A question for Fritz
dick at ling.ucl.ac.uk
Sun Oct 24 19:04:21 UTC 2010
Fritz: Just in case you haven't seen this already. Dick
Ferreira, Fernanda (2005). Psycholinguistics, formal grammars, and
cognitive science. /The Linguistic Review/ *22*. 365-380.
ABSTRACT: In the 1980s, Charles Clifton referred to a "psycholinguistic
cognitive science. During that time, there was almost unanimous agreement
that any self-respecting psycholinguist would make sure to keep abreast
developments in generative grammar, because a competence model was
and the linguistic theory was the proper description of that competence.
But today, many psycholinguists are disenchanted with generative grammar.
One reason is that the Minimalist Program is difficult to adapt to
models. Another is that generative theories appear to rest on a weak
foundation, due to the reliance on informally gathered grammaticality
What can be done to remedy the situation? First, formal linguists might
follow Ray Jackendoff's recent suggestion that they connect their work more
closely to research in the rest of cognitive science. Second, syntactic
should develop a better methodology for collecting data about whether a
is good or bad. A set of standards for creating examples, testing them
on individuals, analyzing the results, and reporting findings in
should be established. If these two ideas were considered, linguistic
might once again be relevant to the psycholinguistic enterprise.
Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm
On 24/10/2010 18:25, Brian MacWhinney wrote:
> By now, Fritz clearly has enough for his brief commentary. Everything mentioned on this issue so far is accurate, according to my knowledge, but let me add a few more wrinkles.
> 1. Regarding cultural anthropology, I always teach my students in Crosscultural Psychology that Linguistics had an enormous influence on the development of both Structural Anthropology and the subsequent Cognitive Anthropology. The influence on structuralism was through views such as Goodenough and others who likened kinship systems to the distinctive feature systems of Prague School phonology. Systems of binary distinctions were at the heart of Herb Simon's EPAM model of thinking and memory. Both Jakobson and Simon thought that the mind could be viewed as a digital computer and so binary features were crucial. Later, with the rise of transformation generative grammar, the emphasis shifted to rules of grammar as models for rules of culture. The major flourishing of this was in the 1970s, a bit later than the 1960s noted earlier. Personally, I thought this stuff was fascinating. My understanding is that the demise of this linguistics cum psychology in cultural anthropology was due not to failures in linguistics, but to the rise of deconstructivism in ethnography.
> 2. Alex is roughly right about Searchinger. Gene spoke to me on the phone about my interests and I explained that I focused on language learning and emergence. He said "thanks" but that this was not what he was trying to develop in this series. Liz Bates and Catherine Snow had the same experience.
> 3. The situation with regard to physics and biology is a bit complex. Often, people in those areas simply assume that Chomsky speaks for linguistics and use his framework for testing of their ideas about system functioning. I often get such papers for review and they do not show any lack of respect for linguistics, just a tendency to not understand the range of variation of analyses within linguistics. Often the analyses they offer in applying ideas from genetic diffusion or statistical physics (Nicolaidis et al.) are more compatible with these alternative views.
> 4. The major area that has been left undiscussed and which in my mind is the potentially most important is computation. Here, there is the famous claim by IBM that every time they fire a linguist they improve their grammar checker. I guess that counts as lack of respect. On the other hand, the basic linkage of generative theory to formal grammars back in the 1950s was a big deal. In automata theory classes and textbooks, students still learn about the Chomsky hierarchy, although much recent work suggests that other characterizations are more effective for resolving issues in grammar induction. More recently, the emphasis on data-mining of the web as a bag of words seems to have hit a bit of a wall and researchers are showing increasing interest in and respect for linguistic analysis. And there is the issue of computational resources for endangered and under-documented languages. Here, people like Lori Levin and colleagues are finding that computational linguists trained only in the use of HMM and SVM are unable to understand the challenges of real linguistic structure. So, there are important areas here involving a beginning of interest in reintroducing linguistics.
> 5. Finally, I wish that I could refer to Conversation Analysis as a part of linguistics. I know that I can't really get away with this, although personally I think it is a part. In any case, I see a lot of interest and respect for CA from areas as diverse as marketing, sociology, politics, aphasiology, and so on.
> -- Brian MacWhinney
More information about the Funknet