LINGUIST List 2.36

Friday, 08 Feb 1991

Disc: Words and Emotions

Editor for this issue: <>

Directory

1. Jean-Marc Fellous, Re: Replies

I would like to thank all the persons who fhad the kindness to answer my 
query for informations on the relations between Words and Emotions.

Following is the summary of the responses I have got. All other comments 
and references are the most welcome and will be sent to this mailing list 
periodicaly .
Please, feel free to contact me for further informations,

Yours,

Jean-Marc
fellousrana.usc.edu

>
In response to J-M Fellous' inquiry about words and emotions. Check the journal
Cognition and Emotion (or is it Emotion and Cognition?). There was a recent
issue devoted in great part to the language of emotion. Work by Lois Bloom and
Johnson-Laird is also relevant. Raoul SMith, in computer science at
Northeastern, has done some work on modeling in this area. You might contact
him directly.

Bill Frawley


>
Hi,
I saw your query on LINGUIST. I'm a linguist writing my dissertation
proposal on the display of affect in oral life histories. So I am
also interested in w emotions are encoded in language. If I were you
I'd start with the 1989 special issue on the Pragmatics
of Affect in the journal TEXT. The people to look for are Elinor Ochs
and Bambi Schieffelin, and Biber and Finegan. They have done much of
the work on this topic. In fact I think Ochs may be at UCLA or USC.
 If you want I can send you my reference list on Language
and Affect. 
 Good luck.
--pam saunders

>
I can only offer one pointer, and that is to Ortony, who's over at ils
(Institute for Learning Studies) in Chicago. 

I'm mainly writing because I'm interested in emotion - I practice
and teach (Re-evaluation) Co-counseling. I'm currently working on my
PhD in computational linguistics (though I'll be working more on
a manageable KR for Natural Language(s)) . If you have any ideas you want to
throw my way to see if they bounce, feel free. I'd be glad to help.

Elissa Feit

>
On your query to the linguist bulletin board about modeling emotions:
I am no expert in that field, but I can refer you to an entire chapter,
precisely on the semantic analysis of emotion words, in the following book:

Wierzbicka, Anna. 1972. Semantic Primitives.

Wierzbicka's subsequent books -- Lingua Mentalis and others -- may also
have something on it; I don't know.

Lee Hartman
Southern Illinois University at Carbondale

>
Dear Jean-Marc,

 William Labov, a socio-linguist, has done some work on the lexical
semantics of the cup/mug distinction which has drawn the interest of lexical
semanticists. Because the distinction is ill-defined and seems dependent on
perception of quite a range of physical and functional features, Katy decided
to attempt to model cup/mug judgments of adult native speakers of American
English. It's not the cup/mug problem itself that makes her work of potential
interest to yours but rather the attempt to model those judgments using
simulated neural networks. The nature of the problem and of the attempted
modeling has similarities to attempts to model emotional information.

 I may be reaching a bit on the relatedness of her work and yours, but
it strikes me that some of the things she's talked about would have bearing on
your interests. Structuring mental and linguistic constructs with serial
processing models hasn't tended to work too well, but some of the parallel
distributed processing experiments have looked promising. I hope this doesn't
seem too far-fetched.

Herbert Stahlke
 University Computing Services
 Ball State University

>
Dear Jean-Marc Fellous,
try Dick Janney. I don't know if he would approve
of your project, but he definietly knows something about emotions and the
relevant linguistic literature. There is also Dick Watts in Berne, Switzer-
land, whose e-mail address I have lost. Dick (Janney) may have it.
Regards,
Hartmut Haberland

>
Il y a des strategies et methodes qui pourraient vous etre utiles dans
_Literary Computing and Literary Criticism. Theoretical
and Practical Essays on Theme and Rhetoric_ (Rosanne G. Potter, ed.,
U. Penn, 1989). Il y a egalement un chercheur que j'ai rencontre
a deux congres des ACH-ALLC. C'est le psychologue Suitbert ERTEL de
l'universite de Goettingen qui travaille beaucoup dans ce domaine
des emotions et des textes.

