LINGUIST List 2.143

Wednesday, 17 Apr 1991

Disc: Intuition, Phonology, Structuralism

Editor for this issue: <>


  1. , Re: intuition
  2. "Michael Kac", The Truth About Everything
  3. Bill Eldridge, Intuition

Message 1: Re: intuition

Date: Tue, 16 Apr 91 13:10:50 PDT
From: <>
Subject: Re: intuition
Richard Ogden challenges my use of 'intuitively' in connection with defending
a linguistic claim. I did make some effort to describe the basis for that
intuition--even going so far as to provide an external 'litmus test' for it.
I did not wish to propose the use of intuition as a sound basis for claims
in linguistic argumentation. So I fully agree with Dr. Ogden's qualification
of his own remarks:
 --- I'm not saying that imagination and hunches
 don't have a place in linguists' arguments, just that they are not enough
 in themselves to explain a position.
I felt that I had inserted the word 'intuitive' at the appropriate place in
the argument, and I apologize if I presumed too much.

 -Rick Wojcik
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: The Truth About Everything

Date: Tue, 16 Apr 91 21:25:46 -0500
From: "Michael Kac" <>
Subject: The Truth About Everything
Some reactions -- brief, I promise -- to a variety of recent postings.

Alexis Manaster-Ramer takes issue with the SPE analysis of words 
like *industry* in part because it leads to underlying representations 
containing unusual consonant clusters. That's a common enough kind 
of criticism of UR's and my inclination is to think that it's valid. 
Ironically, however, Chomsky in *Current Issues* used just such an 
argument to try to subvert the Invariance Principle by arguing that 
one is forced by this principle to e.g. analyze the [D] in an English 
word like *throw* as phonemic /t/, thus leading to phonemic 
analyses in which initial /theta-t/ clusters occur 'counter not only to 
the speaker's intuition but also to the OTHERWISE VALID RULES OF 
CONSONANT DISTRIBUTION [my emphasis -- MK]'. I have always 
been troubled by this case because it seems impossible to have 
nontrivial phonemic or underlying representations at all unless they 
violate just such valid rules -- as SPE-style UR's do routinely.

My question is really this: what are the rules of evidence here? I 
think that this is a relevant question in the context of the animated 
discussion now going on regarding the history of phonology.


Richard Ogden says 'Isn't [intuition] just a way of avoiding a more
rational approach? --- I'm not saying that imagination and hunches
don't have a place in linguists' arguments, just that they are not 
enough in themselves to explain a position.'

Part of the problem here I think is that linguists use the word 
'intuition' in two different ways. On the one hand, they mean the 
native speaker's 'feel' for the language -- e.g. the feeling that there's 
something wrong with *She are here* or that the initial sound in 
*pin* and in *spin* are the same. Exercising intuition in that sense is 
taken to be the way by which we generate data. The second way in 
which linguists use the word 'intuition' is as a synonym for 'common 

I agree absolutely that it is no defense of an analysis that it is 
'intuitive' in the latter sense, nor an argument against an analysis 
that it is 'counterintuitive' (in the same sense). The whole reason that 
we have science in the first place is precisely because common sense 
isn't a reliable guide to the way things really work.

It is possible, however, to accept what I've just said and still believe 
that it's legitimate to rely on intuition in the first sense. Granted, 
there's controversy on this score too (Bill Labov having made some of 
the more provocative statements for one side of the question). I 
won't take a position on that matter here (though I do have one!); I'll 
just say that I think it's easy to get led up the garden path by not 
keeping straight which kind of intuition is being referred to.

By the way, the question of just what kinds of intuitions native 
speakers do an don't have is itself one that I think is interesting and 
not nearly as straightforward as some discussions would suggest.

Vicki Fromkin writes:

'"Taxonomy" is not a pejorative term unless it is used
as a term for "theory". D'Abro's "The Rise of the New Physics" has an
exdellent introductory chapter on the development of any science --
observation as first stage, classification as second, and theor~?
assume that classification isn't a theoretical activity (as D'Abro's 
trichotomy seems to suggest?)

'Evidence to support a theory is not proof in the
sense of a mathematical proof -- and of course there is no such proof 
in an empirical science. In addition, and you know all this much 
better than I do Michael, a mathematical proof is a deductive 
procedure not an inductive one. And therefore should not be seen as 
a parallel to a discovery procedure. Noone can deny the importance 
of rigor -- but that is no~?

Yes, of course. The parallel I meant to draw between discovery 
procedures and the formalization of the notion of 'proof' was 
intended to go only this far: Both were intended to increase the level 
of rigor within the relevant disciplines, and in both cases it's 
recognized that this level of rigor is seldom adhered to in practice.

Adam Kilgarriff alludes to the mothers that bore us. I don't know 
about his mother, but mine is quite interesting!

Michael Kac
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: Intuition

Date: Wed, 17 Apr 1991 13:19:14 SET
From: Bill Eldridge <EXT28%CSPGCS11pucc.PRINCETON.EDU>
Subject: Intuition
In response to Richard Ogden on Intuition:

Intuition is a hypothesis to test and support or refute. Obviously
intuition is a personal experience - Einstein's intuition that "time"
varied according to personal viewpoint was greatly at odds with the
majority intuition in his (and maybe our) day. However, intuitions are
quite often completely invalid, and that's why it's important to stamp
our concepts as "proven", "refuted" or "speculative", to whatever degree
it's possible to be certain in a non-digital world. A good intuition
can provide the starting point for a good theory (or a bad one), and in
a very practical world, the theory is often derived from a functioning
system or notion. The current popularity of neural networks seems to be
a wonderful example of this, since there seems to be little to explain
why the technique works so well (counter to sequential methodologies).
More times than not, we are unable to completely explain phenomena,
and expecting people to wait at the starting gate until everything is
solved is unrealistic and counter to a pioneering spirit (we made a lot
of progress while presuming Euclid's fifth postulate was always true).
 On the other hand, intuitions should not be accorded the sanctity of
a proven or supported concept (and accepted and "proven" ideas need to
be reviewed periodically as well in light of new developments). Many
things are done in the name of expediency or convenience, and part of
the purpose of theory work is to find the more useful and generally
applicable aspects of this. Even in theory work, it seems easier and
more productive to build up straw men to knock down than to continually
struggle against a void (and sometimes these straw men are a lot tougher
to knock down than one would suppose). The bottom line is that one should
try to be aware of underlying assumptions in any endeavor.
 As a possible explanation of why there are so many misinterpretations
of network postings, I can say that right now I'm unable to call up the
orginal message I'm responding to so that I can verify several points,
or at least not conveniently. I'm reminded of the old Saturday Night Live
routine in which a slightly deaf Gilda Radner would lambast an idea for
five minutes before Chevy Chase would inform her that someone was
contemplating a "trade embargo", not "trading Garbo", or some such
misinterpretation (I can't provide any actual examples from the show).
I hope all of this is of some use, even if it's irrelevant to the
original point raised.

 Bill Eldridge
 Czech Academy of Science

[End Linguist List, Vol. 2, No. 143]
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue