Papers
Lee, Naomi. Under review. Learning (im)possible number syncretisms: investigating innate featural representations. [Manuscript]
I present the results of an artificial language learning experiment that explores whether a binary, cross-classifying, or a privative, linear hierarchical system of representing number features better explains participants’ performance on three different grammars. All grammar conditions marked a three-way distinction between singular, dual, and plural on nouns, but each displayed a different logically possible syncretism in verbal number agreement. I find that participants learning the dual-plural (ABB) syncretism perform significantly above chance on referent selection tasks (p < 0.001), whereas those learning the singular-dual (AAB) syncretism fare significantly worse than that baseline (p < 0.001), providing tentative support for a linear containment hierarchy representation for number.
Lee, Naomi & Laurel MacKenzie. 2023. Social role effects on English particle verb variation fail to replicate. Canadian Journal of Linguistics / Revue Canadienne de Linguistique 68, 329–343. [Open Access]
Laurel MacKenzie and I analyze 10,521 tokens of variable English particle verbs from the RadioTalk Corpus (Beeferman et al. 2019) to try to replicate Kroch & Small (1978)’s argument that radio show hosts and in-studio guests’ greater sensitivity to prescriptive norms makes them more likely to use the joined variant of the alternation than listeners calling into the show from elsewhere. Our analysis confirms that direct object length, register, a measure of frequency, semantic compositionality of the particle verb, and the particle’s prosody all condition the alternation. However, the effect of social role does not replicate.
Presentations
Lee, Naomi. 2024. The Nanosyntactic prediction that each language can only have one declension class containing gender doublets is false. Talk, 55th Annual Meeting of the North East Linguistic Society (NELS 55). Yale University, New Haven. [Poster / Handout]
I show how nanosyntactic approaches that seek to encode “declension class without declension features” make a number of empirically falsifiable predictions about the relationship between grammatical gender and declension class.
Lee, Naomi & Laurel MacKenzie. 2021. Does the English particle verb alternation show gradient sensitivity to compositionality? Talk, New Ways of Analyzing Variation (NWAV) 49. UT Austin (online). [Video]
Laurel MacKenzie and I argue that categorizing English particle verbs based on their syntax better captures the variation between joined and split orders than a coding scheme based on particle verbs’ semantic compositionality.
Lee, Naomi. 2020. Learning (im)possible number syncretisms: investigating innate featural representations. [Slides]
+ Talk, North East Linguistics Society (NELS) 51. UQAM (online).
+ Invited talk, Center for Language Evolution, University of Edinburgh (online).
I present the results of an artificial language learning experiment that explores whether a binary, cross-classifying, or a privative, linear hierarchical system of representing number features better explains participants’ performance on three different grammars. All grammar conditions marked a three-way distinction between singular, dual, and plural on nouns, but each displayed a different logically possible syncretism in verbal number agreement. I find that participants learning the dual-plural (ABB) syncretism perform significantly above chance on referent selection tasks (p < 0.001), whereas those learning the singular-dual (AAB) syncretism fare significantly worse than that baseline (p < 0.001), providing tentative support for a linear containment hierarchy representation for number.
Lee, Naomi and Ailís Cournane 2019. The journey, not the endstate: finding innovation in the dynamics of L1A. Talk, Diachronic Generative Syntax (DiGS) 21. ASU. [Slides]
Ailís Cournane and I argue that the character of diachronic innovations is more driven by the learning strategies used by all children than by principles of economy. Specifically, we advance an unraveling approach to learning the morphosyntactic organization of features and show how an “unraveling” child could stay at earlier stages that mirror innovations that are both canonically economical (e.g. Romance determiners (head) from Latin demonstrative ILLE (specifier)) and noncanonical (e.g. innovating Somali’s bundling of gender and number on a single head).
Lee, Naomi 2019. Khoekhoe participant phi-features: evidence from allomorphy and possession. Talk, Penn Linguistics Conference (PLC) 43. [Slides]
Winner of a Best Abstract Prize. I argue that “person” in Khoekhoe is best represented as separate [speaker] and [addressee] participant features, in two distinct syntactic positions within the DP – spec-NumP for [speaker, participant], and spec-DP for [addressee, participant].
Lee, Naomi 2019. Khoekhoe pronominal morphosyntax: gender on Root-attached little n. Talk, Linguistic Society of America (LSA) 93rd Annual Meeting. NYC. [Slides]
I argue that Khoekhoe pronouns and lexical nominals share the same full DP syntax, including a nominalizer head that bears gender features. Lexical Roots appear nestled within pronominal forms, which display syntactically conditioned allomorphy, supporting an identical, articulated DP structure for all Khoekhoe (pro)nominals.
Lee, Naomi 2017. Old English Poetic Alliteration as Correspondence. Talk, 19th SYNC (Stony Brook, Yale, NYU, CUNY) Conference.