home
linguistics
Is Evidence Based Linguistics the Solution? Is
Voodoo Linguistics the Problem?
By Alexander Gross
The author's most recent invited paper, presented
at Dartmouth during the LACUS Conference, August, 2005. It
summarizes the faults of current linguistics and proposes a new
overview for future developments.
This session was chaired and introduced by Professor Sydney Lamb, and
I was particularly moved by the tribute he paid in his introduction
to the decades I have spent pursuing the intricacies of language in
many fields. Three days after presenting this paper, a
follow-up two-hour workshop on these themes also took place.
Its text can be found by clicking
here.
ABSTRACT: A voluntary ten-question true-or-false self-test*
will be distributed, enabling participants to silently
experience for themselves to what extent they may have been
practicing either of these two forms of Linguistics. Of these
two terms, "evidence based linguistics" springs from
evidence based medicine (EBM), **
a movement that has recently made great strides among health-care
professionals to combat many arbitrary clinical procedures,
unexamined assumptions, allegedly authoritative rules, and
institutional shibboleths, along with advertising hyperbole and
unproven claims for impending cures or imminent progress, that have
begun to clutter the study of medicine. "Voodoo linguistics"***
is simply an outgrowth from the book title Voodoo Science, a term
used by University of Maryland physicist Robert L. Park to describe
much contemporary science, including widespread and often
government-supported superstitions concerning the laws of
thermodynamics, the causes of cancer, and the space program.
Although the applicability of both terms to linguistics may
already be clear enough to many LACUS members, the probable results
of this test are likely to demonstrate how deeply comparable voodoo
elements, unsupported by any defensible evidence, have crept
into the study of language. The first quarter hour will be devoted to
explaining the questions in the test, intended mainly as a heuristic
exercise which no one needs to sign or hand in.
A crucial section will explain how Guidelines for EBM function in
practice and attempt to set out a comparable series of Guidelines for
EBL, springing from the basic principles and procedures of
science-method, skepticism, and replicability. EBL
Guidelines will clearly both resemble and differ from those for EBM,
requiring a review as to how the former have been derived and are
likely to be applied in the future.
A final section will examine whether important evidence about
language, whether originated by linguists or rooted in other sciences
or springing from related areas, may have been ignored or suppressed
during the past because of preconceptions related to these ten
questions. The sciences and fields of study discussed will
encompass—but not be limited to—cartography, fractal forms
in nature, translation studies, creating viable laws and measuring
units for language, neurocognitive networks within language, and a
complex of physiological issues related to various language
processes. The methodology of medical diagnosis will be invoked
in an attempt to describe the various stages and degrees involved in
language learning. It is hoped that from these processes there may
emerge a better grasp of the full size and scope of language and a
clearer notion of what the desiderata might be for creating truly
viable and relatively complete theories in our field.
*You may see a copy of this questionnaire by clicking here.
** Those wishing to learn more
about EBM may find the following search pages of use:
http://search.netscape.com/ns/search?fromPage=nsBrowserRoll&query=evidence+based+medicine
and:
http://www.cebm.net/ebm_links.asp
[back to abstract]
*** Park, Robert L. 2000. Voodoo
Science. Oxford: Oxford University Press. Professor Park
also directs the DC center of the American Physical Society and is a
recognized media spokesman for his profession. He contributes a
weekly column on science issues to the APS website at: http://www.aps.org/WN/
[back to abstract]
The full text of this paper follows:
First of all, I feel truly honored to be here, and I want to thank
Sydney Lamb and the conference committee for having invited me.
I'm particularly happy that this is an international conference with
participants from many nations and has been jointly sponsored by both
Canadian and US linguists. I have always welcomed the presence of
Canada as a strong dissenting voice on our continent. And
through Canada it is also a bilingual conference, et c'est la raison
pour laquelle je voudrais aussi, pour commencer, rendre hommage
à la langue et à la culture françaises.
I'd like to start with a few disclaimers, perhaps even something
of an apology, since at least some of what I'll be saying might be
perceived as negative about the last five decades of linguistics,
though I'm not altogether negative about the future. If at
first I seem to be a desecrator of idols sacred to many in our field,
I promise I'll also do my best to suggest a few truly scientific
approaches to take the place of so-much misplaced idolatry. I'm
also keenly aware that some of my remarks might be more
appropriate at an LSA conference than here, but I've become concerned
that a few viruses from the LSA-MIT Petri dish may have become so
widespread that they have to some extent contaminated the study of
linguistics everywhere.
During recent decades we have seen concerted and repeated attempts
to reformulate linguistics either as a branch of logic, or as a
branch of philosophy, or of psychology, or sociology, or
socio-psychology, or even of paleontology. Not to mention
several different branches of math and statistics, especially
computational mathematics, even to the extent of attributing
mathematical precision to grammatical relationships, regardless of
the context such relationships might be attempting to represent.
I hope I won't shock anyone too much if I suggest that linguistics
can be far better described as an offshoot of yet another science,
not the most precise of studies but a science nonetheless, the
science of physiology, itself belonging to the study of
medicine. My claim to competence here is grounded in two areas,
first some four years spent in the company of American doctors,
nurses, and physical therapists in the detailed study of Chinese
medicine, which means that I have to some extent acquired not merely
medical but bi-medical knowledge and experience. And secondly from my
own lifetime of experience with an extremely rare endocrine
condition, which has limited my achievements in some respects but
greatly enriched them in others and has inspired me to seek out
knowledge in a number of nations and cultures and areas of
study. And assuming this all sounds a bit too embedded in
biological variables, during my workshop I'll also be defining some
more precise linguistic values springing from the cartographical
dimension of language, and here too I'll be drawing on my own
real-world experience.
