Bryn Mawr College

Center for Science In Society

To facilitate the broad conversations, involving both scientists and non-scientists, which are essential to continuing explorations of
  • the natural world and humanity's place in it,
  • the nature of education,
  • the generation, synthesis, and evaluation of information,
  • technology and its potentials,
  • the relationships between forms of understanding.

Language: A Conversation

Meeting Notes
1 April 2002

Doug Blank (Computer Science), Carol Bernstein (English), Sharon Burgmayer (Chemistry), Anne Dalke (English), Paul Grobstein (Biology), Ruth Guyer (General Studies, journalist), Eric Raimy (Linguistics), Kathryn Rowe (English), George Weaver (Philosophy), Sharon Burgmayer (Chemistry) 

One Summary View (prepared by Eric Raimy; views by other participants encouraged and can be sent either by email or posted using our working group forum area):

This week's discussion began with Paul clarifying and restating his question/objection to the 'Pinker-style' analysis of language.  This objection is that this approach to language assumes that there is a stable base that underlies the patterns that linguists observe and base their theories/analysis of language on and Paul questions (actually wants to deny) that there is a stable underpinning for language behavior.  A reflex of Paul's objection is that the claim that language is 'rule based behavior' is fundamentally incorrect and leads us to the wrong conclusions about the nature of language.

The alternative to 'Pinker-style' language analysis that Paul suggests could be useful (from an idea originally posed by Anne) is that language is play.  This approach would deny that there is any stable underpinning for language other than a 'desire' to play with the possibilities (or possibly explore if we use a term Paul has used previously) that could be language.  Part of the allure of this position to Paul is that it incorporates as a fundamental aspect of our view of language randomness and non-stability.  Adopting this view would radically change how language is studied because it would create a different set of questions to be asked about language from the set of questions being asked by the 'Pinker-style' research program on language.

Eric was very concerned with the implications and claims made about language that result from the 'play theory' of language.  The first question that Eric had (which was also shared by Ruth) was about the claim that there was no stable underpinning of language behavior.  Eric prodded Paul about how he could be sure of this or what kind of evidence that he had to support this position.  In the end, it was concluded that in order to determine whether there was a 'stable underpinning' for language or not we would have to have basically an entire explanatory theory of language.  Once this goal was achieved, we could then see whether stability was a fundamental characteristic of language or not.  We are obviously no where near this type of understanding of language and language behavior and consequently have to leave the issue of the 'stability of the underpinnings of language' as an assumption.  It should be recognized that at the present time though, because of lack of evidence, both the assumption that language has a stable underpinning and the contrary assumption that there is no stable underpinning are equally viable as beginning points for the study of language.

The conversation continued with Eric still questioning the usefulness of the 'play theory of language'.  Eric did concede that the idea of play would be useful in the area of language acquisition since according to the 'pinker style' analysis of language, acquisition is viewed as the maneuvering through possible grammars until the target grammar of the ambient language is reached.  This view appears (to me at least) to be very much how Paul wants to view language.  Acquisition of language does appear to be an exploratory process where the learner 'plays' with different possible grammars until the target grammar (or a grammar that is close enough to the ambient language) is reached.  Even though the use of 'play' can be seen rather easily here, Eric still had concerns with the position that Paul was espousing.

Eric's primary concern with the 'play theory' that Paul was trying to flesh out was where the primitives of language come from.  For 'Pinker style' theories, the primitives of language are supplied by universal grammar and the exploration process or 'playing' that occurs in language acquisition is limited by what the primitives are.  In other words, the primitives provided by universal grammar delimit a 'possible grammar space' that limits what kinds of grammars that a learner can even conceive of.  Within this type of theory, it is probably correct to view the learner as exploring this grammar space while they are acquiring a language because there is in general uniformity in the end result of acquisition (a child surrounded by English as the ambient language ends up learning English) but there is not necessarily uniformity in the acquisition process (different children start talking at different ages, some babble and some don't, some children [this is rare] don't babble or speak in single words and one day just start talking in complete sentences, etc.).  For Eric though, the role that universal grammar plays in providing the primitives of language is crucial.

Paul would like to allow the 'play' process to derive the primitives of language in addition to explaining  the overall structure of language too.  There is more than this for the 'play theory' though in that it appears that Paul wants to question whether there are primitives of language, whether there is a stable structure of a language and other fundamental assumptions of 'Pinker style' analysis of language.  In support of these questions, we can consider previous discussions of the language group.  We have already discussed the fact that everyone's language is different in that no two people speak in exactly the same way.  We each have different accents, different styles of speaking and can argue over exactly what a particular word means (and in some cases whether the word exists or not!).  These facts are the root of Paul's questioning of the 'Pinker style' language analysis.  When taken at the surface form, the micro-variation between speakers of the 'same' language does raise the exact question that Paul asks.  Is 'language' really as stable as linguists analyze it to be?.....

At this point, time had run out.  Both Paul and Eric apologized to the group for monopolizing the discussion with questions that they can be asking between themselves, and which didn't necessarily represent either the center of concern of the group as a whole or individual interests of many other participants. It was recognized that the group focuses better and produces a much more even discussion when we have a particular reading to discuss.  It was then decided that a selection from George Lakoff's Philosophy in the Flesh would be chosen as the next reading and distributed to everyone in the group.