Fighting Words
Dreitzes
that the research community indulges in way too much infighting, I just wanted
to take slight exception to the notion that we're markedly worse than scholars,
historians and even scientists in general.
Are we worse? Yeah, we're worse. But are we THAT much worse?
This is an exchange culled from a 1991 issue of *Etc.*, a journal of semantics.
Nothing has been changed except the authors' names. Enjoy.
*********************************************
Dear Editor,
Your Summer 1991 issue included C. E. Galbrenner's "The Conventions for
Symbolizing." Space precludes even listing all the flaws and fallacies
perpetrated in this article - let alone providing a detailed analysis. I will
expose just a few.
Galbrenner states on page 173: "Characteristically Western theories
systematically eliminate the observer from consideration." I agree that
statements like "It is green." or "It was found by experiment that water is
composed of two parts Hydrogen and one part Oxygen by volume." do not
explicitly indicate an observer. It is certainly possible to eliminate
reference to the observer, but there is nothing systematic about the possible
elimination. It is equally possible to include the observer. Statements like "I
see that it is green." or "Scientists found by experiment..." explicitly
include reference to observers.
The whole basis of empirical (modern) science depends upon observers; I (and
others) attribute the dawn of empirical (observer "corroborated") science to
Roger Bacon who wrote in 1268 in his *Opus Mias*:
There are two modes of knowledge, through argument and experience. "Argument"
brings conclusions and compels us to concede them, but it does not cause
certainty nor remove doubt in order that the mind may remain at rest in truth,
unless this is provided by experience.
Put simply, the philosophy of science characterizes modern scientific theories
as including law-like statements and observation statements. Observation
statements make
predictions about what may be *observed* by an *observer* upon taking certain
actions. For example, Ohm's law states: "The current flowing in a circuit is
directly proportional to the applied voltage and inversely proportional to the
resistance." One observation statement for Ohm's law could be: "If you double
the applied voltage in a circuit while holding the resistance constant, you can
expect to see the current doubled." Taking action and observing the result
(transacting) forms the basis of the development of, as well as the practical
use of "scientific thought", which according to Galbrenner's quote of Whorf,
"is a specialization of the Western Indo-European type of language." (p. 176)
When Galbrenner claims that "Characteristically Western theories
systematically eliminate the observer from consideration," he is simply wrong.
Not only do characteristically Western theories systematically *include* the
observer, some specific theories explicitly declare the relationship. The
Heisenberg uncertainty principle is a cornerstone of Western physical theory
and can be paraphrased that the observer always interacts (interferes) with the
observed. This principle not only makes observer interaction a basic tenet of
modern Western theory, it explicitly quantifies the result. Quantum mechanics
states that the process of observation "realizes" one potential state; there
isn't even a "what is going on" to find out about - what has happened was
created in the observation process.
Galbrenner, after stating that the observer is eliminated, states that
"physicists ... introduced the notion of 'taking the observer into account.'"
He falsely claims that this makes Western theories "remain inconsistent." In
the process of updating theories we discard older ones. The replacement is
often incompatible with the older theory. That incompatibility does not make
the newer theory inconsistent. It is not Western theory that remains
inconsistent; Galbrenner remains inconsistent when he states that Western
theories both eliminate and take into consideration the observer.
On page 176 Galbrenner asserts:
And since we humans have had no contrasting alternative to science as a
specialization of the WIE (Western Indo-European) type of language, we have
remained unable clearly to perceive the structure of presuppositions which our
dependence on the WIE type of language, and the grammar common to WIE
languages, has entailed.
Whorf equates "scientific thought" as the aforementioned specialization of the
WIE type of language; Galbrenner is therefore stating that we have no
alternative to scientific thought. In one stroke he dismisses all of
philosophy, religion, and a host of other views contrasted to this one
particular specialization of the WIE type of language. In particular he has
dismissed without basis the work of the philosophers of language as well as
linguists, who not only perceive the presuppositions made by language, but
question these presuppositions explicitly. There is simply no basis for
Galbrenner's rejection.
On page 177 Galbrenner states:
They treat any noun or noun-phrase as subject to Aristotle's Law of Identity,
and therefore, as identical with itself or self-identical. Conversely, they
treat any verb or verb-phrase as not subject to Aristotle's Law of Identity -
as not self-identical.