Also, since I'm going to be dealing in some detail today with the
nature of evidence, this necessarily means that I'll also be doing my
best to present credible evidence for every statement I myself may
advance, even about evidence itself.
But even here there are some real limitations involved, which
leads to my next disclaimer. Just this year The Edge Foundation
challenged 120 of the world's top scientists to spell out what they
themselves regarded as true, even though they had no real evidence to
prove it. And not a one of these eminent scientists failed to
come up with one or more notions they subscribed to, even where final
proof was lacking. (Note 1) So
despite the title of this paper, evidence may not always be
everything. Or even invariably easy to gather and verify.
Though it's still also true that it's better to have more evidence
backing up one's claims than little evidence or none at all, since
the latter case, as most of you know, has come to be something of a
norm in our field.
As my final disclaimer I'd like to take a stab at articulating one
of the thorniest problems that faces us whenever we try to talk about
anything, including language. It's simply this: that language,
by its very nature, often forces us—even fakes us—into
oversimplifying complex issues, often by pretending that there are
only two possible views about any subject, or members of only two
camps who are qualified to offer opinions, when there may in fact be
hundreds of such views and/or informed parties.
Whenever you want to say something/slash/anything and want it to
mean close to the same something/slash/anything to others as it does
to you, you are all too often forced to oversimplify. And today I too
may have been faked out, simply by the nature of language, into
something of the same position. It is difficult to emphasize one
facet of reality without simultaneously ignoring many other
aspects. Language does this to us all, scholars and laymen
alike, perhaps far more often than we are aware. I've even
formulated a linguistic law about this and related phenomena.
And by mentioning that "linguistic law." I've given you
a hint about what I'll be covering at the Saturday workshop/tutorial
on this topic. Today I'm doing my very best to be modest with
all manner of disclaimers and apologies, but later this week I'm
likely to be a good deal bolder. †
In fact, I wouldn't be surprised if I can promise you that I may well
be able to tell you a good deal you haven't heard about language and
linguistics before and from several rather unusual viewpoints
† (for text of Saturday workshop,
click here)
But I'm getting ahead of myself. Let's go straight to the
core and look at that little quiz, which you have so far seen only as
a set of true-or-false statements. Here are some copies of the
final version, with the scoring apparatus explained at the bottom.
(to quiz) [handing out copies of quiz with
scoring method] I hope I won't be stepping on too many toes here
by telling you this, but every single statement in that quiz is not
just a little bit false, it is demonstrably and devastatingly false.
Almost all of them are close to being classical examples of
prejudging the results of scientific research before actually
carrying out that research. And to the extent that any of you
may have marked any one of these as true, to that extent you are, in
your mindset and research goals, practicing not Evidence Based
Linguistics at all, but Voodoo Linguistics.
Let's see how I've expressed it: "Score ten points for each
question marked as true...From forty to fifty, the infection has now
spread and presents as Moderate Voodoo Linguistics, seriously
affecting your ability to practice Evidence Based Linguistics."
If you believe even as few as three of these statements, your mind
may have already been partially poisoned. If you believe
as many as five or six, your ability to think clearly about language
may have been seriously impaired. If seven or eight or more
ring true to you—and I doubt if this is true of anyone
here—then there may be no way I can reach you today.
If I were to explain in detail as to why each of these ten
statements is false, it would take me at least twenty hours to
do so, that's how false they are. So what I'm going to do right
now is summarize why I believe they are false and then go through two
or three of the most important ones more thoroughly. But I
don't think I'll be cheating you by this approach, because I'll be
here for not just the next twenty hours but for the next three days,
which if we count our waking hours makes closer to thirty, and I'm
perfectly willing to expand informally on my summary during that
time. I don't want anyone's feelings to be hurt if I have
trampled on any ideas you happened to regard as sacred or even as
merely acceptable. But during this session my time is limited,
and I want to get past the answers to this quiz to more positive
sections about Evidence Based Linguistics and possible insights we
can gain into language from a variety of disciplines.
So let's summarize. Of the ten statements I've asked you to
mark as true or false, only two, Numbers 5 and 6 about Whorfian
theories, qualify as well-formed statements about linguistic ideas,
and they are both blatantly false. All of the other eight
statements are nothing but more or less idle and more or less
ignorant boasts about the nature of language, a kind of superstitious
whistling past the cemetery at night by some three generations of
linguists because they have been afraid or unwilling or unable to
look at the terrain around them.
I'll deal with the first two statements in a few minutes, but let
me start now with number 3. The claim that math and
computational math may be more complex than language does not mean
they can necessarily solve linguistics problems, to do so they would
require not greater complexity but the correct degree of
complexity. And even then they would need to be applied
correctly. As some of you know far better than I, there are
vast numbers of math solutions and formulations in existence for
which no real-world problem has yet been found.
Statement 4 about breaking the genomic code and nanotechnology is
perhaps the emptiest boast in the list, since there is no certainty
that either achievement will have any bearing on language, which may
well stand on its own as a unique network located outside the bounds
of either, owing its composition to roughly equal portions of human
sense and human slime. I'll be discussing this a bit later on,
especially the slime factor.
