[Corpora-List] Quantitative Corpus Linguistics

[J Washtell]

Quoting Wolfgang Teubert <w.teubert at bham.ac.uk>:

> Dear All,
>
> I am very grateful to Linas for his comments on our mental concepts   
> of chairs and tables. He knows that it is 'something more' than what  
>  has been said in the discourse.  I am not so sure. There have been   
> many acts of what I call primordial speech situations in which   
> chairs and tables have been pointed out to me. For me, such things   
> are props that make up an extension of the speech situation. The act  
>  of ostentation make them part of the communication. If I had a dog   
> it wouldn't know what a chair is. It would not even have noticed it   
> had been part of such a speech situation. I recall discussions with   
> people (not with dogs, though) in which we attempted, for instance,   
> to negotiate if barstools with a back rest should be referred to as   
> stools ("Could you bring two more stools over, please?"). Does the   
> result of such speech situations and negotiations mould my mental   
> concept? I think it does. Does each of us have the same (perhaps   
> innate) concept of a chair (as Fodor sees it)? Rather unlikely. Does  
>  everybody have their individual concepts? If yes, how do we   
> understand each other, except by assuming that meaning is in the   
> discourse and not in our heads?

Dear Wolfgang and interested parties,

I have not read and digested the entirety of this thread because there  
aren't enough hours in each day, so please forgive me if I am going  
over old ground (I am assuming here that others do have the time to  
read this stuff :-)). Certainly, most of what I have read, I have read  
with great interest.

You say you recall conversations in which you were trying to discuss  
whether a given object was, in actual fact, a chair or a stool. I  
expect you also recall discussions with people in which you negotiated  
the meaning of a word, or negotiated the word to use for a  
particularly abstract concept, for the purpose of carrying out a  
single conversation.

It seems to me that a very plausible alternative to the options you  
provide above is that the very purpose of (symbolic) language is to  
allow us to transmit information between unlike domains (your head and  
my head). Your mental description of the thing we refer to as chair in  
our conversation will never precisely match mine, or anybody elses,  
because our experiences (and our wiring, perhaps) may be quite  
different. If you had to communicate directly with everybody in their  
native currency of thought therefore, you would have an awful lot of  
negotiation to do. So instead we each learn a single (albeit rather  
dynamic) mapping - that between our thoughts and a common language (in  
the present case  some version of English) - in which the terms are  
well enough grounded (in historic discourse, through colocation with  
events, gestures and other terms) that we can be relatively confident  
that the meaning evoked in my head is *effectively* equivalent to that  
evoked in yours; effective in the sense of allowing us to benefit from  
the conversation.

This is *not* to say that the meaning of these terms in and of  
themselves is absolute, or even existent. Rather, meaning exists in  
the mind of the speaker, and the listener, and in the language  
insomuch as it has a speaker and a listener and (hopefully) a useful  
overlap in their individual understandings. To suggest that meaning is  
absolute in the discourse naked of the participants, or in the history  
of all discourse, is not a necessary step. You might say there is some  
"mean" meaning, or least common denominator of meaning contained  
therein, or that it represents "effective meaning" when it has  
ramifications within the individuals reading it. To say that a chair  
or a table is exactly what people say it is, no less, is to imply that  
the communication between speaker and listener is perfect, and  
perfectly reproducable. This is analagous to saying that a property  
has an intrinsic value, rather than just a market price or a sale  
price (except that with communicable concepts we are dealing with  
something more qualitative perhaps - though some might say it is  
simply something of much higher dimensionality). In the case of table  
and chair this is probably a fair simplification of the truth (the  
buyer's offer readily matches with the seller's expectations), in the  
case of the kinds of elusive concepts being discussed in this thread  
say, it is not very useful at all (I think what we're negotiating over  
is a classic sportscar, you think it is a heap of junk).

As far as this pertains to natural language processing: once a  
computer begins to communicate intent (to request data say), and  
understand and act on users intent, as expressed in a common natural  
language... then I would hazard that there is an exchange of meaning  
taking place, and that the conversation is successfully encoding that  
meaning. To analyse corpora for patterns and clusters of words might  
be a necessary step in getting us there (by way of familiarising  
machines with the common currency, and its surface usage), but it is  
not a practice which, by itself, courts meaning (bring human  
interpreters of those patterns into the equation though, a la corpus  
linguistics, and it does).

If we do not agree over this interpretation, I would suggest that the  
most likely problem is that we have not settled on our currency. I  
note that terms like "meaning" and even "corpus linguistics", at  
present, seem to be particularly heterogeneous in the minds of the  
participants (despite seeming superficially consistent in the corpus!)  
Indeed, some of these terms are ill-defined (not neatly mapped) within  
the minds of individual participants (like me). I expect that if we  
were to resolve some of these differences by building an open-source  
glossary say, and refining its granularity until everybody here could  
more or less agree with it (although some would certainly have to  
sacrafice some of the specific mappings they had grown to love, in  
favour of new terms), then what we would be left with would be our  
genuine disagreements... which we could then set about trying to  
resolve through rather more precise (and perhaps not so verbose)  
discussion.

Justin Washtell
University of Leeds

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora