The Phoneme: An Interview With Elan Dresher

In this episode, we're speaking with Elan Dresher, professor Emeritus at the University of Toronto.

Two things stand out from this interview like a sore thumb. The first is something I said which was at best controversial, and at worst just plain wrongheaded. The second is something which it seems to me was left tragically under addressed.

(1) The Structure of the Phoneme

  

The basic idea is simple (and has its origins in Fodor's Hume Variations) : the notion of “contrast” is of significant vintage in phonology, having its origins in such sources as Sapir, Jakobson, and the Prague school. And yet, it’s also been a central notion in generative grammar where a great deal of other structuralist notions have been eschewed.

In some ways, the idea of individuating (psychological) entities in virtue of their contrasts in a given schema is more in tune with the Pragmatist tradition, and indeed it seems rather isolated in linguistic theory to phonology and lexical semantics.

Thus, the obvious question : why should we continue to act on the belief that the contents of a mental particular is just whatever possible contrasts that it can sustain between itself and every other mental particular in a given system?

As far as I know, (and I know very little) it’s just not a question that bothers the phonologists in my circles -- even phonologists who are committed to explicating their corner of the language faculty in terms of a naturalist, realist psychology of language, who would otherwise consider themselves anti-Pragmatist.

To respond to this question in good faith would mean not just arguing convincingly that if phonology were to pivot around the notion of contrast then the data of externalization is explained with a neat formalism, but also that the concept has some independent theoretical motivation.  

As has been noted elsewhere, a number of questions confront the concept of contrast. Two of them are:

a) Holism: If a unit in a schema is defined solely by its relationship to every other unit in the schema then how can it be learned individually -- since grasping its character is a matter of grasping the relationships it has with all of the other units.

b) Substance: Concretely speaking, what is it that speakers contrast -- abstract mental symbols? acoustic features? articulatory features? 

(2) Phonetics-Phonology Interface

Although we briefly discuss the relationship between the two domains in this interview, we really don’t do the topic justice. For those listeners interested in an in-depth analysis, a good place to start might be Thomas Purnell's (2009) Phonetic Influence on Phonological Operations. 

One issue left unconsidered in this interview is that of the relationship between acoustic cues and the mental symbols they typically token in speakers. As Purnell observes, the relationship between these two entities is neither direct nor predictable simply on the basis of the acoustic information. To see this, consider the observation that acoustic properties are just the physical properties of a stream of sound -- they are layered, continuous, multitudinous -- while the mental symbols they token are in important respects atomistic and categorical. Acoustic cues are quite variable from speaker to speaker, as well as within a speaker, yet the behaviour of interpretation is astoundingly robust. 

Moreover, a single mental symbol (phonological feature) may be tokened by different acoustic cues. This suggests that the ground separating phonetics from phonology has some depth worth considering. 

In essence, this just recapitulates a very old argument about the origin of structured knowledge. Empiricists hold that the origin of structured knowledge is situated in the environment, which the mind reflects; rationalists hold that the origin of structured knowledge is situated in the mind itself, and whose final form is the product of an ongoing interaction between innate schemas and the raw stuff of the world. Doubtless the distinction as described above is whiggish, but it's enough for present purposes.

That's all for now.... 

....

Next Episode: We're speaking with the University of Michigan's Robin Queen. Stay tuned.