Cordialement,
 Joel Goldfield
 Fellow in Foreign Languages
 Institute for Academic Technology
 U. of North Carolina - Chapel Hill
( Translation by Jean-Marc:
There are methods and strategies that might be useful to you in "Literary Computing and Literary Criticism" and "Theoretical
and Practical Essays on Theme and Rhetoric_ (Rosanne G. Potter, ed.,
U. Penn, 1989)". There is also a researcher I met during the ACH-ALLC conferences; the psychologist Suitbert ERTEL from the Goettingen university, who is very involved in the domain of texts and emotions.)

>

You may be interested in our work with the artificial language Lojban,
which is occasionally discussed on sci.lang and elsewhere. Lojban is
designed for research in comp.sci and linguistics, and features an unambiguous
syntax and a predicate grammar.
More to your purposes, though, Lojban has included in its implementation a
method of expressing a variety of emotions far exceeding that embodied in any
written human language. Originally intended to eliminate the need for
encoding 'tone of voice' in a language which must be audio-visually
isomorphic, we attempted to construct a superset of possible emotion indicators
that might be expressed through tone of voice and various inerjections and
ejaculations.
Roughly speaking, Lojban allows encoding of emotions as a set of n-dimensional
scales with 7 values on each scale for intensity. There are about 40 of these
scales all mutually independent, and various modifiers that can expand these
scales into several domains (social interaction, physical effort, even sexual
expression). The system is being tested and to some extent used by myself
and others, who are able to converse in Lojban, if not fluently.
The emotional indicators operate almost independently from the regular grammar,
and can be used to express attitudes about the referent of a single word, a
phrase, or an open-ended scope of 1 or more sentences.


>From Charlotte Linde
To jean-marc fellous

 Subject: Time:11:50 AM
 OFFICE MEMO emotions Date:2/5/91
Try a paper called "goals, events and understanding in Ifaluk emotion theory: 
Catherine Lutz. In: Cultural Models in Language and Thought, ed by Doroty
Holland and Naomi Quinn, Cambridge U. Press, 1987.

>
 Offhand,
I cannot think of very much literature in ling. relevant to your work,
aside from 3 recent monographs by Zoltan Kovecses. They are:
 Metaphors of Anger, Pride, and Love (1986).
 The Language of Love (1988).
 Emotion Concepts (1990).
While I haven't read the latter works, I can say a word about the 1986
work. Kovecses argues that English has hundreds of metaphors -- for
example, "burning with anger" -- that can be explained because of the
physical correlates to emotion. For instance, the feeling of anger
leads to increased adrenaline and metabolism, hence to an increase
in body temperature.
 It might be more fruitful for you to look at work in semantic-
field theory. According to field-theory methodology, lexicologists
analyze words only in related groups. Your interest in the set of all
emotion-related words would seem to call for field-theory. For further
information, the best place to look would be Adrienne Lehrer's Semantic
Fields and Lexical Structure, 1974.
 In general, though, I think you can find your best sources in
psychology and in philosophy. While I am not up on the psychological
literature, I do know that a vast amount of it is dedicated to the
study of affect. In philosophy, you might look at David Perry's The
Concept of Pleasure, 1967. Not that it's particularly well-known or
comprehensive, but it does have a nice bibiliography.


>
 Hi. I saw your query on LINGUIST. I am not the one to reply, though you
r quest fascinates me. I'm not sure that a given set of words will consistentl
y elicit the same emotional response from subject to subject. All kinds of cul
tural stuff mediates between the stimulus (the gestalt of which is itself a cul
tural construct) and the response (the expression if not the experience of whic
h is likewise culturally constructed). Still, I also believe that a biological
 substrate underlies all of this. Eleanor Ochs edited an issue of Text on the
subject of Language and Affect in 1989 and her bibliography in her intro and he
r article co-authored with Bambi Schieffelin is quite exhaustive. It might not
 lend itself to your particular project, and quite likely you are aware of her
work, but I thought I'd mention it. I am a linguistic anthropology grad studen
t at UCLA. I had considered submitting a paper for your upcoming conference, b
ut am too swamped to make the deadline. Still, I take this opportunity to intr
oduce myself and let you know of my interest in your work. My work will hopefu
lly take me to Bangladesh next year on a Fulbright to investigate medical compl
aints as a highly emotional and culturally complex form of discourse.
 All for now. Let's keep in touch. Yours, .... Jim