Computational Psycholinguistics (2015 LSA Summer Institute 2-week course)

1 Course information

Lecture Dates July 7, 10, 14, and 17 (Tuesdays and Fridays)
Lecture Times 3:10pm-5:00pm
Lecture Location Albert Pick Hall for International Studies, Room 16
Office Hours Immediately after class in Pick 16
Class webpage

2 Instructor information

3 Course Description

Over the last two and a half decades, computational linguistics has been revolutionized as a result of three closely related developments: increases in computing power, the advent of large linguistic datasets, and a paradigm shift toward probabilistic modeling. At the same time, similar theoretical developments in cognitive science have led to a view major aspects of human cognition as instances of rational statistical inference. These developments have set the stage for renewed interest in computational approaches to human language use. Correspondingly, this course covers some of the most exciting developments in computational psycholinguistics over the past decade. The course spans human language comprehension, production, and acquisition, and covers key phenomena from both phonetics and syntax. Students will learn key technical tools including probabilistic models, formal grammars, and decision theory, and how theory, computational modeling, and data can be combined to advance our fundamental understanding of human language use.

4 Course organization

Each day of the 4-day course will be primarily lectures with some boardwork, but with ample opportunity for discussion. Don't hesitate to ask questions during the lectures!

Every day of the course has a core reading – a short conference paper describing one of the models receiving primary focus for that day. We encourage you to read the core reading before class each day, as it will prepare you better to absorb the material in the day's class. We've also included a list of other readings to give you a broader range of pointers to key related literature for each day's lecture.

5 Intended Audience

Graduate students and researchers in linguistics, cognitive science, logic, psychology, computer science, and any other discipline who are interested in using computational modeling techniques, especially probabilistic modeling, to study human language processing and acquisition.

6 Start-of-class stuff

Please take a few moments outside of class to complete our Introductory Survey, which will give us a bit of information about the backgrounds and goals of participants in the class.

There is also a mailing list for the course that we will use to communicate with course participants, and that course participants can use to communicate en masse with each other. If you were enrolled in the class as of July 5, you should already be subscribed to the mailing list. Otherwise, please subscribe!

7 Syllabus (subject to modification)

Day Topic Slides Homework Core reading Other readings materials  
Tue 7 July Introductory probability theory; Bayes nets; probabilistic models of human speech perception; the perceptual magnet Lecture 1 (no builds) HW1 (solutions) Feldman & Griffiths, 2007 Feldman et al., 2009a; Clayards et al., 2008; Sonderegger & Yu, 2010  
Fri 10 July Probabilistic grammars, incremental syntactic comprehension, garden-pathing, surprisal Lecture 2 (no builds) HW2 (solutions) Hale, 2001 Jurafsky, 1996; Levy, 2008a; Narayanan & Jurafsky, 1998; Narayanan & Jurafsky, 2002; Demberg & Keller, 2008; Smith & Levy, 2013  
Tue 14 July Optimality in sentence production; phonetic reduction; Uniform Information Density; Noisy-channel models I Lecture 3 (no builds) HW3 (solutions) Levy & Jaeger, 2007; Levy, 2008 Jurafsky et al., 2001; Bell et al., 2009; Genzel & Charniak, 2002; Genzel & Charniak, 2003; Keller, 2004; Jaeger, 2010; Piantadosi et al., 2011; Seyfarth, 2014; Norris, 2006; Norris & McQueen, 2008; Levy, 2011; Gibson et al., 2013;  
Fri 17 July Noisy-channel models II; models of language acquisition Lecture 4 (no builds)   Bicknell & Levy, 2010; Pajak, Bicknell, & Levy, 2013 Lewis et al., 2013; Feldman et al., 2009b; Perfors et al., 2011  

8 Participation requirements (for enrolled students)

There are three requirements for students enrolled in the course to receive credit (regardless of whether you're taking the course for a letter grade or for pass/fail):

  1. You must come to class, and participate (pay attention, and ask questions when you've got them!)
  2. We will give you three short written homework assignments, one on each of the first three days of class. Each one is due half an hour before the beginning of the next class. The assignments will be designed to consolidate understanding of material we cover in class. You must complete two out of these three assignments.
  3. You must write up a short (2-page) final research project proposal on some topic in computational psycholinguistics that connects to your own research or to your intellectual interests. This proposal is an opportunity for you to get feedback from us on your ideas. These writeups can be completed individually or in small groups (no more than 3 people in a group). This final project proposal is due at 11:59pm on Sunday July 19.

Please turn in your homework either (i) on paper in class, or (ii) electronically, by emailing it to Final research project proposals must be turned in via method (ii), in PDF format.

Author: Klinton Bicknell and Roger Levy

Created: 2015-08-03 Mon 12:35

Emacs 24.4.1 (Org mode 8.2.5h)