The Parable of the Man Who Didn't Get the Message

Dec 13 2010 Published by under From the Melodye Files

May I present to you, dear readers, a reading from the Book of Revelations...

Many years ago, the king of Gath called on his scholars, and asked them: "Tell me, what maketh man wise?"

The scholars conjured up many a fine parable, but none of them satisfied the king, who said unto them, "Scholars, what you have told me would help me know a wise man when I saw him, but this I knew already.  I ask you again, what maketh man wise?"

The kings' scholars were much perplexed, until one of them ejaculated, "I know!  We must study the owl, and discover what makes it wise!"

So the scholars captured an owl, and slowly, with the help of an abacus, they devised stories to explain its wisdom.

These stories did not please everyone, however. One day, a particularly clever man asked, "Is a man an owl?"

To which the other scholars replied, "No."

"Then," said the clever man, "you are wasting your time."

Seeing this, the scholars were most vexed, and they rent the leather patches from the elbows of their robes.

"Now now," said the clever man, "Fret not.  Know we not that what maketh man wise is his being a man?!"

"But this will not please the king!" wailed the scholars.

"Oh," said the clever man, "then we must befuddle him with mysticism and bullshit."

The scholars saw that the clever man was right, and they proceeded to befuddle the king.  This worked so well that the clever man was elected to the head of the academy, and given a tablet of his own.

However, one day a stranger came into the town. "Tell me," the king beseeched him, "What maketh man wise?"

The stranger decided to consult the tablets, and found the abacus, and the story of the owl.  He saw that if he used the abacus wisely, he could make new stories, and they would explain what maketh man wise.  And if other men used the abacus, they could see this too.

The stranger was excited, and he sent his findings to the tablets.  When the scholars heard of this, it caused them great anguish - for what if the king were to discover that they had been befuddling him with mysticism and bullshit?

So they beseeched the clever man for help. "Here is what you must do," he told them, "When the keepers of the tablets consult you about the stranger's stories, you must pretend that you have never heard of the abacus!"

And so when the keepers of the tablets consulted the scholars about the stranger's stories, the scholars pretended as they had been told.

When the stranger heard of this, he showed the keepers of the tablets the abacus, and they were much perplexed. "Truly there is an abacus," they said, "and there is the story of the owl. But the scholars cannot see it. What does this mean?"

The keepers of the tablets decided that they should confer at length, and they advised the stranger that in the meantime, he should live in a box.

Here endeth the reading.

(For hints, click the links.)

9 responses so far

  • HAHAHAHAHA. I do love science fables.

  • James Davis says:

    So, let me see if I get this straight. The language acquisition folks decided that animal models weren't working to well so they abandoned them in a push led by Chomsky, who made up some BS to replace the animal models. Chomsky ended up as head of the NAS (um, did this actually happen) and got his own journal (once again, I don't think I've heard of the 'Chomsky' journal, but I concede that I could be confused here).

    Then, when Plum proposed returning to modifications of the animal models, Chomsky advised linguists to pretend they hadn't heard of the animal models before. (This seems like a pretty nasty accusation.) And Nature can't figure out why linguists haven't heard of something.

    This seems to be pretty a pretty damning accusation against Chomsky and Language Acquisition research in general. Am I misunderstanding?

    (I never was good at fables.)

    • melodye says:

      I would adopt a slightly less-literal approach to the parable. I didn't mean for the links to be taken as stand-ins, they're just hints in case you don't already know the story. I should probably change the first line in the piece, since it's really not as specific as this interpretation suggests. I just liked the sound of a "Reading from the Book of..." (I was raised Catholic, after all.)

      Some clarifications:

      1. Prof Plum isn't the only "stranger" out there and he doesn't really live in a box (thank goodness). He's simply a useful figurehead for a whole group of scientists who have been steered rather forcibly away from doing learning-theoretic research in language. It's next to impossible to publish work in that vein in top science journals because of the hold Chomskian-acolytes have on the field. If you want to make an illustrious career in psychology, doing research into language, the message is simple: stick to the nativist story. (Obviously there are other options : you might do 'statistical learning' which is largely model free and answers interesting questions, but none of the hard problems; or do connectionist modeling, which is an excellent idea in principle and often badly applied (and subject to many completely valid nativist critiques); you might be a Bayesian fundamentalist and tell the nativists they were right all along; or a Whorfian, and only symbolically step on anyone's toes. There are options and intriguing work to be done in the middle -- but in terms of establishing a grounding, it's clear where my prejudices lie.)

