
There may be coherence to the concept of FLN, but I feel it belies a very narrow conception of the nature of language. Whether or not it is empty, one needs to look elsewhere for a complete account of why languages are the way they are. So I just don't see the attraction of holding on to FLN and looking away from all the exciting interdisciplinary empirical work that is currently being done. That was where my original comment started: limiting the menu to three options, as Chomskyan linguists do, just seems to be to prejudge the issue and to make the study of language so much less interesting.
Post a Comment