Please join Daniel Siddiqi this upcoming Wednesday, October 4th from 3PM-4:30PM in Dunton Tower 2203 for this interesting talk:
Title: From Homophony to Unity
Presenting work co-authored with Bronwyn Bjorkman (Queens University) and Elizabeth Cowper (University of Toronto)
English word-forms ending with –ing (e.g. running, jumping, singing; variously called gerunds, gerundives, progressive/present participles) have long been a puzzle to grammarians and theoretical linguists. They are interesting for at least two reasons:
- They are perfectly regular (extremely rare in natural language). Every English verb has an -ing form, no -ing form is morphologically irregular, and only a handful of nominal -ing forms have unpredictable meanings.
- -ing forms can be of any lexical category: some seem clearly nominal, others clearly verbal, and others apparently adjectival. A small number are even prepositional.
These two properties have been the focus of much investigation in generative grammar. Chomsky (1970) used them to argue for what became a pervasive and controversial position in syntactic theory called lexicalism, and they represent convincing counter-evidence against a division between inflectional and derivational morphology. Faced with (II) in particular, many linguists have concluded that English must have several distinct-but-homophonous -ingsuffixes, or have suggested that morphology allows totally arbitrary mappings between form and meaning.
Using current Minimalist syntax, Distributed Morphology (DM), and contemporary event semantics, we instead argue for a unified account of -ing: all forms in -ing involve the same affix, which consistently realizes a single morphosyntactic feature. We propose that this feature expresses existential closure of the event argument of its syntactic complement: differences among uses of -ing arise from features that occur in addition to this core meaning.
A further interest of its analysis, beyond accounting for the synchronic distribution of -ing, lies in its implications for human language learning. We discuss the historical development of -ingfrom two historically distinct affixes. These affixes did not merge through otherwise regular sound change; sound change merely rendered them similar. We suggest that the cognitive learning strategies of speakers—in which contrasts can be maintained only if they are beyond some limit of perceptual salience—might play a role in leading speakers to a sound-meaning correspondence as seemingly odd as this one.