Graduate research milestones (PhD generals papers, March 2017)

This is a new blog series profiling graduate research progress in our department, particularly PhD generals papers (each student does two, before starting their dissertation) and MA forum papers. The following are recently completed (within the past few months) or nearly completed GP2s from our current PhD 3 cohort.

Na-Young Ryu (phonology, psycholinguistics)

In my second generals paper, I investigate how Mandarin learners, whose native language has only a binary laryngeal contrast, are able to acquire the Korean three-way laryngeal contrast in stops and affricates, focusing on their L2 language proficiency levels through two perceptual experiments. The results of experiments reveal that most Mandarin learners do not reach the same level of perception accuracy as native Korean speakers; however, advanced learners perceive the Koran three-way categories more accurately than beginning and intermediate learners. In addition, Mandarin listeners are more successful at perceiving Korean aspirated stop and affricate contrasts compared to fortis and lenis regardless of L2 proficiency. Moreover, lenis is least likely to be differentiated perceptually by Mandarin listeners across all proficiency groups, indicating that they lack attention to vocalic f0, which is the most relevant dimension for native Korean listeners to discriminate lenis from fortis and aspirated consonants.

Emilia Melara (sociolinguistics, pragmatics, morphosyntax)

This paper analyzes naturally-occurring uses of the propositional anaphor variants this, that, and it in the English of three Torontonian speaker groups, each of different ethnic and linguistic backgrounds (Anglo, Cantonese, and Italian), together and separately. The variation is analyzed quantitatively, using the variationist framework (Tagliamonte 2002), to examine the spontaneously-produced speech. It finds that on top of the pragmatic conditions proposed by Gundel et al. (1993), specifically that it is the least likely form to be used unless the referent proposition is the most salient one, there are additional constraints on the choice of anaphor used. The study finds that the primary predictor of anaphor choice is the type of clause that introduces the proposition, such that declarative clauses favour the demonstrative that, non-declaratives it. This is the case for all of the groups that were analyzed. A generation analysis of the Cantonese and Italian speakers shows that salience, as proposed by Gundel et al. (1993) to condition anaphor choice, is not employed by first generation speakers, unlike second generation speakers and the Anglo group as a whole. This suggests that if there are effects of language contact between the English of the speakers of Cantonese and Italian backgrounds, salience may not be a conditioning factor in Cantonese or Italian as it is in Toronto English. I explore what morphosyntactic properties of the anaphoric elements may be responsible for their preference with particular clause types and ask what it is about the pragmatic constraints that first generation speakers in the Cantonese and Italian groups do not pick up on in anaphorically referring to propositions in English.

Erin Hall (language acquisition, syntax/semantics)

My second generals paper was a language acquisition study looking at when children understand the difference between sentences with noun phrase embedding (e.g. The cup on the table is green) and coordination (The cup and the table are green). Previous studies have shown that young children have difficulty with the semantics of embedding in the CP domain, as well as with relative clause comprehension, but the PP type of subordination has not been examined in detail. I designed an iPad-based colouring experiment to test children’s comprehension of coordinated NPs as compared with two types of PP embedding, locative (in/on) and comitative (with). Children ages 3 to 5 were found to consistently understand the different interpretations of coordinates and the locative type of embedding, but they had difficulty with comitative PPs like The dog with the bone is blue; children often coloured both nouns (the dog and the bone) in these cases, treating them like coordinates rather than embedded sentences. These results suggest that children have an understanding of the semantic consequences of PP embedding within the NP as early as age 3, but their comprehension depends on the particular preposition involved. With PPs seem to be particularly challenging, likely due to the polysemy of this preposition and/or its closer semantic relationship to the conjunction and.

Ruth Maddeaux (psycholinguistics, sentence processing)

It has been argued that input from both linguistic and visual channels are combined to construct a conceptual representation of an object or event (Knoeferle & Crocker 2006, Altmann & Kamide 2007, Wolter et al. 2011). The relative salience of objects in the discourse is determined by both linguistic representations and visual perception (Wolter et al. 2011). An area of discourse that this research bears on is the processing of presuppositional information; specifically, is there any processing cost associated with adding information to the common ground through different modalities? I use an eye-tracker to determine where people look when they receive information that they are expected to accommodate, and measure whether the manner in which this information is acquired has an effect on the time it takes to process it. I focus on the presupposition trigger return, as in return the glass to the table. The results suggest that listeners have a weaker representation of an object when it is not mentioned by name. Looks to the target object are more likely in this condition, as a result of needing to strengthen the representation; that is, when it is least linguistically established in the common ground. I conclude that the manner of acquisition does have an effect on listeners’ discourse representation. Information that is received through linguistic means is more strongly established in the discourse than information received through non-linguistic means.

Patrick Murphy (semantics, psycholinguistics, sentence processing)

My second generals paper was an eye-tracking experiment investigating the Canadian English "be done NP" construction, which allows speakers of Canadian English (and a handful of U.S. dialects) to say "I'm done my homework" or "I'm finished my homework" (instead of "I've finished my homework" or "I'm done/finished *with* my homework"). A recent paper (Fruehwald & Myler 2015) argued that this construction involves aspectual adjectives directly taking an NP complement, without silent verbal or prepositional structure intervening. The paper also argued that these aspectual adjectives are similar to aspectual verbs (like "begin", "finish", etc.) in requiring a particular interpretation mechanism (complement coercion or type-shifting) for entity nouns (like "the resume" or "the coffee"—contrast with event nouns like "the interview" or "the party"). Previous studies (in self-paced reading and eye-tracking) have found evidence for complement coercion / type-shifting for aspectual verbs in the form of longer reading times for entity nouns than event nouns after these verbs. My own study had similar results: longer reading times for entity nouns than event nouns in the "be done NP" construction. This supports Fruehwald & Myler's claim that these aspectual adjectives behave similarly to aspectual verbs in requiring complement coercion / type-shifting for entity nouns.