I'll also come back to the two about Whorf's theories, which
brings us to number 7 about machine translation. I've written
so often about MT that I don't want to become too embroiled
here. Let me allude to only three possible problems. A
detailed report recently published in Translation Journal (Note
2) carefully reexamines the entire technical
and historical MT background and concludes that even at its
best MT will probably never serve as anything more than a translation
tool for those translators willing to use it.
Just to remind you, it was claimed as recently as 1990, among
other ludicrous boasts, that by using MT systems unaided
monolingual workers will perform "truly automatic translation
.....without assistance from bilinguals, polyglots or
post-editors.....but meeting the quality standards of professional
translators—no less." (Note
3) But I don't think I need to tell you
that this prediction never came true. The man who wrote it now runs a
company specializing in Translation Memory systems, about which he
makes no such extravagant boasts.
The second and far deeper problem is that at present no MT or TM
system (TM standing for Translation Memory) can provide
reliable translations even for set phrases without the editorial
participation of professional translators, those very people whom
Geoffrey Pullum (Note 4) and
others predicted MT would soon be supplanting, 4 besides which
relatively few professional translators are willing to work with
either MT or TM. In any case, MT now accounts for only about one
percent of the roughly 450 millions of pages being translated each
year. The third problem is that the main parties who foresee
any increase in this percentage are, as I have written elsewhere,
"market analysts," individuals with no knowledge of
translation or language worth mentioning who nonetheless cannot
control their imagination whenever they contemplate the multi-billion
dollar translation market. And this is why, as I have written
elsewhere, claims like the one I have cited keep cropping up and why
funding continues to flow to those making them.
As for Number 8, that words and morphemes have "core
meanings" whose study can help create "universal conceptual
glossaries," this claim is totally ludicrous and can only be
advanced by utter amateurs at the language game. It may well
turn out that even the most common of words are not core
distillations of anything but rather highly variable artifacts,
constantly under pressure from the cultures surrounding them.
Number 9, that the 180 worldwide linguistic organizations now in
existence somehow provide assurance of progress in our field, is once
again close to being pure whistling in the dark past the cemetery.
As for Number 10, basically nothing more than an assertion that we
as humans are better than "all them dumb animals," I'll
leave it to you to decide how much scientific value it
possesses. Last I noticed animals were doing just fine
communicating with each other about whatever it is they need to
communicate about.
Now let's take a quick look at Numbers 1, 2, 5, and 6. No,
I'm sorry, linguistics is not the science of language, and it cannot
possibly be. Linguistics is simply not a branch of science, it
is in fact situated, along with language itself, on a far higher
plane. Language is prior to science, both historically and
logically, and to assert otherwise is to commit both a historical
error and the classic Bayesian error of overlooking one's priors.
It's six decades since Louis Hjelmslev, the principal founder of
Glossematics, made this matter amply clear. The primary science is in
fact linguistics, including language itself and even the allegedly
lowly acts of translation and interpreting between languages. All
other sciences have evolved from this original source. In other
words, language is the science of all the other sciences, which in
fact came to be derived from language as what can variously be seen
as offshoots, as branching paths, as sequels and results, perhaps
even as plug-ins, almost as afterthoughts.
Or in Giordano Bruno's words in 1603 at Oxford: (Note
5)
From Translation all science
had its offspring.
Pick up a math book, and you'll discover that many of its
seemingly abstract statements come described in carefully structured
language modules, a point I have made in my work in progress and a
point also made, I am happy to say, by Scott Montgomery in his recent
Science in Translation: Movements of Knowledge Through Cultures and Time. (Note
6) And language is of course historically
prior to mathematics as well, as far as we know language has existed
as long as our species, but most of what we term mathematics, far
from being eternal, has come into use only over the last few hundred
years, at best the last few thousand, and even then only by a small
minority among us.
In other words, all the sciences have in fact sprung out of
language. So how do you use language or linguistics as an
investigative tool, as a so-called science, when it is already the
primary creating, defining, and describing tool? Perhaps this
was what that remarkable German Georg Christoph Lichtenberg, who
somehow managed to be both a physicist and a noted satirist, meant
when he said back in 1799:
Language originated before philosophy,
and that is what is the matter with philosophy. (Note
7)
Perhaps this is also what the father of chemistry, Antoine Laurent
Lavoisier, meant when he observed in his Traité
Élémentaire de Chimie:
It is impossible to disassociate language from science or science
from language, because every natural science always involves three things:
the sequence of phenomena on which the science is based;
the abstract concepts which call these phenomena to mind;
and the words in which the concepts are expressed.
To call forth a concept a word is needed; to portray a phenomenon
a concept is needed. All three mirror one and the same reality. (Note
8)
And yet there are those in our field who suppose linguistics must
become a subclass of philosophy, when it is so obviously the reverse
that is true.
As for Number 2, that the study of grammar has anything important
to tell us about how language truly functions, this notion too is
completely false. I realize of course that I may come across as
extremely rash to make this point at any conference of linguists, and
I also want to apologize to those of you who may have devoted
large parts of your lives to the study of grammar. But just
give me till Saturday, when I'll be explaining this in great detail,
since I believe I may have identified the true organizing principle
of language—and can present evidence supporting its
status—and I will even attempt to show how this principle can be
equipped with a viable quantitative measuring device usable for
making at least some meaningful comparisons. I am happy to
report that my findings in this matter have been confirmed and to
some extent duplicated by another scholar.