Korzybski did and contemporary general semanticists still do complain that the
language is not similar to "reality" in that words, including verb phrases, are
essentially static while "reality" is in constant flux. For Galbrenner to
suggest that verb-phrases are "not self-identical" while claiming that
noun-phrases are "self-identical" violates 59 years of general semantics theory
and practice. A verb-phrase is just as much the same verb-phrase as a
noun-phrase is the same noun-phrase. Galbrenner's suggestion that verb-phrases
are not self-identical confuses the map (the unchanging phrase) with the
territory (the changing action referred to).
On page 183 Galbrenner states that because WIE grammar does not provide a
primary means to distinguish between "map" and "territory" it constrains the
speaker to hold that the world really does not consist of static and unchanging
objects and transient relations between them. In philosophy the notion that
language "cuts reality at the joints" is known as naive realism. Few
philosophers actually hold such a view. Most have argued against it. Linguists
also deny that naive realism is implicit in the grammar of WIE languages. The
world-class linguist, Emmond Bach, addressed this very point in his
Chancellor's lecture at the University of Massachusetts at Amherst entitled
*Natural Language Metaphysics*. Bach concluded that the view - that the
structure of Western languages entails naive realism as a metaphysical system -
is forced on us, not even within "Standard Average European" languages (Whorf).
Bach
explicitly mentioned Korzybski and general semantics in this context as people
who thought the answer was obvious. Galbrenner's assertion is an extreme
version of naive realism. Galbrenner is stating that users of WIE language MUST
believe naive realism. We have over 27 centuries of disconfirmation. In every
age there were many users of WIE languages who argues against naive realism.
On page 178 Galbrenner claims to have developed an axiomatic system and to
have begun to use that system. On pages 187-190 he suggests that the system he
worked on is complete, correct, and tested; it is not.
During the 50s, 60s, and 70s, and even part of the 80s, the intelligence
community spent megabucks attempting to develop automatic translation programs
in order to translate Russian and other (then) "bad guy" languages for the
purpose of more easily gleaning intelligence. These efforts funded and
spearheaded what now has become part of the field known as artificial
intelligence (AI). In the beginning the projects looked doable, but as time
went from months to years to decades, the projects uncovered a depth of
ever-increasing nuances that defied any kind of axiom system or direct
translation. In 1979 I worked on a project which attempted to implement
translation of English formal logic statements in the LISP computer language.
Even that highly restricted project became more and more involved. The
intelligence community has coined the phrase "degenerating research project" to
cover the situation - a project which at first looks doable but gets worse as
more work is put into it. Work still continues in the natural language
sub-field of AI, but no one expects any short-term solution. It is generally
accepted now among AI researchers that no natural language can be axiomatized -
particularly in the manner that Galbrenner suggests.
Galbrenner is looking at only the surface but is drawing conclusions based on
incomplete and inconsistent work on a task that others have shown to be a
degenerating research project.
Space limitations preclude discussing many other flaws.
Cornell J. Ralphs
Alexandria, Virginia
*C. E. Galbrenner replies:*
Once more, Ralphs sounds the alarm against what he perceives as the fallacies
and mistakes of Galbrenner and his collaborators. His focus on our latest
publication misses the mark again.
What Ralphs says about the paper makes very little contact with the paper. He
does not bother to figure out what I say, much less pay attention to how I say
it. Thus his so-called argument not only fails to connect with mine, but the
result does not even make muster as argumentation, in the classical Greco-Roman
sense of the last 2500 years.
I feel somewhat grateful for his scrutiny. The points in the paper with which
Ralphs has the greatest difficulty may also trouble others, even if for
different reasons. And often, in the process of answering his protests, I end
up learning something. So I shall reply, in the first person, speaking for
myself and my co-authors.
Concerning page 173 of "Conventions," Ralphs' objections to my assertion that
"Conventional Western theories systematically eliminate the observer from
consideration" give him the opportunity for a diatribe on the topic of the
noun-phrase *the observer.* He says in effect that we know whether or not we
have taken the observer into account by whether or not we have mentioned
her/him by that name. This topic deserves a more ordered examination than
Ralphs gives it.
By 1905, after some 200 years of cumulative experience with Newtonian
mechanics, physicists had discovered that Newton and his successors had
included a fundamental theoretical error in the logical structure of Western
physical theory - they had left something crucial out of account. Einstein
found a way, with his theory of relativity, to take this "forgotten factor"
into account, by making a new distinction: He posits that light propagates at a
finite velocity, which remains constant for all observers (regardless of their
position, their velocities with respect to one another or to the light source).
When an observer occupies one position rather than another, with relation to
some event(s) s/he intends to observe, s/he introduces a bias into her/his
observations. Einstein's theories allow us to compensate for this bias.
