The weekly newsletter of The Department of Linguistics, The University of Massachusetts, Amherst

WHISC
What's Happening In South College

March 25, 2004
Issue 2:13

Archived at http://www.umass.edu/linguist/about/whisc/

OVERVIEW

Brian Joseph Colloquium
Speech and Prosody 2004
Student work
Report from Prague
A new whimsy from Hubert Haider
Solutions to last week's whimsies
A postcard from Francesca Foppolo

In The Guardian:

Noam Chomsky: MIT professor, writer and activist

"in this system of immense power, small differences can translate into large outcomes"

Also: Chomsky now has a weblog --- or, perhaps more accurately --- someone is posting Chomsky's comments on the Net for him:

Turning the Tide

All the entries read as though someone started recording Chomsky mid-lecture. For a few days, one could post comments. Not surprisingly, that did not last.

[via Brian Weatherson's Thoughts Arguments and Rants]


Reminder

The first meeting of SPRIG will be this Friday, 9 AM, in the Phonetics Lab, Bartlett 6 (in the basement of Bartlett if you've never been there). Download the newest version of PRAAT beforehand.


Help us reach #1

At present, WHISC is merely Google hit #3 for WHISC. Ahead of us are
  1. Western Hemisphere Institute for Security Cooperation
  2. Women's Health Information and Support Centre
If you have a website, you can help. Just link to WHISC using WHISC as the link name. Like this:

<a href="./"</a>.

Though WHISC has a tiny readership, we feel that this should not stop us from being #1. Anyway, the Western Hemisphere Institute for Security Cooperation seems to prefer WHINSEC.


Post festi

Robert Stalnaker spoke earlier this evening at Amherst College:

What is de re belief?

* * * * *

The Department of Communication Disorders cordially invited you to attend a presentation on Thursday, March 25th, 4-5 p.m. in 321 Arnold House:

Dr. LaVae Hoffman, Research Scientist

University of Texas at Austin

'Clinical Implications in Specific Language Impairment: Bridging Theoretical Investigations into Clinical Intervention Research'

* * * * *

the semantics reading group has met again, at uri's house, to continue the discussion of rullman and you.

* * * * *

[If fewer interesting people with stimulating questions had stopped by the WHISC editorial offices today, we would have gotten this issue out in time for these notices to be of use to you.]


BRIAN JOSEPH COLLOQUIUM

The GLSA welcomes

Brian Joseph
Ohio State University

Optimality, Optimization, and Analogy Reconsidered:
Back to Basics (and Beyond)

Friday, March 26, 3:30 pm, in Machmer W-26


SPEECH AND PROSODY 2004 REPORT

by Shigeto Kawahara
Lisa, Taka, and I presented one oral presentation and two posters at the Speech and Prosody conference in Nara, Japan, March 23-26, 2004. Lisa did an astounding job of delivering the talk in a very limited time (12 minute presentation, 3 minute question period). The oral presentation was particularly well received, inviting many interesting questions from linguists as well as engineers. The posters attracted the attention of many people as well, and the discussion with people from different backgrounds (such as JTobi people from Ohio State and engineers who are mainly interested in speech synthesis) was particularly fruitiful.

Mariko Sugahara, who graduated from UMass last year, also presented a paper. Tateishi Koichi (who graduated around 1990) was also there and engaged in discussion with us on a variety of topics. The picture of mini-Umass-reunion should be sent to you from Taka shortly.

A mini-Umass-reunion

STUDENT WORK

Luis Alonso-Ovalle has posted his forthcoming NELS paper:

Luis Alonso-Ovalle. 2004. Simplification of Disjunctive Antecedents. In Keir Moulton and Matthew Wolf, eds., 2004), Proceedings of the North East Linguistic Society 34. University of Massachusetts, Amherst: GLSA.

BARBARA PARTEE FROM PRAGUE

Dear WHISCERS,

Volodja and I got back Saturday evening from a week in Prague at the Mathesius Lecture Series (probably the last one ever), where we gave four lectures and attended lectures by Manfred Krifka and Stuart Shieber, and spent time with a number of our Prague friends and with 5 Georgian friends and colleagues who were attending Mathesius.

