Apr 5, 2002:
The slides of Kathrin's talk (the 12th meeting) is posted. We're still looking for the next speaker so please stay tuned.

Mar 28, 2002:
The announcement of the 12th meeting is posted. Katharina Probst (LTI) will discuss her work on semi-automatic learning of syntactic transfer rules for machine translation.

Mar 15, 2002:
The slides of Brian's talk (the 4th special meeting) is posted. Please stay tuned for the next talk (tentatively by Kathrin, on learning transfer rules for machine translation).

Feb 28, 2002:
The announcement of the 4th special meeting is posted. The invited speaker is Prof. Brian MacWhinney (Psychology Department at CMU), on "Rethinking the Logical Problem of Language Acquisition".

Feb 22, 2002:
We had a wonderful talk by Christian. I need to say sorry to him because I mistakenly put him under HCII in my email announcement (stupid copy-n-paste error) - he is an LTI PhD student!
Please stay tuned for the next talk (tentatively by Brian MacWhinney from Psy@CMU).

Feb 15, 2002:
The announcement of the 11th meeting is posted. Christian Monson (LTI) will discuss unsupervised morphology learning.

Feb 14, 2002:
Happy Valentine's day - for Spring 2002 semester Ben ( will be the webmaster and main organizer, and Kenji ( will be the list maintainer and co-organizer. We expect to announce the first talk shortly before the end of this week so stay tuned.

Nov 13, 2001:
The announcement of the 10th meeting is posted. Sonya Allin (HCII) will discuss language and sensory-motor processes.

Nov 7, 2001:
We regret to announce that the 4th special meeting has been cancelled. We will have Prof. Brian MacWhinney as an invited speaker in December, or early next semester.

Oct 31, 2001:
The 4th special meeting, featuring Prof. Brian MacWhinney, has been rescheduled to Nov 8. The announcement has been updated.

Oct 24, 2001:
The announcement of the 4th special meeting is posted. The invited speaker is Prof. Brian MacWhinney (Psychology Department at CMU).

Aug 30, 2001:
So it's the new semester now! As you know LARG has been dormant for almost 2 months, but now we're back in business. For the new semester Kenji ( will be the webmaster and main organizer, and Ben ( will be the list maintainer and co-organizer. We're now looking for speakers, so if you're interested in giving one please contact either of us.

Also note that the LARG website is now moved to its permanent location: (we promise not to change it for at least a few years!) - please update your bookmark.

July 5, 2001:
The slides of Hal's talk (the 9th meeting) is posted.

June 25, 2001:
The announcement of the 9th meeting is posted. Hal will give a talk on "Language Learning in Optimality Theory".

May 22, 2001:
The slides of Erik's talk (the 8th meeting) is posted.

May 22, 2001:
The announcement of the 8th meeting is posted. Erik will give a talk on "Learning to Read a Non-alphabetic Script - Chinese".

May 10, 2001:
  1. The slides of Guy's talk (the 7th meeting) is posted.
  2. The Resource section is updated.

May 2, 2001:
  1. The LARG Anthology (in LTI library) is now updated up to Guy's talk (will take place on May 10).
  2. Erik's talk is now pushed back one week to May 24 (Thu) - the details of the talk will be announced soon.
  3. The first round of intro-level talks now officially came to an end. We're now seeking speakers for the summer meetings - anything related to LA is welcome! Please email to either Kenji or Ben, thanks.

Apr 26, 2001:
The slides of Michael's talk (the 6th meeting) is posted.

Apr 21, 2001:
The slides of Klaus' talk (the 5th meeting) is now added.

Apr 19, 2001:
The slides of Lori's talk (the 3rd special meeting) are posted. The slides are no longer available.

Apr 16, 2001:
Due to the final exam schedule the 7th meeting (given by Guy) is rescheduled to May 10.

Apr 12, 2001:
The announcement of the 3rd special meeting is posted. The invited speaker is Dr. Lori Holt from Psychology, CMU.

Apr 10, 2001:
Thanks to Guy's earlier-than-anyone abstract - now we have the announcement of the 7th meeting posted.

Apr 9, 2001:
Due to the speaker's schedule conflict the 6th meeting is postponed to Apr. 26, at NSH 4513. The topic, abstract and the readings remain unchanged. The readings will be available later today or tomorrow.

Apr 6, 2001:
  1. The slides of Janyce's talk (the 2nd special meeting) is posted.
  2. The announcement of the 6th meeting is posted. The papers will be available after Apr 9 (Mon) at the LTI front desk. Note for this talk the meeting place is changed to NSH 4632.
  3. The LARG Anthology in the LTI library has been updated up to the 2nd special meeting.

Mar 26, 2001:
  1. The 5th meeting is now rescheduled. This is also the last talk given by Klaus Ries before he leaves for Germany - so come while you still have the opportunity!
  2. The announcement of a special meeting is posted. This invited talk is given by Janyce Wiebe, and is a joint presentation with LTI Seminars.
  3. The details of the 6th meeting, the talk given by Michael Kohlhase on Kowledge Representation/inferences for LA (T4), will be posted soon.