Statements number 5 and 6 about Benjamin Lee Whorf contain so many
false elements that it is difficult to know where to begin. No,
we won't settle the question of free will versus determinism today,
but trained linguists and anthropologists have long observed that
most people in most cultures have never encountered problems living
within the bounds of painfully narrow linguistic and social
structures while rarely questioning the limits they impose. In
fact, it happens all the time. And to state as in Number 6 that
in all cases "translation is in fact perfectly possible" is
a totally absurd claim.
Translation is frequently less than perfectly possible, not merely
in the domain of high poetic visions but even while translating from
one allegedly precise technical vocabulary to another.
Remarkably few translations can be totally faithful to their
originals, just as remarkably few paraphrases within a single
language can be completely faithful to the text being
paraphrased. Conscientious technical translators are
continually adding footnotes, introductions, and technical addenda to
their work or may even request to discuss the meaning of the text
with the authors or editors of the original. Those who continue
to spread this misconception have obviously never bothered to consult
with a translator or translation editor, or they would know
better. In the words of Miguel de Unamuno:
An idea does not pass from one language
to another without change. (Note
9)
I believe he would have been equally correct if he had added that
ideas have frequently been known to change shape when transmitted
even within a single language.
Let me summarize this section about the quiz by pointing out three
major errors made by professional linguists over at least the past
five decades:
First, they have attempted to reify language, to turn it into a
thing, an object of study, when it is in fact a complex process that
develops within each of us over time with a slightly different
learning curve for each individual. And language is not just
any process but the primary process through which almost all
knowledge, including knowledge of its own workings, can be gained.
Second, they have attempted to impose the quite erroneous
observation that language is generic, homogenized, even universal,
when language in fact varies from people to people and culture to
culture and even within a single people or culture and can far better
be described as sui generis, multifaceted, rooted in
particular circumstances and limited contexts.
Thirdly, they have tried to impose the even more dubious
proposition that human cognition is generic, united, relatively
unvarying between and among individuals, and/or at the very least
subject to a bell-shaped curve with a predictable range of
variations. But the growing evidence about cognition favors an
markedly different outlook. The recent discovery of the
Asperger's/Autism Spectrum, along with accumulating evidence
that a comparable Bipolar Spectrum may also exist, suggests that the
totality of human cognition may not fit neatly into a bell-shaped
curve at all but occupies a far broader spectrum of conditions,
including even so-called normal cognition. If true, this could
well transform the results of much recent research into so-called
cognitive branches of learning. It also may mean that the 350
million dollars granted to MIT a few years ago by the McGovern
Foundation for research into neurolinguistics could end up being wasted.
But it's time to say something about Evidence Based Medicine and
whether it can serve as a model for building truly Evidence Based
Linguistics. If you've been to those two websites I provided in
my abstract, you already know that Evidence Based Medicine, while it
may not be perfect, is the runaway success story in today's medical
arena. Which means, to me at least, that if we could evolve an
equivalent version of Evidence Based Linguistics, we might
indeed have prepared the groundwork for some real advances in
our field. But before we reach out for any such conclusion,
let's make sure we have some under-standing of what Evidence Based
Medicine is all about.
There can be no doubt that EBM is the major medical paradigm of
the last few decades. Although it began in England in 1972 with
the work of Doctor Archie Cochrane, its major impact in this country
has been during the last ten years. If you go to amazon.com,
you'll find that some 325 books about EBM are listed, almost all of
them published since 1995. If on the other hand you go to the
NYU-Bobst Library, which is attached to a major medical teaching
college, you'll find that they have bothered to collect only about
eighty of these volumes, about a quarter of the total, which actually
turns out to be a fairly respectable percentage, especially when
compared to linguistics.
Parenthetically, let me explain what these figures mean.
About fifteen years ago university librarians publicly announced that
they could no longer go on collecting the many repetitious,
ponderously terminological, self-serving volumes that were being
published as a form of career advancement in several of the so-called
social sciences, psychology, sociology, and linguistics among
them. For instance, if you go to amazon and search out books
under the heading "constraints linguistics," you'll find 94
volumes listed. But if you go to either Bobst-NYU or the New
York Public Library, you'll discover that they have both
bothered to collect only two or three of these, in my opinion a
wise policy decision.
That's why 80 out of 325 books on EBM is actually a rather good
score, and all criticisms aside there can be no doubt that EBM is the
wave of the present and the future in most medical fields, though if
you look at the web sources I mentioned in my abstract, you will find
that it too has critics. They make the point that EBM methods
have sometimes been used to short-change patients in the medical care
they receive. It's also probable that recent reports that quite
a few well-known medications and therapies are less effective than
previously believed may spring from EBM-based research projects,
leaving it somewhat problematic, despite the excellence of the
theory, as to whether we truly know enough in all cases to set up
valid testing procedures for this new approach. But the
question remains as to whether and how something like EBM can be best
applied to the study of linguistics.
This topic could quickly become so complex as to convert my
presentation into a highly technical medical discussion, so I will
limit this section to a consideration of EBM as presented by
wikipedia.org and the Centre for Evidence Based Medicine and then
move on to how EBM principles might best be converted into a set of
Guidelines for Evidence Based Linguistics.
Here in the following extended quote is how wikipedia.org and the
Centre for EBM sum up this field:
"Evidence-based medicine is a medical movement based upon the
application of the scientific method to medical practice, including long-established
existing medical traditions not yet subjected to adequate scientific
scrutiny. According to the Centre for Evidence-Based Medicine
`Evidence-based medicine is the conscientious, explicit and judicious
use of current best evidence in making decisions about the care of
individual patients.'
