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
23 September 2002

Doug Blank (Computer Science), Carol Bernstein (English), Anne Dalke (English), Paul Grobstein (Biology), Mark Lord (Theater), Sharoukh Mistry (Biology), Eric Raimy (Linguistics), Kathryn Rowe (English), George Weaver (Philosophy)

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):

The week's readings were Bickerton, Derek. Creole Languages from Scientific American and Crane, Mary Thomas. Introduction from Shakespeare's Brain: Reading with cognitive theory.

This week's discussion began with Kathryn raising a few points of contention with the Bickerton article.  The first one was to question whether certain constructions found in Hawaiian Creole were actually lacking in English.  The second point was questioning exactly what language was used in 'bilingual' families in pidgin situations.  On the first point, Eric suggested that the particular examples chosen by Bickerton may be incorrect but that other more robust ones can also be found.  Consequently, this point was still valid in Bickerton's argument.  The second point was much more unclear from Bickerton's perspective.  In familiar cases of bilingual families the parents know enough of the two languages that they can communicate with each other using one or the other language.  This provides enough input for the child to learn at least one of the languages as complete system.  The Bickerton article suggests that parents in a pidgin situation do not know each other's language and only communicate with each other through the use of the pidgin.  Thus, the child is forced to creolize the pidgin because they do not get enough input from one of the parent languages to learn it instead.  This is a sketchy situation because we would assume that the parents have a full language as their native one.  So this raises the question as to why the children of these parents do not become bilingual instead of just monolingual in the creole.  This issue was not really settled or investigated further by the group.

Paul highlighted his interest in the Bickerton article that the similarities in all creole languages suggests that there are 'defaults' (such as double negatives) in human languages.  When a child does not receive enough input due to being only exposed to a pidgin, they rely on these defaults to fill in grammatical aspects of the new creole.

Doug pointed out that the notion of 'default' is being rethought in some areas of cognitive science.  Instead of a 'hard wired and specific' default as suggested by Bickerton's biological language approach, defaults can be produced using connectionist modelling.  This raises the question as to the origin of these defaults.  Are they actually innate instructions provided by genetic sources or do they emerge from the interaction of more basic genetic instructions that determine how language is analyzed?  This topic was also not settled or investigated further.

Another discussion of the idea of 'defaults' especially related to double negatives was whether the use of double negatives in English was determined by learning to speak in a 'civilized' manner.  The idea is that children are not fully acculterated for a period of time and maybe this is the window of time where they 'default' to the use of double negatives.  As time progresses and the child learns the practices of culture in English they learn to not use double negatives.  A question was raised about how to define 'civilized' because French uses double negatives (more correctly termed 'negative concord').  Is French then an 'uncivilized' language?  The group thought 'no' and placed the difference between English and French as a difference in culture.  Its 'civilized' to use double negatives in French but 'uncivilized' to use them in English.

The discussion then moved onto the Crane article.  In general the group was interested and happy enough with the introductory chapter to agree to read one of the other chapters that provides a case study of a particular term in a particular play.  Various lines of critical discussion of Crane's ideas did emerge though.  (In no particular order.)

One issue was whether choosing Shakespeare as the object of study was a good idea.  The general question was whether we would get a representative picture of the average human mind by analyzing Shakespeare.  Everyone agreed that this was an important question because although the group as a whole would not accept applying the term of 'genius' to Shakespeare everyone did agree that he was exceptional.  One problem that the group discovered was that no one could come up with an example of a 'typical' author.  Mark even suggested that no author is 'typical'.

This lead to a discussion about whether using the methods outlined in Crane on a different author would produce similar results.  Paul and Anne had discussed earlier about doing a 'cognitive' analysis of Emily Dickinson.  Paul was very optimistic that we could learn new things about cognition and Dickinson through this endeavor.  Anne was pessimistic about this proposal though.

Carol raised the general question as to whether Crane's couching of her analysis as 'cognitive science' served any purpose.  She could not tell whether the analysis that Crane produces of Shakespeare's work would really be any different than the Derridaian or Foucaultian analyses that Crane was arguing against.

Another important issue that was identified from Crane's work was the difference between 'reading' and 'interpreting' texts.  Anne pointed out that this distinction was the probable source of many times where the group talked past each other.  What was noted was that it is difficult if not impossible for the 'humanist' approach to not 'interpret' a text.  This was seen very clearly in last years discussion of Stephen Pinker's work.  Both Anne and Kathryn were 'interpreting' Pinker's writing while others in the group attempted to insist on only 'reading' the text.  Because these two useful but distinct approaches weren't overtly identified and appreciated at the time, the discussion of Pinker's work was not very useful or successful.  Now that the group has identified this issue we can move to discussions where each tool is used at appropriate times to create more productive exchanges of ideas.

The group agreed to read another chapter of Crane's work (to be chosen by Kathryn) and a selection from Beckett.