to home
to summary of reasons

Forty-Four Reasons
Why the Chomskians Are Mistaken

Reason 11

11. The continuing inadequacy—often approaching abject failure—of the one practical application that ought all along to have been a showcase for the validity of the TGG approach: so-called "machine translation." Just recently the heir presumptive for TGG doctrines went on record in the New York Times that the output of "translation engines" was definitely improving. This is certainly news not only to professional translators but also to many of those who have tried out machine translation applications on any extended basis. Outside of a few rigidly defined areas, MT contains just as many problem areas as it did several decades ago, as this author made amply clear at the panel he chaired during the Machine Translation Summit III conference in Washington (also documented on this website).

TGG theorists suppose that conquest of MT remains a slow and steady advance of their forces across an even, unobstructed terrain. In reality, they have been fighting a desperate and continually losing uphill battle for over fifty years against an enemy hidden and well-entrenched in mountain fortresses they haven't even discovered exist yet. The growing acceptance for TM (Translation Memory) over MT—though even here only for severely restricted subject matters—represents only the latest defeat for their forces.

The continued enthusiasm over MT may be more the result of the continued availability of funding for MT research rather than any deep or dramatic progress in this field. And the availability of that funding may be entirely in the hands of a few bureaucrats and engineers who understand little of linguistics but are acting under the advice of these very TGG linguists. If so, then we may be dealing with something like a vicious circle of ignorant boasts begetting substandard results begetting further boasts and claims that more funding is needed. While Chomsky himself has made the odd declaration doubting the feasibility of MT, the collected mass of his prose largely suggests the contrary, and it is scarcely surprising that most MT workers revere his work as both ground-breaking and inspirational.

Several of the preceding points can be summed up at this juncture as follows: essentially, the followers of TGG are living in a dream world where a number of truly remarkable discoveries are just about to be discovered and a number of final resolutions are certain to be resolved. Caught up in this dream world, they naively but passionately continue to believe:

that some miraculous new scholarship will prove once and for all the existence of a "universal grammar;"

that some miraculous new psychological finding will demonstrate the "poverty of stimulus" doctrine for all time to come;

that some miraculous new medical breakthrough will reveal the existence of a hitherto unnoticed organ in the body: "Pinker's Plexus" or "the Innateness Gland."

that some hacker in some garage somewhere will all at once hit upon the key permitting perfect and instantaneous machine translation in all directions between unlimited texts in all languages, whether living, dead, or yet to arise.

to the next reason

to the previous reason

to the summary of reasons

COPYRIGHT STATEMENT:
This piece is Copyright © 2000
by Alexander Gross, with specified
portions Copyright © 2000 by
Sergio Navega. It may be
reproduced for individuals and for
educational purposes only. It may
not be used for any commercial (i.e.,
money-making) purpose without
written permission from the authors.
All Rights Reserved.

to top
to summary of reasons
to linguistics menu
to home