      2. As you noted, Chomsky doesn't really run the NAS and doesn't have his own journal; he does exert massive control over the field of language acquisition, and has used his influence to more or less halt learning-theoretic research into language. If you look at NAS membership, it's actually surprising how many members in language have been part of nativist research programs (it's nearly all of them).

      3. W/r Nature & Science -- neither of them have published any research on learning or information theoretic approaches to language in at least two decades (and I haven't searched prior to then). In fact, most researchers who still do work in this vein are relegated to Behaviorist journals that have impact factors of less than <1. I don't think this is necessarily the fault of the journals. It's a problem with the system. Unless editors decide to outright ignore the comments of powerful and politically-motivated reviewers, research like this will continue to be censored by peer review.

      But now this is becoming far less amusing, and you're a linguist -- shouldn't you be an expert at reading between the lines? ;) I don't mean this piece in particular as an indictment; I've written a series on the politics of ideas in the field and have often touched on what I see as logical flaws in the nativist position: see e.g., here, here and here.

      The question is, is this the message the field should be sending its young?

      • It's interesting that you feel nativism has such a strong hold. In my experience nativists are a minority and Chomskian nativists have nearly disappeared. Unless you are calling all the Construction Grammar folk and Statistical Learning folk nativists. I doubt they'd agree.

        • melodye says:

          Whether or not the majority of cognitive scientists doing work in language would self-describe as 'Chomskian nativists,' per se, is really not at issue. For example, Pinker probably would not describe himself in that way, but his research program over the past several decades owes nearly everything to the arguments made by nativists in the late 60's and early 70's. If you look at the basic assumptions of many in the nativist program (e.g., category learning via induction; competence/performance distinction; generative grammar; productivity; and so on and so forth), all of these ideas come out of the Chomskian program. I'm not saying that contemporary nativists are homogenous in their beliefs and research questions; what I am saying is that many of the ideas Chomsky articulated (including his logical arguments and criticisms of learning theory) are now implicitly accepted by nearly everyone, *including* researchers in, for example, statistical learning or connectionism. You don't have to believe in a universal syntax to have been significantly influenced by Chomsky, I can tell you that much.

          For example 'statistical learning' approaches trade in all the modeling power and precision of learning theory for the somewhat vague notion that children can track probabilities. If you read FLO, we talk extensively about why statistical learning (as usually construed) would fail to capture how children learn verbal categories. A simple, probabilistic model would never capture how children distinguish a fox from a dog, for example. While there has definitely been some interesting work done in 'statistical learning,' the approach, as a whole, dramatically undersells human learning capabilities. This is particularly weird, given that learning models are widely used in the behavioral neurosciences (e.g., in modeling decision-making, executive control, addiction, mental illness and so on). But we don't use our most powerful, neurally-supported tool for investigating the workings of language. Why?

          The only reason I can grasp, is because many scientists (nativists and non-nativists alike) still buy into an outdated critique of learning theory, which many at the top of the field have staked their careers on selling.

  • "Shecky R." says:

    ...but did anyone live happily-ever-after?

    I gave up on the nativist vs. learning theory debates long ago, ultimately no longer believing the human brain can successfully decipher itself (though it can easily fool itself into thinking it can). I still love to read cognitive psych. (finding much of it fascinating), but doubt that many of the questions therein can be resolved with any certainty.

    • melodye says:

      There's actually a great group on Facebook, "I picked a major I liked, and one day I'll be living in a box." I'm not sure if boxes are happy endings. Perhaps--?

      Fwiw -- contemporary learning theoretic approaches aren't about resolving anything with certainty. They're about discovering the extent to which general learning mechanisms can explain language acquisition empirically -- i.e., experiment by experiment -- using learning rules with a firm neurobiological grounding. That seems to be the way the scientific method should work; as far as I'm concerned, the logical 'certainty' Chomsky claims is pseudoscience.

  • LadyLobo says:

    A couple years ago I heard Chomsky give a talk about the history of the Phil and Linguistics dept at MIT. During the question period someone stood up and asked him a question related to animal language(/communication?) . Unfortunately, I no longer remember the wording of the question. What I do remember is the dismissive reaction he received by Chomsky to the effect of: animals do not have language; your question is irrelevant. I was an undergrad junior at the time. I couldn't believe scientists would speak to each other in such a manner.

Leave a Reply

Bad Behavior has blocked 393 access attempts in the last 7 days.