Thus the construct of *the observer* stands for a CORRECTION FACTOR, which
compensates for and corrects that fundamental theoretical error encoded in the
standard frame of reference, the Newtonian world-view. Furthermore, the inquiry
into the foundations of physical theory did not stop with Einstein. The quantum
theorists disclosed further fundamental theoretical error, and proposed further
correction factors. I and my collaborators have continued the process. On pages
192-3, we summarize the epistemological bedrock which underlies our assertion
that the way of relativity and quantum theory handle the topic of *the
observer* makes those theories inconsistent.
Roger Bacon did indeed contribute to the tradition that eventually received
the contributions of Newton, Einstein, and others. But to quote Bacon's AD 1268
views apropos of a discussion of a paper which makes proposals at the "cutting
edge" of human knowledge in 1992 implies that the detailed issues addressed by
creative workers have not changed in over 600 years. Although Newton, like
Bacon, did practice alchemy, neither Einstein nor Galbrenner do.
Ralphs also misses the point the point that within the frame of reference I
have created, I make a number of corrections to the standard Western
Indo-European worldview. For example, I address - CAN address - one and only
one topic, namely, *one particular
organism-as-a-whole-transacting-with-her/his-environment at a date.* With Dewey
& Bentley, I use the term *transacting* to signify "back-and-forth 'doings'
between organism and environment which leave both fundamentally transformed."
With Sherrington, I regard this organism as "designed to do one main thing at a
time." In other words, I regard her/his "doings," as s/he transacts with
her/his internal and external surroundings, as *biologically "purposive";* and
I model such apparently "purposive" "doings" by means of Sommerhoff's construct
of *directively correlated.* Further, within this apparently "purposive"
setting, I specifically include the transactings of this organism with what we
commonly call "culture", "language," etc., as s/he functions as a member of
what Korzybski calls *the time-binding class of life.* By so doing, I do not
pre-judge concerning what CONVENTIONS FOR SYMBOLIZING this organism utilizes in
framing her/his "worldview," "language," etc. These conventions form the main
topic of the paper under discussion (although in his critical comments Ralphs
gives no indication that he has noticed the topic of my paper).
This focus seems sharply limited - shall I say delimited? - but speaking as a
student of "human behavior," I find that it includes most of what I want to
study.
Concerning page 177, where I point out my own view that native speakers of
Western Indo-European (WIE) languages treat *noun* and *verb* differently: I
propose one way to account for this difference. In his refutation, Ralphs
indirectly quotes Korzybski as complain(ing) that the language is not similar
to 'reality' in that words... are essentially static while 'reality' is in
constant flux." When I looked in the index of *Science and Sanity,* I found no
entry for *reality;* and found no such passage in the text, nor have the
resource persons whom I've consulted. I request that Ralphs provide a reference
to at least one place where Korzybski says that.
Meanwhile, Ralphs implies, e.g. in the paragraph preceding the spot where he
cites my page 177, that he has competence in linguistics, and in the philosophy
of language. The linguists in my research group aver that linguists have
standard ways of finding out how people handle various terms, various parts of
speech, etc. For one example, they propose "test sentences" for their
informants to read, and to judge on the grounds of whether or not the informant
would actually say things that way. In the context of a sentence like *Alice in
Wonderland's* "The cat grinned," people do, I maintain, say things like:
A cat is a cat.
Having agreed that that sentence seems well-informed, we can conclude that the
noun-form *cat* gets treated as *self-identical.* However, people do NOT say
things like:
Grinned is grinned.
Hence, we can conclude that the verb-form *grinned* does NOT get treated as
*self-identical.*
Furthermore, my proposals for how to account for this difference in how we
handle *nouns* and *verbs* have proved fruitful.
Concerning page 183, Ralphs objects to my observation that the WIE languages
have no PRIMARY way of distinguishing between "map" and "territory," and my
conclusion that therefore they tacitly constrain their speakers to posit their
identity. In discussing this point, Ralphs restricts his remarks to a
consideration of VERBAL levels - to what philosophers or other WIE speakers say
they believe. He mentions "views," what people "argue," what they "deny," the
points they "address," what they "conclude," etc. He does not refer to
non-verbal behavior or "doings." In particular, he does not use, mention or
refer to Korzybski's suggestion that we regard what we humans DO as analogous
to a formal deductive system - AS IF what we DO follows strictly from what we
ASSUME. For Ralphs, apparently, "assumptions" exist only on verbal levels,
perhaps as "sentences." But the psychoanalysts showed, almost a century ago,
how feeble a force in human behavior the verbal-level "beliefs," "opinions,"
etc., exert - as compared with that of the non-verbal logic of the unspoken. I
take "doings" as primary evidence. The evidence often indicates a profound
split between what we do and what we say.
