Mon, November 23, 10am – 11am
Glaser Auditorium (2nd floor), LRDC 3939 O'Hara St. (map)

EFFECTS OF WITHIN AND CROSS-LANGUAGE SEMANTIC AMBIGUITY ON LEARNING AND PROCESSING

Chelsea M. Eddington, M.S.

November 23rd, 2015 10:00 am Glaser Auditorium (2nd floor),

LRDC 3939 O'Hara St., Pittsburgh, PA 15260

The majority of English words have more than one meaning, which is referred to as semantic ambiguity. Cross-language translation ambiguity occurs when a word in one language has more than one translation in another language. Within and cross-language ambiguity affect the learning and processing of words in the first language (L1) and the second language (L2) (e.g., Armstrong & Plaut, 2008; Degani & Tokowicz, 2010). This dissertation examined how semantic ambiguity and meaning/sense dominance affects learning and processing across three vocabulary-training experiments. Experiment 1 examined how semantic similarity impacts the learning of novel meanings for previously-known word forms. We found that participants correctly recalled more associates for words with related meanings than words with unrelated meanings. However, no differences were found between words with related and unrelated novel meanings on a primed lexical decision task. Experiment 2 examined how semantic similarity of L2 vocabulary (German) affected the learning and processing of these words using vocabulary tests and a semantic relatedness task. Participants were slower and less accurate at recalling and translating words that had less related translations (e.g., Trunk – Rüssel (elephant), Kofferraum (car)) than words with more related translations (e.g., Sheet – Laken (bed), Blatt (paper)) and were slower and less accurate responding to subordinate translations than dominant translations. Experiment 3 examined how L2 learners mapped meanings from ambiguous English words to L2 vocabulary by teaching participants only one translation that corresponded to one meaning of a semantically-ambiguous word (e.g., Trunk – Rüssel (elephant)). Using a translation-recognition task in which the critical “no” response items were semantic distractors (e.g., responding “no” that Nose is not the correct translation of Rüssel (trunk)) we examined how the trained (e.g., nose) and untrained (e.g., car) meanings interfered with semantic processing of the L2 vocabulary by comparing responses between semantic distractors and controls (e.g., responding “no” that keyboard is not the correct translation of Rüssel). Results demonstrated that participants extended the trained and untrained meanings to words with related meanings (e.g., wrapping paper vs. academic paper) but only the trained meaning for words with unrelated meanings (elephant trunk vs. car trunk). Overall, this dissertation sheds light on the interplay between meaning dominance, meaning similarity and context and how these factors influence how meanings and words are connected in the brain. The results of this dissertation provide a better understanding how monolinguals and L2 learners process and learn semantically-ambiguous words, and informs models of monolingual and bilingual semantic memory. Chelsea Eddington PhD Candidate University of Pittsburgh and the Center for the Neural Basis of Cognition