Get real, George

[Tom Givon]

                        1-7-97

Dear FUNK people,

I was going to hold my peace on the parser issue, which I suspect has
provided all of you with much merriment, having noted that Matt
Dryer basically got it right: Neither the extremist position of
the Syntax-for-the-love-of-syntax Chomskyites, nor the extremist position
of syntax-doesn't-exist it's-all-lexicon-and-discourse Functionalists
are really consonants with the facts of language as we all know them.
In other words: Yes, Virginia, there is syntax.
                But no, Virginia, it is not there
                as an autonomous, non-adaptive flower
                of a genetic megamutation.
Honest, I was going to bite my tongue and leave be this time -- till I
read George Lakoff condescending, gratuitous intervention.

Even then, I would have still preferred that someone else do the honors.
But then it dawned on me that perhaps I am in the best position than most
to call George's bluff. You see, I've been watching George in action for
thirty years now (La Jolla, Spring of 1967, "Is deep structure necessary?"
boy the years sure fly). And over the years I have seen George get away
with similasr fetes by remarkably similar means. That is, by insinuating
that somehow the rest of y'all ignorant slobs better join his elect group
of with-it cognoscienti or else you'll miss the (latest) boat.

So let us see what it would means to the many of us who have been looking
at grammar/syntax from a variety of perspectives if it now turned out tha
syntax does not really exist. And that at any rate, it plays no functional
role in mediating between the cognitive-lexical- communicative levels and
the phonetic output. To wit:

1. GRAMMATICALIZATION AND SYNTACTIC CHANGE: We have been describing how
pre-grammar or non-grammar (lexicon cum parataxis) changes into grammar
(syntax, morphology), how lexical items grammaticalized into morphology,
how clauses that used to come under separate intonation contours somehow
condense themselves under a single contour. We thought we were studying
something real, poring over successions of older texts in arcane languages,
wretling with internal reconstruction in unwritten languages. But we've
been deluding ourselves all along -- says George. Forget it, good non-with-it
FUNK folks. It's not really there. And if it is, no matter, it DOES NOTHING.

2. ACQUISITION: We have been describing how children move gradually
from pre-syntactic (pre-grammatical) communication to grammaticalized
communication, adding a morpheme here, an embedded construction there;
stabilizing rigid grammatical(ized) word-order where previously only
semantically-based (AGT oriented) or pragmatically-based (TOPIC oriented)
word-order could be found; gaining embedded constructions, gaining de-
transitive clauses, gaining fancy subject-inversion with auxiliaries;
etc. etc. Forget it, folks, -- says George. It's been all for naught,
what you've beel studying is plainly a mirage. And just in case it did
exist, it doesn't matter. Because -- it turns out -- it serves no function
whatever. "We", the cognoscienti, have already "demonstrated" that we
"can do it all" without grammar.

Nice try, George. But how come the kids are still insisting on doing
what they're doing? Are they, like us, deluded too?

3. PIDGINS AND CREOLES: We have been observing for a long time the
peculiar consequences of having no morpho-syntax; that is, of pre-
grammatical (pidgin) communication. The halting, repetitious, error-
prone, frustrating communicative mode of pidgin is familiar to many
of us from early childhood studies. Doesn't Sue Erwin-Tripp work at UC
Berkeley? Doesn't Dan Slobin? Isn't Ron Scollon's dissertation still
available? Or Liz Bates? Or Eli Ochs' early works? Or Developmental
Pragmatics (1979)? How about the extensive literature on Broca's Aphasia
communication? Lise Menn's/Loraine Obler's magnificent 3-volume collection?
And isn't the data of second language pidgin available? What is it that
makes grammaticalized language -- such as the Creoles of children of
Pidgin-speaking parents -- so much more fluent, fast-flowing, streamlined?
What has been subtracted between the Creole and the Pidgin? According to
George, nothing. But just in case it was something, forget it too. It
serves no purpose. "We" can do without it. Well, as a person who have
gone through the agony of moving from pidgin to grammaticalized commu-
nication five distinct times (and hated it with passion every time...),
I'd like to know why nobody had ever told me that I was wasting my
precious time? That I was much better off mapping directly from cognitive
to phonetic structure? Might have saved me years of toil and agony.

4. CROSS-LANGUAGE TYPOLOGICAL VARIATION: We have been observing how
the very same cognitive-semantic-communicative function can be executed
in different languages by a (relatrively small, mind you) number of
syntactically distinct constructions. We've also noted that those cons-
tructions represent distinct diachronic pathways of grammaticalization.
We've seen this with complementation, with relative clauses, with passives,
inverses, anti-passives, clause-chaining types, tense-aspect-modal systems,
negation -- you bloody name it, we've been observing it. But, sorry you
simpleton FUNK folks. George says that -- it turns out -- what y'all been
documenting so laboriously is -- right, folks -- nothing. And if indeed
it is still something, nevermind; because you see, it is there for no
purpose whatever. "We" can do without it.

