Friends of Semantic Compositions

January 2009

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31

Site Statistics

Blog powered by Typepad

« Self-eating food | Main | I would never say that »

December 28, 2004



Just FYI, the article in question is in Volume 35 Number 1 (Winter 2004), entitled "A Note on the Weakest Crossover" by Eduard Gerbrandus Ruys. Not versed in Minimalist jargon or theory, I can't (or is to 'don't want') to wade through the article for the actual argument made, though that's not really the point as far as the anti-Chomsky argument goes.

Semantic Compositions

Thanks for the reference. I'm getting less and less current on Minimalism as grad school recedes into the past, so I might not be able to decode the article at this point. It's safe to guess that casual readers just coming across the article that prompted this post won't fare much better. Which is bad for us as a field.


Why is it bad for linguistics as a field if people with no training in it can't read technical articles? Surely some background knowledge should be assumed of readers of Linguistic Inquiry?

Semantic Compositions

Oops, bad organization on my part. I didn't mean that it's bad that people without technical knowledge can't read LI; that ought to be expected. What's bad is that they're left with the bald assertion that terms like A-position and c-command are just so much specious academic puffery.

Whatever my misgivings about the specific theories in question, I would never defend the notion that theoretical syntax is a dubious subject in principle. Unfortunately, I don't see how a lay reader could come across this article and not be tempted toward that conclusion.


The main problem with modern linguistics is that it has not been "heuristic" in decades. It builds clever models but discovers nothing but the useless and trivial. It reminds me of the old Junggrammatiker line: "mehr scharfsinnig als wahr."


I presume you mean trifles such as investigating into the possibility of a universal base-order for all languages (as the Linear Correspondence Axiom holds), which is tested for numerous and an ever increasing number of languages since 1994. By this program, all languages basically behave in parallel, differing only in the way functional projections are active or not. Or such trivial ideas such as that the language faculty is perfect, one tenet of a research program since the mid 1990s. Or such uninteresting discoveries that there are covert elements that don't appear at the phonological surface of a sentence: pretty trivial and rather old stuff indeed.
Concerning the Junggrammatiker: what is at stake is not the truth of a theory alone, but whether or not it is feasible. In current syntactic theorizing researchers constantly test the limits of theories and test whether they are true or not. What is interesting is not so much if they are true or not (truth is a concept that has a weird place in science, in my opinion), but what follows from theories, and if the consequences are desireable or not.

By the way: if by heuristics you mean "a method to help solve a problem, commonly informal. It is particularly used for a method that often rapidly leads to a solution that is usually reasonably close to the best possible answer." (wikipedia) This is exactly what modern linguistics has. Unlike traditional grammar, however, many of these guesses can be empirically and precisely tested.

The comments to this entry are closed.