I nearly failed out of grad school, defending Chomsky's theory of syntax. Half a decade later, I'm done pretending it was worth it.
Chomskyan generative grammar — X-bar theory, Government and Binding, the Minimalist Program — was taught to me at the University of Pennsylvania as the only legitimate science of language. It was the gatekeeper, the screener, the thing students were washed out of linguistics PhD programs over. As I've come to discover, decades of work in dependency grammar and construction grammar — frameworks I was told didn't exist, didn't matter, or had been "subsumed" — were doing better empirical work the whole time.
In this video:
- What Chomsky actually got right (the cognitive revolution, generative grammar as discrete infinity, the takedown of Skinner's Verbal Behavior)
- Where transformational grammar, deep structure, movement, empty categories, and Universal Grammar go wrong
- The "poverty of the stimulus" argument and why Pullum & Scholz's critique matters
- How construction grammar (Adele Goldberg) handles the active/passive, the dative alternation, "the more the merrier," and coercion — without movement
- How dependency grammar (Lucien Tesnière) handles headedness, raising vs. control, and cross-linguistic data — without phrase structure trees
- Why long-distance reflexives in Mandarin, Icelandic, and Japanese broke Binding Theory
- Why Minimalism's proliferation of functional projections (TP, vP, FocP, ForceP) looks an awful lot like Ptolemaic epicycles
- Usage-based linguistics (Tomasello, Bybee), psycholinguistics (Levelt, Ferreira), and what kids actually do when they learn language