Erin's Psycholinguistics Page
University of Arizona, Tucson

Curriculum Vitae
Event Structure in Comprehension
Coyote Papers
Ling 300 Intro to Syntax
My Dissertation
CUNY 2005 Conference
Teaching
Event Structure Reading Group
Other stuff I like

Psycholinguistics
My doctoral research in linguistics focused on the area of psycholinguistics, specifically language comprehension. Broadly, the goal is to find out what's happening in the brain when humans understand language. What are the roles of syntax (structure), semantics (meaning), prosody (intonation, stress, & rhythm), and real world knowledge in understanding a spoken sentence? How much of our linguistic knowledge is stored with words in the mental dictionary (lexicon), & how much of our knowledge results from general processes for identifying structure & meaning?

My Dissertation Committee: Thomas G. Bever, Heidi Harley, Merrill Garrett, Ken Forster, Audrey Holland

Recent Collaborators: Thomas G. Bever, Heidi Harley, Raffaella Folli, Connie Clarke, Ken McRae, Dave Townsend, Selene Gardner, Matt Jarboe, Janet Nicol, Katrina Treadwell, Christopher Nicholas, Ben Jones, Roeland Hancock, Amanda Wixted, Janice Scott, Mike Metz

Not so recent collaborators: Shari R. Speer, Amy J. Schafer, Steve Perry

Teaching Experience (follow links for on-line teaching materials)

My research
The role of event structure in language comprehension
Some verbs, such as "arrive" and "notice", include the completion of the event in their meaning, while other verbs, such as "dance" and "run" do not. During the time that the sentence, "The woman is dancing", is true, it is also true that "The woman has danced." In contrast, during the time that "The woman is writing a book" is true, it is not true that "The woman has written a book". Rather, it will be true only after she has completed the writing of the book. Thus, completion is a critical part of the meaning of "writing a book". My research investigates whether the information about completion or non-completion in a verb or verb phrase's meaning is used immediately during the process of understanding language. Research has shown that within the first 500 milliseconds after hearing or reading a sequence such as "the woman noticed", the language processing system begins determining whether "the woman" is the "noticer" or the "noticee". My experiments show that the information about an event's completion or non-completion is one of the first pieces of information used to determine the role of the other words in the sentence in relation to the verb.

This project began as my doctoral thesis. It was supported by an NSF Doctoral Dissertation Improvement Grant. This project now has a web site of its own, where you can download materials and data. Click here to go to the Event Structure in Language Comprehension page. Also, I invite you to visit my dissertation page.

Minimalist derivation in comprehension project
The purpose of this project is to present an account of how Minimalist derivation (Chomsky 1995) can be embedded in a comprehension model, the Late Assignment of Syntax Theory (LAST) (Townsend & Bever, 2001). I am concerned with issues regarding the interface between the first step of the model, in which heuristic strategies apply to the utterance, and the second step, Minimalist derivation. Thus far, I have focussed on two questions about the interface: 1) How are features in the numeration needed to begin a Minimalist derivation chosen? 2) What dictates which units Merge in the derivation? Chomsky (1995:226-227) claims that we do not need to ask either question. In my paper in Coyote Papers volume 12, I review his reasons and argue that we can and should answer these questions in a workable comprehension model. In response to the first question, I demonstrate that heuristic strategies applied to the utterance determine which features enter the numeration. In response to the second question, I discuss how heuristic strategies combined with lexical information determine which items Merge. Thus far, I have considered the processing of English reduced relative clauses sentences and Spanish sentences with unaccusative verbs. Future research will involve testing predictions made by the model for other structures and other languages. This paper is published in an electronic journal (Coyote Papers 12: Language in Cognitive Science). Click here to go to the e-journal from which you can download the paper.

Auditory reduced relatives & filled gaps project (in collaboration with Janet Nicol, Dave Townsend, & Tom Bever)
The purpose of the auditory reduced relatives and filled gaps project was to test for three classic sentence processing effects, the reduced relative effect, the effect of animacy on reduced relatives, and the filled gap effect, using spoken sentence stimuli in the voice change monitoring task (Townsend & Bever 1991) as opposed to a reading task. The results differed from similar experiments that used text stimuli in self paced reading and eye tracking (Trueswell, Tanenhaus, & Garnsey 1994, Stowe 1986, Crain & Fodor 1985). The reduced relative effect was found in the on-line measure but not in the off-line measure. The filled gap effect was found off-line but not on-line. There was no significant effect of animacy on the reduced relative effect. These results were presented in a poster at the CUNY Human Sentence Processing conference held by UC San Diego in April 2000. A paper on this research is being prepared for submission to a psychological journal. Future research will investigate whether rate or control of presentation accounts for the different findings.

Sentence comprehension of DECtalk project
The purpose of the DECtalk project was to see whether similar sentence processing results are found using sentence stimuli produced by the speech synthesizer DECtalk compared to stimuli produced by a human speaker. Twenty-two English monolinguals participated in the experiment. Subjects made speeded grammaticality judgments to two kinds of auditory reduced relative clause sentences. Type 1 sentences had animate initial NPs and by-phrases. Type 2 sentences had inanimate initial NPs and prepositional phrases. Each subject heard one of two counterbalanced presentation lists. Each list included both natural speech and synthetic (DECtalk) sentences in a mixed order. The results showed a main effect of speech type: subjects made more errors to DECtalk sentences than natural sentences overall. There was no main effect of sentence type. There was a significant interaction: subjects made significantly more errors to type 2 sentences than type 1 sentences in the natural speech condition only. Grammaticality judgment errors for the two types of sentences were not different in the DECtalk condition. Two conclusions were drawn: 1) some processing effects found using natural stimuli may not be found using DECtalk stimuli 2) the significant difference between the two sentence types in the natural speech condition shows that reduced relative clause sentences with by-phrases are easier to process than those with other prepositional phrases. The results were presented at the Student Conference in Linguistics held at the University of Arizona in February 2000. The proceedings are published in MITWPL. Click here to download the paper as a pdf.

Dialects and sound change
Sound change refers to significant changes in the phonological features used in spoken language by a defined population of speakers over a period of time. An example of a major sound change would be the Great Vowel Shift in English. During the past twenty-five years, there has been a lot of research which supports the theory that we can detect sound change in progress by analyzing the speech of sample groups of speakers of different age groups in a given region.

In 1995, an Undergraduate Research Award (UGRA) from the University of Kansas Honors Program enabled me to conduct some sound change research of my own with advice and guidance from Professor Jim Hartman of the KU English Department. I interviewed speakers over the age of fifty and under the age of twenty-two in Wichita, Kansas, for the purpose of studying a feature which may represent a sound change in progress, the laxing or lowering of tense vowels before /l/. This feature has also been documented in some other regions of the United States.

If it becomes feasible, I would love to do a follow-up study to see if the sound change has progressed. If this project interests you, please feel free to write to me at the e-mail address at the bottom of this page.

Tech-y stuff: Experimental Hardware and Software Links

Some Linguistics links

Professional Organizations Psycholinguistics and Cognitive Science Journals CUNY & AMLaP websites for Tom Bever's sentence processing seminar, Fall 2004 Linguistics Jobs in industry Some online dictionaries, encyclopedias, lexicons Money for Grad Students Statistics References Link pile

Please note that this page is out of date and will probably not be fully updated any time soon.

Click here to go back to my main page.


Last modified 8/10/2007