Deixis, Buhler and the Problem of Ambiguity

[Tahir Wood]

I agree with the broad notion of deixis; I have never thought of it as a small class of linguistic expressions. But I don't agree that it has much to do with ambiguity. It seems to me that deixis is more like the pole of concrete as opposed to abstract in language, or specific as opposed to universal. So a linguistic expression will tend to have a deictic content as well as an ideational content, but the one can predomiante over the other in a specific instance.
Tahir

>>> <Salinas17 at aol.com> 05/17/09 9:09 PM >>>
(First, my deep appreciation to this forum for an opportunity to try out
these ideas and to those kind enough to consider them.)

It’s striking how the term "deixis" has shrunk in meaning since it was
first coined in linguistics.

It's common these days to say that deictic words are a special class of
words or utterances that "need external context", or "require contextual
information" or "depend on an external frame of reference."   And to use
demonstratives (this, that) as examples.   And to make a comparison to anaphora,
another "special class" of expressions.

But in reading Karl Buhler, whom I think first used deixis as a formal
linguistic term, you don't find those kind of limitations.   It seems that
originally deixis was not just a class of expressions but the “fundamental” base
of all "linguistic messages" or all "intercourse using language."

This is from Buhler's The Theory of Language (the English translation from
1991):
"...in linguistic messages there are two closely interlocked fundamental
processes which we can and must distinguish in order to understand what is
going on.   In intercourse using language there is first pointing: things and
processes are indicated.   That is demonstratio; I prefer the Greek word
deixis.   Second, there is also representing... Objects and states of affairs
are given a formulation in language and are symbolized by words that designate
them in the symbolic field of language."

It's not easy to read the segment above and think that Buhler was writing
about a "special class" of expression.   And it's also difficult to think
that Buhler meant "pointing" literally.   (The Greek word deixis in its most
basic sense meant the act of bringing something to light, un-hiding,
revealing; more abstractly showing proof, demonstrating, or more concretely an
imaging.)

How deixis got linguistically limited to where the pointing has to be
almost explicit, I'm not sure.   But that contraction in definition might perhaps
have been unfortunate, as it perhaps made deixis a bit more difficult to
explain.

The authors of the Handbook of Pragmatics (2006) in introducing the subject
write:
“One persistent complication for any theory of reference is the ubiquity of
deictic or indexical expressions.   Deixis characterizes the properties of
expressions like I, you, here, there, now, hereby, tense/aspect markers,
etc., whose meanings are constant but whose referents vary...”

Now, there's a peculiar linguistic beast -- meaning stays constant, but
referent varies!   Some would say that a different referent equals a different
meaning.   How is “meaning” defined here, if it doesn’t include reference?
 How do we separate the two ideas for the sake of defining deixis?

Here’s an example of the difficulty. Recently, I've met different persons,
each one introduced with the phrase, "This is John."   The referents
definitely varied. Does "John" have the same meaning but varying referents * more
than one John?   So therefore is “John” a deictic?

Stephen Levinson, in the same volume, goes on to explore in detail the many
paradoxes and ambiguities of this narrowly defined linguistic “deixis”
(along with the akin philosophical “indexicality,” using Charles Pierce’s
terminology.)

Levinson, points to the many ambiguities that deixis can create in
describing time, place, space, person, etc.   And they CAN be confusing!   Like:
today referred to today yesterday. But, today, today refers to a different day.
  Or, of course, I is I if I say I, but if you say I, you mean you, not I.
There are many more, needless to say.

And because of this, we learn from Levinson that “semantic deficiency” is
a defining characteristic of deixis (or “indexicality”.)

Now, this might seem contrary to our more common understanding * deixis
would seem in fact to remove ambiguity, not add it.   Even definite articles
are not as definite as some deictics.   "The car is mine" could refer to any
one of millions of cars.   But "THAT car is mine" suddenly becomes more
specific, less ambiguous.   "Truths" can refer to millions of different truths,
but "we hold THESE truths to be self evident" tells us that all the truths
are going to be narrowed down.   “I’ll get it to you today,” promises a more
specific time than “I’ll get it to you.”

Even when the deictic must be ambiguous to be accurate -- “Somewhere out
there.” -- it still contributes to a greater specificity by the most elemental
logic, being unambiguously ambiguous * “Not here, but somewhere out there.”
   The absence of the object, not “being here,” is quite definite.

How can this be?   How is that this special class of expressions can
linguistically increase ambiguity, when we seem to use it for the exact opposite
purpose? Is it that deictic language, in promising to be specific, is held to
a higher standard than language that doesn’t make reference to “external
context”?

The answer, according to Levinson, seems to have something to do with what
linguists traditionally study and some kind of a conclusion from somewhat
dubious animal psychology:
“Students of linguistic systems tend to treat language as a disembodied
representational system which is essentially independent of current
circumstances, that is, a system for describing states of affairs in which we
individually may have no involvement* It is these properties of language that have
been the prime target of formal semantics and many philosophical approaches to
language * and not without good reason, as they appear to be the exclusive
province of human communication. The communication systems of other primates
have none of this “displacement” as Hockett (1958: 579) called it.”

So, compared to “language as a disembodied representational system” * one
with no or at least relatively few ambiguities * deixis seems replete with
them. Here we are using “this” and “that”, “I” and “you”, “today” or “
next week” to be more specific, and it ends up we are being more “
semantically deficient,” according to this point of view.

Obviously, one way to find deictic expression guilty of ambiguity is to
assume away the ambiguity in non-deictic expression.   Say “Socrates is a man”
and you are fine, but say “Socrates is that man” and you are in trouble.

Of course, these ambiguities just get an awful lot worse if we literally
follow what Buhler seems to be saying and consider that all "linguistic
messages" as grounded in deixis * not just a special class of expressions.   (cf.,
Lyons (1972) “Deixis as the Source of Reference”)

It should also be pointed out that Buhler makes deixis just part one in a
two part “interlocking” process.   The second part is “representing...
Objects and states of affairs are given a formulation in language and are
symbolized*”

So that in Buhler’s system, deixis is not even a class of expressions at
all, but what happens before something * whatever it is, an object, a process,
an idea or even another word * can be translated into a representation, a
symbol or a “linguistic message.”   If that’s the case, then deixis is
pretty much an individual event, even a private one, since it occurs before
language can be used to share it.

There are a number of interesting ways to look at ambiguity in language *
even ambiguity in syntax.   But one way of looking at it is a problem created
by a hundred million personal deictic events becoming representations,
symbols that can be in some way and to some degree unambiguously shared by a
hundred million speakers of the same language.   How is that accomplished?

Thus our focus shifts from ambiguity as a problem -- the common approach --
to how language solves that problem.

steve long












**************
A strong credit score is 700 or above.  See Yours in Just 2
Easy Steps!
(http://pr.atwola.com/promoclk/100126575x1222585011x1201462751/aol?redir=http://www.freecreditreport.com/pm/default.aspx?sc=668072&hmpgID=115&
bcd=Maystrongfooter51709NO115)