Mar 19, 2001:
Due to the speaker's schedule conflict, the 5th talk given by Klaus Ries on March 22, Tursday (Discourse Learning Track) will be postponed to the next week. Details will be announced later this week.

Mar 12, 2001:
  1. The announcement of the 5th meeting is posted. This is also the last talk given by Klaus Ries before he leaves for Germany - so come while you still have the opportunity!
  2. A binder "The LARG Anthology vol. 1", which collects all of the papers read so far, has been placed in the LTI Library for reference. The binder will be updated periodically.

Mar 8, 2001:
  1. The slides of Ben's talk (the 4th meeting) is posted.
  2. The speaker of the next regular meeting (Mar. 22) is Klaus Ries (on discourse learning) - details will be posted soon.

Mar 7, 2001:
Great news: finally we have our mailing list up and running! Take a look here for details.

Feb 28, 2001:
  1. The announcement of the 4th meeting is posted.
  2. One interesting talk follows.

    Exemplar Model Accounts of Dissociations Between Categorization and Memory

    Dr. Robert Nosofsky
    Indiana University

    Monday March 5, 2001
    4-5:20 p.m.
    A51 Baker Hall
    Carnegie Mellon University

    According to exemplar models of classification, people represent categories by storing individual exemplars in memory, and classify objects based on their similarity to these stored exemplars. Beyond accounting for categorization, a goal of such models is to account for relations between categorization and other fundamental cognitive processes, such as old-new recognition memory. Achieving such a goal is important because it suggests the possibility that a single representional system may underlie diverse cognitive tasks.

    Recently, however, such single-system exemplar-model accounts have been challenged on grounds of certain "dissociations" between categorization and memory. Researchers have demonstrated statistical dissociations involving low correlations between categorization and recognition, in which individual items that are not well recognized are nevertheless well categorized, and vice versa. They have also demonstrated experimental dissociations in which amnesic patients with exceedingly poor recognition memory nevertheless learn to categorize objects as well as do normals. Such dissociations have motivated the development of alternative multiple-system models, in which separate representational systems are assumed to mediate categorization and recognition judgments.

    In this talk I argue that such dissociations are compatible with the predictions from single-system exemplar-memory models and that such models provide viable accounts of the phenomena of interest. I also report on some preliminary attempts to distinguish the single-system exemplar-model view from the alternative multiple-system accounts.

Feb 22, 2001:
  1. The slides of Rachel's talk (the 3rd meeting) is posted.
  2. The speaker of the next regular meeting (Mar. 8) is Benjamin Han (on grammar induction) - details will be posted soon.
  3. Good news: the mailing list saga is going to its conclusion: we'll have a list running in the next week!
  4. Information about two interesting talks follows:

    Deborah Chen Pichler (U. of Connecticut), 15:15, Feb 28 (Wed), Room 144 of the Cathedral of Learning (The English Nationality Room): Word Order in Early ASL

    Children acquiring languages with variable word order must not only set the word order parameters correctly, yielding the basic or underlying order, but also learn the syntactic operations that will allow them to derive variant word orders. This study examines the word order patterns produced between the ages of 20-30 months by 4 deaf children acquiring American Sign Language (ASL) as a first language from deaf, signing parents. Although the basic word order of ASL is SVO, verb+object combinations produced by these children occur in both VO and OV word orders. Examination of the noncanonical OV sequences reveals use of verbal morphology that licenses word order variation in the adult language. On the assumption that noncanonical surface orders are derived by syntactic operations applied to the basic order, I propose that the adult-like variation in the child data is evidence that the children have correctly set the head-complement parameter by the age under investigation. In addition to using OV-licensing verbal morphology, one child also produces what appear to be topicalization structures. Acquisition of topicalization is generally thought to be acquired late in ASL (Reilly et al., 1991), due to the absence of an adult-like topic nonmanual (raised brows, head tilt and pause after the topic) in early production. I argue that the standard description of the adult topic nonmanual is too restrictive, and that children mark their early topics with a more broadly defined nonmanual in the form of a general prosodic break between the topic and the rest of the sentence.
    Jeffrey Mark Siskind (NEC Research), 15:30, Feb 27 (Tue), NSH 1305: Grounding the Lexical Semantics of Verbs in Visual Perception using Force Dynamics and Event Logic

    In this talk, I will present an implemented system, called Leonard, that classifies simple spatial motion events, such as `pick up' and `put down', from video input. Unlike previous systems that classify events based on their motion profile, Leonard uses changes in the state of force-dynamic relations, such as support, contact, and attachment, to distinguish between event types. Since force-dynamic relations are not visible, Leonard must construct interpretations of its visual input that are consistent with a physical theory of the world. Leonard models the physics of the world via kinematic stability analysis and performs model reconstruction via prioritized circumscription over this analysis. In this talk, I will present an overview of the entire system, along with the details of both the model reconstruction process and the subsequent event-logic inference algorithm that can infer occurrences of compound events from occurrences of primitive events. This inference algorithm uses a novel representation, called spanning intervals, to give a concise representation of the large interval sets that occur when representing liquid and semi-liquid events. I will illustrate how Leonard handles a variety of complex visual-input scenarios that cannot be handled by approaches that are based on motion profile, including extraneous object in the field of view, sequential and simultaneous event occurrences, and non-occurrence of events. I will also present a live example illustrating the end-to-end performance of Leonard classifying an event from video input.

Feb 17, 2001:
  1. Finally the Resource section is updated (thank you Natasha and Rachel). Please contact the webmaster to add new links.
  2. Paul Bloom's new book: How Children Learn the Meaning of Words, MIT Press, 2000 (thank you Rachel). Here is a review.

Feb 15, 2001:
  1. The announcement of the 3rd meeting is posted.
  2. The two invited talks are now settled: the first one (Apr 6, Friday) will be given by Dr. Janyce Wiebe from CS, UPitt on subjective/objective classification, and the second one (Apr 19, Thursday) will be given by Dr. Lori Holt from Psych, CMU. For Dr. Wiebe's talk we've also made it an LTI seminar so the entire LTI will be invited to the talk. Stay tuned for more details.

Feb 1, 2001:
  1. The slides of Eric's talk (the 2nd meeting) is posted.
  2. The speaker of the next regular meeting (Feb. 22) is Rachel Chung (on lexical acquisition) - details will be posted soon.

Jan 26, 2001:
  1. The announcement of the 2nd meeting is posted. Note that Rachel's talk on Lexical Acquisition is postponed to the 3rd meeting (Feb. 22, see below) due to scehdule conflicts.
  2. Due to the schedule conflict with LTI IR recitation, which takes place every other Thursday at roughly the same time as LARG, we need to skip one week to 'synchronize' with it. Therefore there is no meeting on Feb. 15, i.e., the 3rd meeting will be on Feb 22, and we'll resume the regular meeting time afterwards, namely every other Thursday at 1630-1830.

Jan 22, 2001:
  1. The announcement of the special talk given by Allen Gorin is posted.
  2. The speaker of the next regular meeting (Feb. 1) is Rachel Chung (on lexical acquisition) - details will be posted soon.

Jan 18, 2001:
  1. The slides of the 1st meeting are available now in the archive section.
  2. Now it's official: the meeting time for LARG is every other Thursday at 1630-1830, and the place is at Newell-Simon Hall (NSH) 4513 (direction).
  3. Thanks to Eric Nyberg for pointing out the gapping hole in the coverage - "Linguistic perspective" is now added and the Definition of LA section is revised.
  4. We're now seeking speakers for the intro-level talks on three tracks: Lexical Acquisition, Cognitive Science perspective and Linguistic perspective. Volunteers are welcome!
  5. Please suggest useful web resources - we need to expand our Resources section. Also we're constantly seeking more reading material. So fire away!
  6. The mailing list is coming soon.

Jan 14, 2001:
The announcement of the 1st meeting is posted.

Dec 21, 2000:
A questionnaire has been sent out to all prospective participants. The questionnaire is posted below in case you have not received it - you can send the completed questionnaire to us. Please do so as soon as possible since we're relying on it to schedule the first official meeting in Spring 2001.


  1. List the *bad* time slots for you to meet bi-weekly in Spring 2001.

  2. What are your research interests/specialization?

  3. The subject areas of LARG include:
    (a) Lexical Acquisition
    (b) Grammar Induction
    (c) Discourse/pragmatics Learning
    (d) Knowledge Representation/Inferences for LA
    (e) Cognitive Science Perspective

    Please choose the areas you are already familiar with enough to give an intro-level presentation. This is to schedule for the 1st round of the presentations in order to get all of the participants up to speed. Later we will let you know who chose the same subject area as you - further coordination might be necessary to make a joint presentation, for example.

  4. List anyone you know who might be interested in participating LARG.

  5. List anyone you know who might give an invited talk at LARG (we will tentatively have two invited talks: in March and April).

  6. List any online resource that you think is relevant/useful to LARG.

  7. List any reading material (together with the corresponding subject area) that you think is good for reading/discussions.

  8. Take a look at the member list and verify that the info about you is correct - otherwise please give the corrections below.

  9. Any other suggestions? Write them down here.

Dec 21, 2000:
The Pre-meeting is past. A summary of the meeting is here. Questionnaires will be sent to the prospective participants later.

Dec 20, 2000:
Now LARG is organized jointly by Kenji Sagae and Benjamin Han. Feel free to send us any comments you have.

version 0.2 ; htmlinst date: Fri Nov 7 16:49:04 2003

Webmaster: Benjamin Han (benhdj at cs dot cmu dot edu)