"Overview
"Using techniques from science, engineering and statistics,
such as meta-analysis of the existing literature, risk-benefit
analysis and randomized controlled trials, it aims for the ideal that
all doctors should make `conscientious, explicit, and judicious use
of current best evidence' in making decisions about the care of
individual patients.
"Evidence-based medicine holds that testimonials, hearsay and
mystical arguments have little value as proof because the placebo
effect, random events and observer bias distort perceptions.
"Practicing evidence-based medicine implies not only clinical
expertise, but expertise in retrieving, interpreting, and applying
the results of scientific studies, and in communicating the risks and
benefit of different courses of action to patients.
"For all its problems, evidence-based medicine has very
successfully demoted the ex cathedra statement of the "medical
expert" to the least valid form of evidence and all
"experts" are now expected to be able to reference their
pronouncements to the relevant literature." The preceding
extended quotation ends here.
Based on these criteria for EBM, I will now attempt to derive a
set of comparable Guidelines for Evidence Based Linguistics,
springing from a set of questions about linguistics today and their
probable answers, which I believe will prove acceptable to most of us
in this room. Here are those questions:
Is scientific observation truly being applied in our
field? Here I believe the only answer arising from the
results of our quiz—and much else going on in our field—is
clearly No.
Do we find that aesthetic considerations have been imposed over
thoughtful investigation? Here I believe the answer is clearly Yes.
Have the opinions of alleged "experts" been favored over
careful investigation & the testing of hypotheses? Yes, in
so-called mainstream linguistics this is indeed the case to a
remarkable extent.
Precisely how qualified have these "experts" been in
dealing with multiple aspects of language? The answer here
would need to be "not very qualified at all."
Have mentalistic, formalistic criteria taken precedence over
physiological realities? The answer here is an emphatic yes.
Are the views of alleged authorities—even ancient authorities
long abandoned by other sciences—held to be more important than
actual observation of how language works? Once again, clearly yes.
Have data and other sources of information been actively solicited
from all groups working professionally within the language
field? No, not in the slightest.
Have specific inputs or types of information and/or data been
explicitly excluded? Yes again, they certainly have been, and
in what follows I will be able to describe only a few of these
excluded resources.
Here my evidence springs from the questions I have just asked and
the answers they have evoked, which I trust you will agree are
reasonable ones. I believe that certain conclusions spring from
these questions and answers, and provided most of you find some
justice, at least in general terms, in the preceding set of questions
and their answers, then I would like to suggest that the following
may prove workable as our Guidelines in shaping Evidence Based Linguistics.
1. In all our work we must make sure that scientific
observation is being followed as rigorously as possible.
2. We must take great care to ensure that esthetic
preferences do not take precedence over scientific conclusions based
on genuine observation.
3. We would do best to favor measurable physiological data,
where it is available, over formalistic, mentalistic suppositions.
4. We must make sure that no relevant information is
excluded from our investigations, even if it comes from hitherto
untapped sources.
5. Towards this aim we need to actively seek out the input
of professional language users, translators, interpreters, bilingual
editors, writers, and language teachers among them, even if their
testimony goes against previously accepted conventional wisdom.
6. Above all, we must carefully weigh the views of all experts and
authorities in our field, whether ancient or modern, and be prepared
to reject their views if they do not satisfy the conditions of science.
I have a fair amount more I want to say about applying these
principles in practice, especially concerning how to detect the
differences between esthetic preferences and scientific conclusions,
and I'll start doing so at my workshop session later this week.
At this point I believe I have covered the first two sections of
this presentation as summarized in my abstract. I have
explained at least in part the ten questions of my quiz, and I have
also presented what the principles of evidence based linguistics are
likely to be, with more still to come. As for the third section
of the abstract, time limitations will permit me to introduce today
only one of the many directions that linguistics might have followed
if these superstitions had not obstructed it for so long a period
from doing so.
For now I'd like to sum up those other directions as follows.
One important insight into linguistics springing from translation
studies—I believe many others are likely to arise—is
discussed in my paper published in print and soon online, Translation
as the Prototype of all Communi-cation. (Note
10) Here I describe a Gedanken-experiment
demonstrating that there is essentially no difference between the
problems of translating from one language to another and those
involved in paraphrasing the text of a single one. My
description presents credible evidence proving the truth of this
thesis. To the extent that these results are correct—and I
believe it is a great extent—there is no way that we as
scientists can discuss language unless we also discuss translation,
and vice versa.
And a dialogue between translators and linguists could be helpful
for another reason as well: so many experts in so many of the
sciences so often seem to believe that they are always and
unconditionally right in their opinions. Many translators I have
known, by contrast, dwell in a realm where they continually believe
they may be mistaken. It seems to me that the latter attitude
is a more appropriate one for the study of linguistics.
It was none other than Georges Mounin, (whose reputation and
prolific output made him the French equivalent of David Crystal) who
observed that there can be no valid theory of linguistics that does
not also provide a workable theory of translation, and since Mounin
is the only linguist who has written both a history of
linguistics and one of translation, surely his opinions and works
ought to better known among Anglophone linguists. I'll
try to tell you something more about that Gedanken-experiment
on Saturday.
In my book in progress I have also shown how a form of
differential diagnosis can be used to identify and distinguish
between a far wider range of common language processes and disorders
than the field of medical linguistics has thus far recognized.
I'll also be able to show you some of that material on Saturday.
The viable laws and measuring units for language, as well as its
possible fractal dimension, will all spring forth quite
spontaneously, though unfortunately not today, from what I will soon
be telling you about linguistic cartography.
