Any event can be construed from a variety of perspectives. While this flexibility is fundamental to human ingenuity, it poses a challenge for language learners who must discern which meanings are encoded in their language and by which forms. The papers in this dissertation focus on verbs encoding directed motion (e.g., a girl runs into a house) and caused change-of-state events (e.g., a boy blows out candles). Both classes of events can be expressed by verbs that lexicalize different components of the event, namely Manner-of-motion (e.g., run) or Path (e.g., enter), and Means (e.g., blow) or Effect (e.g., extinguish), respectively.
Papers 1 and 2 examine the representation of higher-order generalizations about the meanings of directed motion and novel caused change-of-state verbs. Both studies use a novel verb-learning paradigm to manipulate the meanings of novel verbs in the input and then assess how learners interpret subsequently encountered novel verbs (measure lexicalization bias). The results indicate that learners rapidly use semantic regularities to form higher-order generalizations about verb meaning.
In Paper 1, adults taught Manner verbs construed new directed motion verbs as lexicalizing Manner more often than those taught Path verbs. Moreover, changes in verb learning bias were accompanied by shifts in visual attention: Manner-verb learners fixated on Manner-related elements of visually-presented events more than Path-verb learners. These results indicate that previously observed cross-linguistic differences in verb lexicalization biases are unlikely to stem from the restructuring of semantic representations along language-specific lines and more likely reflect the operation of a flexible, inferential learning mechanism that monitors the input and updates beliefs accordingly.
Likewise, in Paper 2, adults taught Means verbs interpreted unknown verbs for caused change-of-state events as encoding the Means more often than those taught Effect verbs. Unlike directed motion verbs, the encoding of these events is not characterized by marked typological variation and the availability of Means and Effect verbs does not appear to vary appreciable within or across languages. Our results, then, suggest that the formation of higher-level generalizations about meaning is a fundamental property of the processes that undergird lexical acquisition.
Paper 3 focuses on the representation of the event concepts that underlie verb meanings. Specifically, we examine the possibility that Manner-of-motion and Means are actually instances of a broader semantic category, MANNER, whereas Path and Effect are instances of a different semantic category, RESULT. Adults were taught novel verbs for either directed motion or caused changes of state and subsequently presented with novel verbs from the other semantic class. The results revealed that adults transfer newly-learned higher-order generalizations about the meanings of directed motion verbs to caused change-of-state verbs (and vice versa), providing support for the psychological reality of superordinate event concepts. / Psychology
| Identifer | oai:union.ndltd.org:harvard.edu/oai:dash.harvard.edu:1/17467171 |
| Date | 17 July 2015 |
| Creators | Geojo, Amy Celine |
| Contributors | Snedeker, Jesse, Pinker, Steven, Spelke, Elizabeth, Carey, Susan |
| Publisher | Harvard University |
| Source Sets | Harvard University |
| Language | English |
| Detected Language | English |
| Type | Thesis or Dissertation, text |
| Format | application/pdf |
| Rights | open |
Page generated in 0.0131 seconds