Grammatical relation probability: How usage patterns shape analogy

Abstract

Abstract It has been argued speakers' knowledge of the probabilities of certain phones, words, and syntactic structures affects language production (Bell, Brenier, Gregory, Girand, & Jurafsky, 2009; Tily, Gahl, Arnon, Snider, Kothari, & Bresnan, 2009). This study provides evidence for effects of grammatical relation probabilities by identifying significant effects on verb morphology in the Spanish presentative [haber‘there (be)’+ NP] construction stemming from nouns with varying proportion of use in subject function. In addition to this novel type of probability (grammatical relation), we present calculations that are not context-dependent but cumulative, reflecting speakers' overall experience with these nouns in the grammar. We conduct variationist analyses on corpora of spoken Puerto Rican Spanish. Our results reveal that nouns with a high probability of subject function promote the analogical leveling ofhaberby increasing the likelihood of reanalysis of the object as subject of the construction. We interpret these results as suggesting speakers possess lexicalized knowledge of grammatical relation usage patterns.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 26 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.