I want to emphasize that while much of what I have so far
described springs from my own work in progress, I do not for an
instant exclude work by others that bases its findings in human
physiology and demonstrable workings of the mind. In other
words, the system I am attempting to build is not a reductionist,
exclusionary one, like a certain one I am sure you can all readily
identify, but an expandable, welcoming one, so that there is plenty
of room, say, for Sydney Lamb's Neurocognitive Linguistics (Note
11) as well as a theory being developed by Price
Caldwell, an American professor currently working in Japan. And as
time goes on, probably many others as well. (Note
12)
But let me start by describing in some detail the physiological
side of language. I'm aware that many linguists working today
cherish the belief that the properties of language are necessarily
exalted, deeply intellectual, and sublimely metalinguistic ones, so I
hope it won't come as too great a shock to anyone if I state that
language can also be seen on its most basic level as primarily a
physiological function, even (dare I say it?) a bodily function.
I could probably get away with simply being a bit poetic here and
stating that language is the musical instrument that uses the air
from the lungs to project the sounds of the stomach through the
amplifier of the larynx as modified by the keyboard of the uvula,
palate, tongue, and lips. I could become a bit coarser by
adding that language is a lot closer to the sounds of belching and
burping from the stomach and gasping and wheezing from the lungs than
it is to Aristotelian logic, Boolean logic, or synthetic syntactic
restructuring. But that's still not coarse enough for me.
Let me explain quite clearly—perhaps a bit too clearly for
some—what I mean when I call language a bodily funtion.
Not just human beings, not just all animals, but even plants as
well—in other words, all of life itself—relies on one (and
truly just one) clever trick to maintain itself: the exchange of
gases and fluids, sometimes with the odd bit of so-called solid
matter thrown in. This is true of breathing in and out, of
ingestion and excretion, even of the exchange of sexual fluids. If we
start from this primitive level and work our way forward from there,
not only will we be able to see how language fits into this process,
but we may also gain some unexpected insights into the field of
evolutionary linguistics, a rapidly expanding growth industry among
us that has encouraged specialists from many fields but almost no
professional language users to hold forth on a great variety of
theories. The rule encapsulated by a German proverb still holds true:
Jeder Mensch glaubt, weil er spricht, über die Sprache
sprechen zu dürfen.
Which translates as: Every human being believes that because
he speaks he is also entitled to speak about language.
One step upward from the stage of merely exchanging gases and
fluids between organisms would be to add a distinctive chemical that
other organisms of the same type might recognize while at the same
time repelling certain different organisms. Already a form of message
is being conveyed. One step beyond that would be to employ a
chemical that other life forms could recognize, so as to attract or
antagonize them. The message has become a bit more complex. The
next step, well-known among biologists, would be to make this
exchange even more noticeable by adding a chemical aroma.
This process is of course well-known as scent markings. A
parallel step, taken by many species, eventually including our own,
would be to add what I call sound markings.
On my website you will find three and soon four different pieces
describing the relationship between scent markings and sound
markings. (Note
13) In a very real sense all of human
language is nothing more or less than a highly elaborated form of
sound markings. Where other animals spray the scent markings on
their environment, we spray sound markings on or in the direction of
almost everything around us. I am spraying sound markings at
you right now, and soon you will have a chance to spray a few of your
own sound markings back at me and at everyone else. And we carry out
this process just as insistently and just as unstoppably as any pussy
cat hard at work marking our carpets, our armchairs, and our trouser
legs as its own.
From at least one aspect, human language can be viewed as merely
an advanced form of spray, sometimes even spewing or sputtering.
If we have failed to notice this process before now, one reason is
likely to be pretty obvious: our abiding cultural reluctance
bequeathed to us by Puritanism cum Victorianism to look at—much
less discuss—how our bodies really work. Or perhaps it's
merely squeamishness—well, I can assure you that you can't learn
anything useful about medicine or human physiology without finally
abandoning almost every last vestige of squeamishness.
(a free copy of this program can also be obtained on this site
by clicking here)
Just as marine life has been observed repeating evolutionary
phases in the seas around us, so we can observe various life forms
continually recapitulating various forms of communication around
us. What further evidence do we need that human language and
sound markings come close to being synonymous? Can anyone state
with absolute certainty that what we do with our sound markings is
all that different from what sea lions do with their barks, crickets
with their chirps, or bees with their cute little dances?
(Though bees are also known to employ scent markings to convey their messages.)
At some point, possibly among higher primates, an evolutionary
step was clearly taken to favor sound markings and gradually phase
out scent markings. I've been in contact with Jane Goodall
about this topic, and she tells me that even chimpanzees rarely use
odor as a major form of communication, though male chimps do give off
a strong scent at mating time. (Note
14)
Which could also provide an explanation for the near simultaneous
surging of our language skills along with the gradual atrophy of our
sense of smell as compared with other animals.
If this finding holds up for other hominid apes, it means that
both humans and their closest relatives began to give up scent
markings at a fairly early stage in their evolution. This would
mean that the period when higher primates began to lose the vast
animal sensorium of aromas coincided with the period when humans
first began to develop language. But vast numbers of other
animals still use scent markings, and the precise
evolutionary—much less historical—events involved, assuming
they occurred as described, are of course still unknown. For
instance, Frans de Waal has recorded instances of Bonobo chimps
recognizing each other through their scent after long absences. (Note
15)
Why do I keep insisting on this connection between scent markings
and sound markings, why have I stressed the common thread of an
exchange of gases and fluids through a whole phylogeny of plant
and animal species? Because these gases and fluids unavoidably
furnish the basic substance of all human communication, the carrier
wave so to speak. I've been told that the following point is so
obvious that perhaps I shouldn't be making it at all, but I still
believe it needs spelling out. So if you want to flinch at what
follows, you can start flinching now.
