So, last semester I ran into some sleeping issues that ate up the time I might have spent blogging my morphology course. Since these issues are finally resolving, I hope to catch up and review some of what I learned last semester teaching the class. However, morphology continues this semester in a seminar I’m co-teaching with Stephanie Harves. Topic: Argument Structure and Morphology. While, inevitably, the seminar will end up discussing the biggies – causatives, applicatives, and nominalizations – we do intend to focus on particular issues to frame our reading of some current literature, particularly research of past and present students. Let me outline some of these issues in the interest of drumming up comments and suggestions from any blog-readers (and, yes, synthetic compounds should come up as well).

A key topic for the connections among syntax, morphology and argument structure is whether roots take arguments/complements, the alternative being that roots must be categorized as nouns, verbs, adjective or prepositions prior to the resulting head merging with complements. Here I see three major proposals, with some mixing and matching of subproposals across the three general lines of research. The approach associated with the Hale/Keyser position has roots in the position of arguments (and/or predicates) below a v/V head, into which they incorporate or amalgamate/coalesce. The second approach (H. Harley as a proponent) sees roots as argument-takers such that a vP would consist of a v and a rootP containing the internal arguments of the verb. The third approach, which I have been investigating lately, assumes that roots must be first-merged with a categorizing head (or another root) as an adjunct to that head such that any effects of a root on argument structure must be mediated by the categorizing head.

A related topic, one that we actually began the seminar with, although obliquely, is that of the nature of the categorizing heads n, v, and adj. We revisited a paper by Paul Kiparsky (Kiparsky, Paul. “Nominal verbs and transitive nouns: Vindicating lexicalism.” On looking into words (and beyond)(2017): 311) that argues against a view of the distinction between gerunds and complex event nominalizations that ties the difference simply to the height of attachment of a categorizing n head. On this view, the gerunds have the n head high, at least above voice and perhaps above aspect, while the complex event nominalizations have it low, say attached to the vP. Kiparsky’s main argument is that there is a sense of “noun” in which the event nominalizations are nouns but the gerunds aren’t. Although in English gerunds can take subjects with the possessive clitic (John’s running the race), and although in many languages the gerund itself may be a case-marked word that, morphophonologically, looks like a noun, Kiparsky argues that in English and other languages, the gerunds (which are part of the inflectional paradigm of a verb) don’t take adjectives or, for that matter, anything else associated with DP structure. He suggests, covertly following the analysis of Reuland (Reuland, E. J. (1983). Governing-ing. Linguistic Inquiry, 14(1), 101-136), that the gerund head has case attracting features but is not of category N/n.

We’ll be discussing Jim Wood’s recent book ms. on Icelandic nominalizations soon. Jim argues that Icelandic nominalizes verbs, not vPs, for complex event nominalizations. I think, given Kiparsky’s cross-linguistic insights about gerunds, this has to be true in general. That is, the category heads n, v and adj attach to categorized stems and/or to roots and begin the extended projection associated with their heads. Of course i*, then, must be treated as “derivational” in some sense, not part of the extended projection of a category head, since one can nominalize or verbalize or adjectivalize a structure with an i* (voice, etc.) head.

I’ll write more about these ideas soon, but in this context we must tip our hats to Hagit Borer, who argued that the category heads like n and v are really associated with the higher functional structure and not substantive morphemes. So, nouns are in some sense induced by higher DP type heads in a way that points to Kiparsky’s generalization about the distinction between gerunds (no DP structure) and complex event nominalizations (yes DP structure).