I especially enjoyed Manfred's lectures. Besides his 3-lecture series, he also delivered the annual Roman Jakobson lecture for the Prague Linguistics Circle, which all Mathesius participants were able to attend. (And then he was at a conference with Angelika in Leipzig at the end of last week.)

Continue reading Barbara's report from Prague


WHISC WHIMSY #3 (FROM HUBERT HAIDER)

GEN and EVAL, but be very, very quick! ---
Especially if you are a syntax processing human

Background:

The Gen and Eval mechanisms are central for OT. Gen is a pattern generator. It creates the set of possible structural realizations of an input --- the candidate outputs. For our purposes, candidate outputs can be strings. Eval evaluates the entire set to determine which candidate output is optimal with respect to a finite set of constraints.

Here is your task:

Someone (not necessarily B.G.) offers you US$1 billion if you fully and successfully apply GEN+EVAL to the following string of words, judging (as human evaluator) for each candidate whether it is a grammatical string:
We the people of the United States, in order to form a more perfect union, establish justice, insure domestic tranquility, provide for the common defense, promote the general welfare, and secure the blessings of liberty to ourselves and our posterity, do ordain and establish this constitution for the United States of America.

Assume that the candidate set is given by the set of all permutation variants of the words in the quotation. This simplification (to a finite candidate set, with no symbols added or taken away) is not inherent in OT, but it is a live possibility that the set of potential winners is finite. With just this restriction in place, is it reasonable to accept the deal, given that you will get the money only after delivering?

Here are some clues:

  1. Calculate the candidates. How many possible candidates are there?
  2. Approximate your time calculation, allowing yourself the following time scheme for your future tough, but well-paid job: 1 second per eval operation; 14 hours per day, 7 days a week.
  3. Guess: When will you be finished and get your money?
    a.within 10 years
    b.before your retirement age
    c.at the age when Jeanne Calment died (121 years)
    d.before the end of this millenium (i.e. Dec. 31st 3000)
    e.later, but still in the life time of our solar system
    f.later

SOLUTIONS TO LAST WEEK'S WHIMSIES

Trafficking in verkehrts

The task of WHISC Whimsy #1 was to insert one word per box such that the example becomes a grammatical German sentence (inserting commas to taste):

         
         
         
verkehrt verkehrt verkehrt verkehrt verkehrt verkehrt

Jan Anderssen was the first and only WHISC reader to supply a correct answer:

Werweiss,werverkehrtverkehrtverkehrtverkehrt,verkehrtverkehrt.
Whoknowswhotrafficswronglytrafficswronglytrafficswrongly.
'The one who knows that whoever traffics wrongly traffics wrongly is one who traffics wrongly himself.'
Hubert Haider's solution:
Werwo,manverkehrtverkehrtverkehrtverkehrt,verkehrtverkehrt.
Whowhereonetrafficswronglytrafficswronglytrafficswrongly.
cf. 'Wer dort, wo man andersherum verkehrt, andersherum verkehrt, der verkehrt verkehrt.'

* * * * *

Birthday, PIE

In brief, WHISC Whimsy #2 told the story of a professor who valiantly displayed 19 CVC roots of proto-Finno-Ugric that he claimed to be derived from proto Indo-European (PIE). The question: Can we be sure that these are not accidental matches?

Hubert's answer begins from the solution to the Birthday Paradox: How many people must be in a room for the chances of two of them having the same birthday to rise above 50%? The answer is 23. Hubert then reasoned that 19 CVC roots out of a basic vocabulary of about 3000 places the chance level at 190 correspondences.

Hubert continues

A week later, the professor for Indo-European Studies called me into his office. Look, he said to me, 35 Students are enrolled in my class, and there are no two with the same birthday. (He was not impressed by me pointing out that I had said: chance greater than 50%)

A POSTCARD FROM FRANCESCA FOPPOLO

Postcard from Francesca

Click the postcard to see its other side.


To The WHISC archive      To The UMass Department of Linguistics      To UMass Amherst