If anyone truly requires compelling evidence for the gaseous
content of language, let them simply hold their hand in front of
their mouths while they speak. If they need evidence for the
fluid content of language, let them keep that hand in place a bit
longer, say for about two minutes, while they count in a whisper from
1 to 200, after which they will discover a small damp patch in the
very center of the palms of their hands. Quod erat
demon-strandum. You can prove it to yourselves right now while
I am talking, provided you keep it to a whisper. On its most
basic level language is composed of gas and fluids. With
sometimes a further embrocation of thought matter.
And as soon as we grant language recognition as a physical
process, we also open the door to the possibility that language in
itself, like all other physical processes, may be open to various
irregularities in growth and development and even various pathologies
not normally studied under language handicaps.
Here I believe a further disclaimer is definitely in order, since
I have now further opened the door to a number of topics not
currently covered under linguistics. As I have written
elsewhere, I do not mean for an instant to debase human speech to
nothing more than animal sprayings nor to suggest that language does
not also possess other more abstract properties. But would not
such an evolution explain much about how human beings still use
language today? Do we really require "scientific" evidence
for such an assertion, when so many proofs lie self-evidently all
around us? One proof is that normally human beings do not use their
nether glands to spray a fine scent on their environment. They do,
however, undeniably talk at and about everything, real or imagined.
Sometimes the volume of spray can be so pronounced that it becomes
inadvisable to stand too close to the person speaking.
And as I have also written elsewhere, would not such an evolution
also aptly explain the attitudes of many "literal-minded"
people, who insist on a single interpretation of specific words, even
when it is carefully explained to them that this is mistaken?
Does it not suggest at least one possible reason why many
misunderstandings fester into outright conflicts, even physical
confrontations? Assuming the roots of language lie at least
partially in territoriality, would this not also go some distance
towards clarifying some of the causes of border disputes, even of
wars? Perhaps most important of all, does such a development not
provide a physiological basis for some of the differences between
languages, which themselves have become secondary causes in
separating peoples? Would it not also permit us to see different
languages as exclusive and proprietary techniques of spraying,
according to different "nozzle apertures,"
"colors," or viscosity of spray? Could it conceivably shed
some light on the fanaticism of various forms of religious,
political, or social fundamentalisms? Might it even explain the
bitterness of some scholarly feuding?
Perhaps most important, at a time when the majority of Americans
has apparently decided that evolution is no more than a theory, does
this discovery that human language is clearly allied to animal scent
markings not present important new evidence that evolution must in
fact be true? In addition to offending today's puritan
attitudes, does it not at least for some evoke much the same sense of
shock that our ancestors felt a century ago when told that humans
might have at least something in common with monkeys? It was
the hope of linguists like Whorf, Sapir, and the younger Hayakawa
that our field of study could provide a continual flow of insights
for our citizens about the values we live by. Wouldn't it make
a great deal more sense if leaders of our profession, instead of
wasting their time on warmed-over cold war causes, were sharing this
important truth about evolution with the general public?
I want to make it clear that even before today, one or two of you
have taken pains to inform me that this truth is simply too obvious
to bear such repeated emphasis as I have given it. I don't want
to sound stubborn, but I can't help wondering if such a criticism may
not be prompted by a barely concealed form of theism holding that we
as humans can occupy only a sublimely elevated station, just one
degree below the angels, or at least something intellectually quite
comparable, and that our form of communication must also be equally
exalted in character. I find no evidence for this view at all,
and I wonder if the seemingly perpetual debate carried on since 1984
first by the Language Origins Society and now by Evolang may be a
prolonged attempt to avoid the obvious connection between language
and evolution, simply that human language IS evolution in action,
through its transformation from scent markings to sound markings.
In other words, language is both wet and soft, just as we
ourselves as human beings can also be best described as wet and soft,
as what is sometimes called wetware. And being as we are both wet and
soft, since I'm hearing this term "hard science linguistics"
at this conference, I'm not entirely sure how well something
relatively hard and presumably dry is going to is going to solve very
many linguistic problems. I agree completely with Prof. Yngve's basic
premises that grammar has been a false start, leading us into
centuries and even millenia of failed linguistics, and obliging us to
find a new direction. But I doubt if that direction can emerge
from ever more rigorous logic-chopping and sentence-diagramming,
rather I see its origins in the many new insights we may gain by
leaving logic, mathematics, and philosophy behind and returning
language to its birthplace as a branch of human physiology.
In other words, we can turn linguistics into a hard-ER science,
but never a fully hard one, or at least no harder than physiology and
biology themselves. Which is where language ultimately belongs,
and not in all those other domains of knowledge where so many in our
field are trying to enshrine it. And in the totality of its
diachronic and social development, I believe language belongs under
the general heading of natural history. I once tried to design a
museum exhibit about translation for a natural history
museum—parts of it are up on my website (Note
16)—and I'm sure I could design an
appropriate natural history exhibit for language and linguistics as well.
Well, I'm at least headed in the direction of an ending now.
I hope at least some of what I've presented has made more than a
fleeting impression on you, since I'm afraid the best I've been able
to do today is to provide you with a few glances at the rather large
number of aspects of language I've been trying to examine.