5. DISCOUSE: Here's the real bad news, folks. You've been studying for
30 years now how syntactic/grammatical constructions have specific
communicative function paired with them, systematically, intimately. What
a cxolossal waste of energy. You see --it turns out, George says -- that
morpho-syntax doesn't really exist. What you should have been really
studying all along, it turns out (you hear this -- Wally? Liz? Sandy? Russ?
Jack? Barbara? Matt? John? Bob? Anna?) is how communicative function maps
directly onto phonetic structure. Directly folks, directly.

6. COGNITIVE PSYCHOLINGUISTICS: Forget it Russ Tomlin, sorry, Morti
Gernsbacher, your loss, Brian MacWhinney, butt of, Walter Kitsch,
you're past, Tony Sanford, get lost, Liz Bates. All your labor has
been for naught. George has decreed your experiments null and void,
whatever it is your studying just doesn't exist. Or worse, you
simpleton fukn-folks, it is there for no purpose.

6. NEUROLOGY: We all know localization is a complicated, that grammar
in adults is distributed across many "modules". Sure, the modules bear
little resemblance to their Jerry Fodor name-sakes. They interact, they
talk to each other, they are NOT encapsulated, they collaborate with
"cognitive" modules (attention, activation, memory, intention, pragmatic
zooms, etc.). But however widely distributed, portions of this complex
mechanism can be knocked out selectively by lesions. What is it, George,
that aphasics have lost, exactly? You study their transcribed discourse
(Menn and Obler, eds 1990, e.g.), and you notice that
  (i) the lexicon is there. nouns, verbs, adjectives.
  (ii) the coherence of discourse is still there (referential coherence,
       temporal coherence, all the measurables).
So what is it that is NOT there? We used to think it was morphology and
syntactic constructions. But  it turns out, George now says, no go.
Whatever we thought it was is not really there, never really was to begin
with. And if for some reason it turns out it was, still no matter;
it performed no function. So, one wonders, how come the poor slobs in the
wards are having such a hard time stringing lexical items together into
clauses and clauses into discourse?

The most gratuitous insult, I must confess, is George's reference to "neural
networks" and connectionism. This is a rather poor substitute for real
neurology, which is vast, complex, frustrating and cannot be practiced by
erzatz experts in search of the latest fad. When I ask the real neurologists
I know what they think of connectionism, I get an incomprehension response.
Never heard of it. Con what? So the exhortation to join the bandwagon before
it leaves the station and we're stranded for good rings rather hollow.
Especially this undignified business of "in real time". So, if nobody else
has yet, I guess I must tell George that the "thing" that makes it possible
for humans to process language at the rate of, roughly, 250 msecs per word
and 1-3 seconds per idea (clause, proposition, event/state frame, intonation
unit...) is called grammar/syntax.

7. EVOLUTION: I have saved this one for last, since in some funny way it
remains the crux of the matter. Here's the real puzzle: Why should this
manifestly existing, acquired, diachronically-changing, cognitively manipou-
lable, neurologically-based entity called "grammar/syntax", with its
complex, imperfect but nonetheless clearly manifest ICONICITY to semantic
and pragmatic function(s) -- why should it ever evolve? With its marvelous
hierarchic design, parts fitting into larger parts; why should this
extravaganza ever evolve in the first place? According to George, the
extravagance doesn't really exist. Presumably then, there's no difference
between lexicon and morphology (forget what you've been observing, you gramma-
ticalization hounds); no difference between parataxis and syntax (again, forget
your puny facts you poor un-enlightened souls, you un-cogged simpletons).
But, just in case it did exist -- nevermind. It's there for no reason,
it just happened to evolve, somehow, for the love of God or Descartes.

For those of us with ears that are yet undulled by the clamor to be
with-it, the strage ghost deja-vu is now creeping in: Hey, but this is
what Chomsky has been saying all along about language evolution -- that
it is a mysterious saltation, unguided by adaptive (communicative, cognitive)
behaviour.

I have probably said too much already. I always live to regret getting
involved in these silly affairs. But I think if there is one lesson to
be learned from this, it is perhaps that nothing comes cheap in real
science. You can't do complex biologically-based science on the fly.
If one intends to talk to only God or yourself or the Elect, that's
certainly one's privilege. But if you want to be taken seriously, well
get serious first. Like, get real.

Happy New Year y'all, TG