difference in form without difference in meaning

[Frederick J Newmeyer]

Dear Florian, et al.,

Thank you all so much for your contributions to the line of discussion that I initiated. There is no way that I can give point-by-point commentary on all of the postings, but then nobody would expect that I should do so. Just a few comments.

First, it's clear -- and Florian cited several references -- that variants might differ not so much in their meaning (even broadly defined), but rather in *how relatively effectively* they can convey a particular meaning given particular discourse and other background conditions. So subject sentences ('that he'll go home is likely') may or may not have identical meanings as extraposed sentences ('it's likely that he'll go home'). But clearly, conditions that are to a degree meaning-independent are at work in speaker choice of one variant over another: the length of the subject, the stylistic register, and so on. One could make the same point with respect to heavy-NP-shifted items versus non-shifted ones. The different focal properties associated with the different positions (which we can think of as aspects of meaning) are relevant, but do not suffice to explain fully why some NPs are shifted and some are not.

One of the most frustrating facts for the theoretical linguist is that the analyses that we come up with are not always (possibly not often) confirmed by particular psycholinguistic studies. And here the problem cuts across theoretical frameworks. Consider for example the abstract generative phonological analyses based on alternations; the minisculey-fine semantic distinctions posited by cognitive linguists as a basis for syntactic structure; and the functionally-motivated hierarchies that form a basis for a lot of functionalist theorizing. The conflicting experimental results with respect to the 'psychological reality' of these various analyses have led a lot of grammarians to be cynical about what psycholinguists can offer them as an aid to or as a check on theory construction. That's lamentable of course.

In his second posting, Florian referred to 'functional theories of meaning differences', citing work by Fox, Thompson, and Mulac. These are really at one extreme end of the functionalist spectrum, given the role that they impart to 'fragments' and 'memorized formulas' as being at the centre of language, as opposed to grammatical processes (as the term is understood within whatever framework). These fall down in explaining how languages users have the ability to *interpret* input that they have not previously encountered. As I argued in Newmeyer 2010, this interpretive capacity (among other things) points to the need for a stored grammar.

Newmeyer, Frederick J. 2010. What conversational English tells us about the nature of  grammar: A critique of Thompson's analysis of object complements. Usage and structure: A Festschrift for Peter Harder, ed. by Kasper Boye and Elisabeth Engberg-Pedersen, 3-43. Berlin: Mouton de Gruyter. The paper is also available on LingBuzz: http://ling.auf.net/lingBuzz/000679.

--fritz


Frederick J. Newmeyer
Professor Emeritus, University of Washington
Adjunct Professor, University of British Columbia and Simon Fraser University
[for my postal address, please contact me by e-mail]