LINGUIST List 4.296

Fri 23 Apr 1993

Qs: Stemming, Taps, Frequencies, Klingon

Editor for this issue: <>

Directory

1. Tom Donaldson, Word "stemming" methods request
2. Joe Stemberger, query: taps
3. , Part of speech Frequencies
4. Yuphaphann Hoonchamlong, klingon

A computational linguist suggested that someone on this list might be
able to point me in the right direction.

A common feature for full text retrieval engines is the ability to do
"stemmed" searches. For example, a stemmed search on "computer" might
be functionally transformed into a search for: "compute", "computers",
"computing", and so on.

I am told that "word stemming" falls under the liguistic rubric of
"morphological analysis". I am looking for algorithms or existing
software to perform this small subarea of "morphological analysis" for
a variety of languages, but am especially interested in German and
Japanese. Can someone direct me to software vendors or other sources
of software, or to relevant literature?

Thanks,

While reading the most recent book by James Herriot, in which he
frequently uses non-standard spellings to give the flavor of Yorkshire
speech, I came across this:

 spelling: gerrim for: get him

I assume that the {rr} spelling here represents the tap that British
dialects have for intervocalic /r/. But it's not for /r/ here, but for
an alveolar stop.

Now, I've always thought that taps for /t/ and /d/ was a North American
phenomenon, but here it is for a British dialect. Are there dialects on
THAT side of the Atlantic that have taps for /t/ and /d/, then? And is
there any historical connection with the North American English tap (such
as: North America got a lot of immigrants from Region X in England, where
taps are used for /t/ and /d/)? Do any of you British folk know?

This makes me wonder about the Jethro Tull song, SKATING AWAY ON THE THIN
ICE OF A NEW DAY. Ian Anderson sings with a blatant British accent, yet
taps the /t/ in SKATING. I always thought he was intentionally doing a
North-Americanism, but it's bothered me.

---joe stemberger

Does anybody out there have any hard data on the following?

 1) frequencies (types and tokens) of parts of speech (specifically nouns
 and verbs) in adult-to-child speech in English or other languages;
 2) relative frequencies with which nouns and verbs appear at the BEGINNINGS,
 MIDDLES, and ENDS of sentences in adult-to-child speech or even in
 casual adult-to-adult speech in English and any other languages?
 3) the above categorized by utterance types?

I am looking for something with which I can compare my own findings
for adult-to-child speech in Mandarin Chinese. Specifically, I found that
8 out of 10 of my Chinese-speaking toddlers (22-26 months, each less than
200-word vocabularies from 1-hr samples) use more verb types than noun
types and 9 out of 10 of these children used more verb tokens than noun
tokens. Interestingly, the caregivers in all 10 families ALSO used more
verb types and tokens than noun types and tokens AND were more likely to
place transitive verbs in sentence-initial position than any other word
class (about 20% of the time for multi-word utterances, compared to just 5%
for nouns). Given that the children's propensity for early verbs is VERY
UNLIKE English and acquisition data from other languages (except perhaps
Korean, cf. Gopnik & Choi), I am very interested to know what the frequency
of adult verbs vs. nouns (types, tokens, and position X frequency) is for
the other languages that we have acquisition data on.

Any and all responses will be much appreciated and I will happily post a
summary if there is enough interest!

Twila Tardif
Department of Psychology
Yale University

I read in Time (Asia Edition) April 9th, in an article about Klingon language,
that there's Klingon info available somewhere via internet. Does anyone know
the internet address for that?
Please reply to: yuiipied.tu.ac.th
Thanks.
Yuphaphann.