Roll out the Barrel!

OK, folks, the moment you've all been waiting for...

As SC noted before, this topic owes its existence to Geoff Pullum's glorious decision to coin it. In an essay published originally in Natural Language and Linguistic Theory, and later reprinted in "The Great Eskimo Vocabulary Hoax", Pullum remarked that "[S]ince I wrote the 1978 review, we have witnessed a new flowering of methodological moaning and self-serving cracker-barrel philosophy of science in the work of people who actually do produce publishable work in descriptive and theoretical linguistics" (p. 125).

Thanks go to David Elworthy for pulling up the quote.

Anyway, the reader may be saying "cracker barrel? huh?", and so the ever-helpful Semantic Compositions research staff provides the following background:

From Merriam-Webster Online: "cracker barrel, adj., Etymology: from the cracker barrel in country stores around which customers lounged for informal conversation : suggestive of the friendly homespun character of a country store "

A search on Google turns up, at this writing, 147 hits for "cracker barrel philosophy".

So what we're talking about is the tendency of people to be unable to resist -- especially in the context of the best human-temperature-raising device ever invented, the e-mail list server -- making statements about the scientific method which, at best, are quite intemperate, and at worst, are egregiously wrong. Cracker barrel philosophy isn't about being right, it's about the much-more-important feeling right ([which makes SC the Aristotle of the field -- ed.]).

Semantic Compositions spent a lot of time thinking about how to kick this off. On the one hand, SC is not an academic, and does not intend to become one in the future. On the other hand, SC would like to be able to visit conferences in the future without being garroted at the registration table. Thus, for the very first discussion of "cracker barrel philosophy of science", we are kicking things off with a topic which never causes any linguist, especially not any syntactician, to fly into a rage -- the question of what makes a theory "generative". SC has no plans to answer that question, just to illuminate it under the oil lamp (presumably hanging right over the cracker barrel).

Even the most casual non-linguist reader probably figured out from that last passage that "generative" is to linguists as a raw steak is to a pit of lions who have been deliberately starved for three weeks. SC prefers to think of it as a scene from The Princess Bride, where a character played by Billy Crystal is hectored by his wife, who runs around shouting the name of his enemy until he finally agrees to do what she wants. Those who have seen the movie will get it; it's not worth reconstructing in full for those who have not.

Once upon a time, Noam Chomsky coined the term "generative grammar" to refer to a grammar which either: 1) produced all and only the valid strings of a language, 2) was explicit in its characterization of the rules, 3) did one and two, or perhaps 4) had been blessed by Noam Chomsky or an anointed disciple thereof. Semantic Compositions does not actually subscribe to the notion that (4) was originally intended to be the case, but as this description of one 1999 conference's proceedings makes clear, some refereed papers lean in the direction of (4) even if they don't say it in exactly those words. And many linguists would argue that there has been a definite shift from (3) to (4) as the field has developed. Personally, SC subscribes to Groucho Marx's dictum (perhaps apocryphal; this search fails to confirm it reliably) that he "wouldn't want to be a member of any club that would have him".

Nevertheless, as an example of what the "generative" grenade can do when tossed into the right circles, SC presents this discussion from a mailing list he enjoys, HPSG-L. One day, an innocent undergraduate said: "I wonder what the people on this list have to say about the relationship between HPSG and generative grammar". Little did she know that she was about to spark a flame war where a single e-mail made reference to Pythagoras, Rembrandt, Euclid, Newton, and the Big Bang, all while laying out the answers to the vexing questions of "what is a theory?" and how to distinguish a theory from a model or framework (with the caveat that they were strictly one man's view). And another one threw down a gauntlet.

Semantic Compositions actually believes that insofar as any modern-day research program is close to developing a True Theory of Language, it's the one being followed by the people above, most of whom would refer to themselves as "lexicalists" (those aren't scare quotes, it's just not a common term). Anti-lexicalists are also guilty of overheated rhetoric; one conference paper contains the following: "Lexicalism is dead, deceased, demised, no more, passed on..." (ellipsis in original) and promises to replace it with "the alternative that allows us to dump lexicalism once and for all". SC wishes to further point out that the matter above did have a graceful resolution, but that it took place over too many messages to want to link them all -- just follow the thread and you can see for yourself. Committing an act of Cracker Barrel Philosophy of Science doesn't mean you don't know when to cool off.

Prof. Pullum also wrote that he believed this sort of discourse to be unique to linguistics, and that other disciplines were somehow free of it. In another essay from The Great Eskimo Vocabulary Hoax, he shared the results of his discovery that mathematicians were no better (SC is speaking from memory; he might be conflating essays). The truth is, Cracker Barrel Philosophy can be found anywhere that multiple sides want to lay claim to some truth. Or even just some term. And despite the intention of Semantic Compositions to stick to covering this sort of dispute in scientific disciplines, there are Cracker Barrel Philosophers in all walks of life. And Cracker Barrel Philosophies...by the barrel.

UPDATE: On further reflection, something could be clearer about why SC considers a flame war about generativity to be a matter of Cracker Barrel Philosophy: the episode in question didn't actually turn on any relevant facts about linguistics. Despite nominally being a discussion about whether or not HPSG and GB/Minimalism were mathematically equivalent, the most heated rhetoric all had to do with an at best tangential issue, namely which theory got to wear the name "generative". Getting into a dispute over a theory is not Cracker Barrel, but getting into a dispute over a term whose meaning nobody agrees on...well, like we said before, Cracker Barrel Philosophy isn't about about being right, it's about feeling right.

Readers might also object that the dictionary takes a rather less sour view of Cracker Barrel thinking than Geoff Pullum's original article, or SC's. And it's true, thinking in Cracker Barrel terms doesn't require you to be angry. It does, however, require you to insist that the terms be understood in a way that favors you, and from there, it's only a short hop to getting angry when disagreements arise. (added January 23, 2004 at 10:26 PM)

Comments

[language hat]

Groucho Marx's dictum (perhaps apocryphal; this search fails to confirm it reliably)

That's because you were using the wrong search string (a common problem). Here's a reference for you.

[dan]

I can't get used to this. Cracker Barrel is a brand of cheese where I am. This distorts the meaning of this somewhat (or perhaps adds to it).