Martha Hinrichs: The role of surprisal in L2 syntactic priming
(2014/2015 University Scholars Program Fellowship undergraduate study project)
In this project, Martha investigates the double object alternation in L1 Korean L2 English written production data. In Korean, all ditransitive verbs can be used in the prepositional object construction, while only some verbs such as cwu- (give) can also be used in the Korean double object construction (Jung & Miyagawa 2004). This raises the question whether advanced Korean learners of English would exhibit the same kind of constructional priming effects observed for other L2 English learners at advanced levels of proficiency, and if so, to what extent these prioming effects are modulated by verb-specific knowledge that is aligned with the constructional verb preferences of native speakers (Gries & Wulff 2005, 2009). This study uses the Young English Learners Corpus (YELC), a compilation of essays written by students in South Korea. Distinctive Collexeme Analysis (DCA; Gries & Stefanowitsch 2004) will be employed to to measure each verb’s associative bias towards either construction in the learner data. These statistics will then be compared to native English speakers’ preferences documented in previous research.
Rebecca Morris: L1 vs. L2 idiom processing: investigating the role of morpho-syntactic variation
(2014 undergraduate individual study project; completed)
In this project, Rebecca will test native and non-native speakers’ sensitivity to different variants of V NP idioms. Adopting a usage-based perspective, the hypothesis under investigation is that native speakers should be faster and more accurate in determining whether a given phrase constitutes an idiomatic or literal meaning depending on the surface form that the phrase is presented in. More specifically, native speakers are expected to identify idiomatic and literal senses faster when the phrase is presented in its most typical, i.e. frequent, surface form. A second hypothesis to be tested is that non-native speakers should exhibit the same qualitative behavior, yet less pronounced than native speakers. The underlying assumption is that since non-native speakers have had less input, and therefor weaker mental representations of what constitutes the most typical variant forms, they will be less able to rely on their knowledge of these more or less fixed assemblies as they make judgments and or give reaction times.
In order to test these two hypotheses, Rebecca will first of all identify the most frequent as well as less frequent variant forms of a set of 60 V NP idioms (which are available as a data sample for previous and ongoing research of Dr. Wulff) and use these as stimuli in a combined judgment and RT task.
Dylan Attal: Cognitive determinants of blend formation: an experimental approach
Dylan completed his USP Fellowship project and his honors thesis in April 2014. His honors thesis earned highest honors.
(2013/2014 University Scholars Program Fellowship and honors thesis project; completed)
In a television commercial broadcast at the 2013 Superbowl, the sandwich company Subway let its customers know that throughout the month of February, any sandwich would cost only 5 dollars. In order to make this promotion more memorable, they referred to it as Februany, a blend of February and any [sandwich]. Blending is an extremely popular word creation process, especially for advertisement campaigns and newspaper headlines – both genres in which space is limited and publishers compete for consumers’ attention. Blends fit the bill because they are compressed language, and they are catchy.
From a cognitive-linguistic research perspective, blends raise one major question: what factors impact the way in which a speaker blends two words together? For example, what makes brunch, a blend of breakfast and lunch, a better blend than breakfunch? Previous research on the basis of large collections of blends suggests that speakers take a variety of cognitive determinants into consideration in order to achieve the ideal balance between economy (the bigger the overlap of words, the better) and recognizability of the source words (the more material of both source words remains intact, the better). These cognitive determinants include various characteristics of the source words, such as their phonetic, phonemic, graphemic, segmental, and semantic similarity as well as their frequency in language. How exactly these characteristics interact in the online production and comprehension of blends, however, remains largely unclear to date. Gries (2012: 166) correspondingly points towards the dire need
to leave behind purely descriptive linguistic accounts and turn to psycholinguistic concepts, notions and methods instead... With regard to experimental approaches, it would be interesting to have speakers coin blends of source words while controlling for many of the factors known to influence blending.
In this research project, we aim to take a first step towards addressing this gap by conducting an experimental study in which native speakers of English are asked to blend to source words together. The source word stimuli will be systematically controlled for the different cognitive determinants mentioned above. The results will be statistically evaluated both monofactorially (means, interquartile ranges, and exact tests) and multifactorially by means of a linear model that identifies which factors contribute to an increasing distance of the chosen cut-off points to the ideal ones as determined by the predictors (Gries 2006). The findings of this study stand to make a valuable contribution to our understanding of subtractive word formation processes by providing us with first clues regarding an online production and comprehension model of blending and by informing our understanding of the differences between creative and conscious word formation processes such as intentional blending compared to involuntary and unconscious word formation processes such as speech errors.
José Molina: Constructional priming as a function of L2 proficiency and L1 background
José completed his honors thesis project in December 2013 and was awarded highest honors.
(2013 Honors thesis project; completed)
In this project, José elaborated on a previous study by Gries & Wulff (2005) that tested advanced German L2 English learners’ knowledge of verb argument structure constructions such as the ditransitive construction (José gave Steffi the paper) and the prepositional dative construction (José gave the paper to Steffi). José replicated two experiments, a syntactic priming experiment and a semantic sorting experiment. Rather than investigating only advanced learners of English from one L1 background only as in Gries & Wulff (2005), José elicited data from L2 learners at low-intermediate levels of proficiency, and from different L1 backgrounds. He found that the main controbutor to priming was the verb provided in the sentence fragment to be completed (as opposed to the verb presented in the prime sentence or the construction provided in the prime). Furthermore, he observed pronounced verb-specific effects such that certain verbs primed either construction significantly more often than others.