Concerning page 178, Ralphs objects to my claim to have developed an axiomatic
system and to have begun to use that system. In fact I have developed, and
published, three successive systems - one framed in ordinary scientific
English, one framed in the mathematical theory of sets, and one framed in a
non-standard notation built up on a grammar derived from Korzybski's
non-aristotelian premises. I have spent some twenty-nine years creating and
using this increasingly rigorous series of systems. One of the products of this
work consists of the paper "The Conventions for Symbolizing," which the editor
of this journal published and which Ralphs here criticizes.
Concerning pages 187-190, Ralphs attributes to me the view that "the system he
worked on is complete, correct, and tested" (but, Ralphs asserts, "it is not").
I do claim the system - systems - seem "complete" enough to elicit continued
study and efforts to go on developing them. AS FOR claiming my theoretical
system(s) as "correct," to do that would violate my own chosen assumptions. I
feel so sure that my system(s) remain(s) inaccurate, incomplete, and
self-referential that I have built it (them) up on the non-aristotelian
premises of Korzybski - the Postulates of Non-identity ("inaccurate"),
Non-allness ("incomplete"), and Self-reflexiveness ("self-reflexive"). AS FOR
claiming my theoretical system as "tested" - yes, I and various of my
colleagues have done some preliminary testing; but no, I don't consider the job
of testing as finished and done with. I follow Popper, Godel, and others in
doubting that we can ever find any system 100% "proved." Indeed, I don't regard
the job of testing the First Law of Thermodynamics as finished and done with.
Concerning page 176, I develop the view that our science still grants a
privileged position to the grammar common to the WIE discursive and formalized
languages. In passing, I support my view by quoting Whorf: "What we call
'scientific thought' is a specialization of the Western Indo-European type of
language." Ralphs interprets Whorf's remark as what, elsewhere, he calls a
"law-like statement" - as if Whorf maintained that *scientific thought* can
exist ONLY as a specialization of the WIE type of language. In sharp contrast,
I read Whorf's comment as a historical observation: what we call *scientific
thought* actually arose within the cultural and linguistic setting of the
Western cultures and the WIE languages (and the time has come to transcend that
limitation).
Here, then, we have a vivid example of how different the observations made by
two different observers appear, as they consider a single text - a single
sentence. This suggests that Ralphs' assumptions differ sharply from my own.
I maintain that twentieth-century science has broken with the WIE tradition in
certain ways, but not in others; and that Korzybski and his allies have made
explicit the revision of assumptions entailed in that break with tradition.
Furthermore, I and my collaborators have built on the revised assumptions which
Korzybski provided, until we have generated an entirely new frame of reference
- thus making systematic our break with tradition.
In the paper under discussion, I examine the process of making such an
alteration of premises, and discuss how this process alters our view of the
uniquely human environment, the matrix we often call *language* and *culture.*
I propose a general structure of assumptions which has a certain number of
"blanks" in it. With the blanks filled in one way, you have a subject-predicate
language like English or the mathematical theory of sets. With the blanks
filled in another way, you have a language like the non-standard notation that
I and my colleagues have begun developing. Presumably, with the blanks filled
in yet another way, you would have a language like Swahili, or Haida, or
Japanese.
I expected these proposals to seem uncomfortably unfamiliar - though I hoped
that people competent in general semantics might find them less troubling than
might the general public.
On page 184, I point out that neither English nor any other WIE discursive
language arose as a deliberately generated axiomatic system. To treat English
as if it functions as one - as we avowedly do - involves a deliberate pretense,
a conscious AS IF. Perhaps, when I undertake to build on this acknowledged
pretense, I utilize different axioms than those which Ralphs and his friends
used. Does Ralphs really want to go on record as arguing from the unworkability
of one or more sets of axioms for one task to the inevitable unworkability of
another set of axioms for another task?
As for Ralphs' failed efforts to axiomatize discursive languages for the
purpose of facilitating machine translation, I can empathize with anyone whose
work has come to nothing. However, I wonder about any comparisons between those
efforts and my own.
C. E. Galbrenner
Columbus, Ohio
Posted by Dave Reitzes
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * \8^)
For my series on the research and theory of John Armstrong, please see my
directory at Deanie Richards' JFK Place:
gopher://freenet.akron.oh.us/h0/SIGS/JFK/Only/JA/DR/daver.html%09%09%2B