With more to come soon. Quite simply, I've been trying to
cover more bases than a system that deals mainly with networks or
phrase parsing or syntactic juggling. This is crucial since
language lies so squarely at the foundation of so many of our
sciences. In fact I'm fairly sure that there must be a number
of bases I've left out entirely or that supplement my own objectives,
and since I'm not building a reductionist, exclusionary system I more
than welcome their introduction and inclusion.
Perhaps the main point I've been trying to get across is how
extraordinarily complex and multi-faceted language truly is, and one
further way I'd like to stress this is through a slightly adapted
version of a famous nineteenth century American poem, which you may
have encountered before. I've only needed to change about four
words to make it fit rather aptly into today's world of
linguistics. Here it is:
The Blind Linguists and the Elephant
Six gurus wise of Linguastan
To learning much inclined,
Went forth to see the Elephant
(Though all of them were blind),
That each by observation
Might satisfy his mind.
The First approached the Elephant,
And happening to fall
Against his broad and sturdy
side,
At once began to bawl:
"God bless me! but the
Elephant
Is very like a wall!"
The Second, feeling of the
tusk,
Cried, "Ho! what have we
here?
So very round and smooth and
sharp?
To me 'tis mighty clear
This wonder of an Elephant
Is very like a spear!"
The Third approached the
animal,
And happening to take
The squirming trunk within his hands,
Thus boldly up and spake:
I see," quoth he, "the Elephant
Is very like a snake!"
The Fourth reached out an eager hand,
And felt about the knee.
"What most this wondrous beast is like
Is mighty plain," quoth he;
"'Tis clear enough the
Elephant
Is very like a tree!"
The Fifth who chanced to touch the ear,
Said: "E'en the blindest man
Can tell what this resembles
most;
Deny the fact who can,
This marvel of an Elephant
Is very like a fan!"
The Sixth no sooner had
begun
About the beast to grope,
Than, seizing on the swinging
tail
That fell within his scope,
"I see," quoth he, "the
Elephant
Is very like a rope!
These hoary seers of
Linguastan
Disputed loud and long,
Each in his own opinion
Exceeding stiff and
strong,
Though each was partly in the right
And all were in the wrong!
Envoi:
So oft in wild linguistics
wars,
Most disputants we have seen,
Rail on in utter ignorance
Of what each other mean,
And prate about an Elephant
Not one of them has seen!
This version is close to the text by the American poet John
Godfrey Saxe, which was itself based on a Han Dynasty Chinese
poem. A century ago it was believed that these verses shed
light on the various claims of conflicting religious faiths.
But it requires only the slight degree of adaptation found above for
the poem to work equally well for the so-called science of linguistics.
I'll be quite pleased if I've succeeded today in merely suggesting
to you how many sides to the elephant there just might be.
Today I'm afraid I've only been able to concentrate on the bodily
fluids of the elephant as an exercise in physiology, which I hope no
one has found too upsetting. On Saturday I hope to get around
to the elephant's skin surface as seen through cartography, to
the differential diagnosis of the elephant in health and
disease, its various fractal extensions, what elephants look like in
other languages, and a few other aspects, including what I call the
Six Laws of Language and Linguistics in Draft Form, the basic laws
that keep the elephant going. And I'll give all of you who
attend that session a copy of those laws to take home, treasure, and
criticize. And in all those cases I will do my best to present
the evidence that stands behind each one of these various
aspects. The one aspect I will definitely not be addressing is
the syntactic structure of the elephant, simply because syntactic
structure is almost totally useless in understanding either elephants
or languages.
One question of course piques my curiosity: Is there any hope that
such complexity can ever be integrated into existing departments of
linguistics at our universities? I'm a bit skeptical that this
can ever become possible, at least not without substantial
reorganization of the university. More to the point, one
delightful citation, which I first found attributed to Freud but is
more likely to be an urban folk saying among college administrators,
makes the following observation:
Reorganizing the university is like rearranging
the graves in a cemetery. (Note
17)
In any case, I cannot stand still waiting for such a
reorganization to happen or not happen. In the meantime I'm
pleased to tell you that I have been invited by an international
group to speak on the topic Practical Linguistics for Translators
(And for the General Public) some time later this year or early next
year. So however many schools of linguistics we may now
have—or perhaps how few—this just might be a new one.
While I was in college and over several decades before and after,
the back cover of the journal The American Scholar in all its issues
always featured prominently Ralph Waldo Emerson's five-word
definition of what a scholar is, of which only two words comprise the
actual definition: "The Scholar Is Man Thinking."
Yes, I know, sexist as hell, but if we just change one word and say
"The Scholar Is Humanity Thinking," I believe it still
works admirably well. In other words, the scholar is someone
thinking for and on behalf of all humanity.
Today I fear that we might need to revise that definition, and not
merely in linguistics, and place the following text on the back
covers of our journals: "The Scholar is only interested in
communicating with other Scholars for the purpose of advancing his
own theories and/or discussing other similar theories of little
interest outside his profession and in enhancing his own status
within that profession."
I suppose this could change, and I'd truly like to believe that it
can change, and that your having invited me here today might be one
small sign of such a change. But I'm not the least bit sanguine
that this can happen, at best I'm guardedly optimistic. The
gargantuan pressure from the sheer dead weight of false ideas
blocking such a change, both in linguistics and probably other areas
as well, was perhaps best summed up almost two thousand years ago by
the neo-Platonist philosopher Epictetus when he said:
