Let's put all the comments I've posted here on the Fiona Cowie book together, to see if I managed to write a coherent review piecemeal. That'll be fun. I'll keep the transitional changes as slight as possible.
---------
FIONA COWIE, WHAT'S WITHIN (1999)
So I've bumbled on a discovery. I've discovered that a certain 18 year old book seems to be important to controversies that are in turn important to me.
This is a contribution to the old debate between rationalism and empiricism. Cowie says that empiricism, with its blank-slate mind filled by experience (or, as behaviorists came to say, by conditioning) was regnant in the Anglo-American world in the late 1940s. This was the era of Skinner's rise to prominence. It was also an era when a lot of ideas seemed to have been discredited by the recent war, by having a Teutonic sound to them, and innate ideation was a casualty.
Later, Chomsky and Fodor turned the tide: Chomsky as to language skills, Fodor as to ideas proper. Cowie takes them both to task, and wants her book to be a contribution to another such turn.
As I expected she would, Cowie charges that both of her Ur-nativists have a difficult relationship with the theory of evolution by natural selection. She thinks that each of them owes us a theory of how the capacity for the self-conscious development of language or ideas developed over time, amongst a certain set of primates. If they are in any important non-tautological sense innate then they somehow GOT to be innate, and unless either thinker is willing to make a clean break with Darwin (which has not been forthcoming) there ought to be at least a schematic explanation of how that happened, how one particular species acquired this equipment which is now pre-installed in each of its tokens.
And if either does want to make a clean break with Darwin, that needs explanation too and (since both are naturalists) sky-hooks can't be part of the alternative account.
Cowie notes somewhat drily that Plato is a rare example of an innatist who DID give an explanation of how the innate stuff got inside the head: metempsychosis! His followers have not been as helpful.
The book, which began life as a doctoral dissertation, is a look at Fodor and Chomsky in particular, and their revival of elements of classical rationalism, especially with regard to a priori ideas.
For the next few paragraphs I'll focus on an early point, within Cowie's historical discussion. She asks: what exactly was the classical debate, the one featuring Descartes and Leibniz on one side, figures such as Hobbes and Locke on the other, really about? It is not easy to pin in down. The empiricists sometimes wrote as if the rationalists were saying that a baby comes into the world already knowing, say, what a triangle is, or what constitutes a prime number. Babies clearly don't. And the old rationalists denied this is what they meant.
Descartes wrote that he and others on his side of the dispute "simply mean that we have within ourselves the faculty of summoning up the idea" of a triangle, a prime, or a Perfect Being.
On the other hand Locke etc. weren't denying the tautology that we are capable of summoning up the ideas that we really have. We were then born with the possibility of knowing whatever we later came to know. That even stubborn Anglophones don't deny. Is that all the "faculty" within means?
Cowie argues for what she calls a "special faculties" interpretation of the dispute. What Descartes meant was that the acquisition of some of our ideas, particularly our geometrical ideas, were due to the activation of certain features in our minds. As is well known, he thought of "mind" in an intangible sense here, he wouldn't have said, "certain wiring configurations in the brain." Still, he saw the mind as divided into units, some with quite special/limited purposes. The modularity of the mind was critical to what he meant by innateness.
In his COMMENTS OF A CERTAIN BROADSHEET, Descartes makes what Cowie takes to be a revealing analogy. "This is the same sense as that in which we say that ... certain diseases such as gout or stones are innate" in certain families. People with a genetic disposition may have the faculty of contracting gout. This doesn't mean that they have gout as infants. Nor does it simply mean that gout is a possibility for humans. It means that there is a specific gout contracting ability -- that which as we in the 21st century would say is passed along by the genes of the families Descartes had in mind. The possession of this faculty makes it neither more nor less likely that the possessor will someday get the bubonic plague, or the flu.
So what Locke was really objecting to, when he objected to innatism, was the compartmentalization of the mind. He believed that "what's within" is a quite general learning capacity. When he spoke of the blank slate of our mind at birth we are to understand the singularity of that image as critical to it. It is all ONE slate, and the same chalk that writes geometrical demonstrations there is the chalk that generalizes about the behavior of squirrels, or contemplates whether there must be a First Mover. There is a learning faculty in the human mind, so to speak, there is no congeries of special faculties.
I find this a helpful clarification as to what that debate was about.
Cowie distinguishes this from what she calls the "mystery" hypothesis in innatism, aka rationalism. Rationalists, like empiricists for that matter, "understand that both experience and our innate endowment play critical roles" in our development, learning, etc. The first issue that divides them, as we have seen, is whether the innate endowment is to be understood in a unitary or in a modular manner. The second issue that divides them is the idea of mystery, the idea prominent in the writings of the classical rationalists that there is no natural explanation of these special faculties, that they resist empirical investigation.
Descartes' inability to conceive of mind-body interaction is notorious. Obviously, if I (a mind) learn about the physical world by observing it, performing laboratory tests, etc., then there must be mind-body interaction. The bodies that I observe and generalize about must be impinging upon my own body, thus upon my sense, and thus in turn upon ... my mind. But given Cartesian dualism, that is arguably impossible.
Leibniz was explicit about this impossibility. My mind is my "dominant monad" a world apart from all the other monads that may roughly be said to constitute me, and of course worlds apart from all the monads that aren't part of me. There is no real interaction between them, because as Leibniz said, the soul does not have a window.
So how does it end up acting like it has a window? One can only answer by invoking God, and God's mysterious ways.
Much of the book focuses on two 20th and 21st century figures: Jerry Fodor (1935 - 2017) and Noam Chomsky (1928 - ). Fodor's philosophy revives the impossibility arguments of Descartes and Leibniz: there can be no coherent explanation of human learning, all or almost all ideas must be presumed to be innate. Chomsky, portrayed above, is more interested in reviving the special faculty argument with regard to language than in stressing the mystery of it, but his arguments too may have a non-naturalist side.
In discussing Jerry Fodor, Fiona Cowie uses the term "innatism" in a sense that doesn't seem to imply anything about the modularity of special faculties, although in the earlier section she had concluded that this was what the CLASSICAL controversy over innatism was really about. And this is so despite the fact that the title of Fodor's 1983 book was in fact THE MODULARITY OF MIND.
In discussing Fodor, Cowie uses innate simply to mean the opposite of learned. Whatever we mean when we say we have learned a concept, we simply negate that which we mean when we claim that it is innate. She explains that, in this sense, Fodor believed that most concepts are innate. A very few concepts are compound, and THEY must logically be learned. We must know how to use the word "man" and the word "unmarried" properly, and then we may learn to combine those meanings into that of a new word, "bachelor." But at some point (Fodor and Cowie both take this to be a very common claim) one reaches atomistic, primitive concepts. At that point (Fodor says, Cowie seems to dispute it) one reaches innatism, since primitive concepts can't be learned. Further (Fodor says and Cowie vigorously disputes this) MOST concepts are primitive and as a consequence innate in this sense.
On Fodor's view, every human mind has a concept of a PLATYPUS built in. (Fodor and Cowie both use ALL CAPS as a way of signalling that we are discussing a concept, rather than a word in the vocabulary of a specific historical language on the one hand or a fact in the outer world on the other.) Even if you have never seen a platypus, you have PLATYPUS, and this idea is triggered, it becomes conscious, if you are ever brought to a zoo and have a platypus pointed out to you. The protoconcept becomes a proper concept when it is triggered in this way.
Cowie says that it is an embarrassment for Fodor "that the relations between concepts and their causes typically make a good deal of intentional sense. The fact that we get EMU from emus and PLATYPUS from platypuses and DOORKNOB from doorknobs has an 'intentional integrity' about it of exactly the kind that should not exist, were Fodor right about the arbitrariness of the inputs to the concept-forming mechanism."
Fodor calls his view of concept acquisition "brute causal." There is nothing psychological to say about it, it simply is the case that when I see a dog it triggers in me the protoconcept that then becomes the full-fledged concept DOG.
Cowie contends that this is irrational. Any view of concept acquisition of any value must involve "doing psychology," using "intentional" rather than brute-causal mechanisms.
A story that she tells about this involves the movement of the mind from a certain medium level of abstraction both up and down, toward greater and toward lesser generality. A child typically grasps the idea of "dog" fairly early in life. The movement up in generality, toward "mammals" and "vertebrates," comes later, and the movement is accompanied by a movement in the other direction, toward a concept of "chihuahuas" for example.
Likewise, a child's mind will grasp "chair" fairly early on. Later there will be a movement toward specific sorts of chair, like "recliners," and a simultaneous movement toward categories such as "furniture." On the scale of abstraction there is in both cases a medium that seems to be a developing mind's comfort zone, and various highs and lows working from there.
Such facts indicate to Cowie that there is an intentional process at work in concept acquisition, a deliberate struggling by a mind with the outside world: this is not brute causality. It is not right to say that a person is simply a furniture detector, like a certain machine might be a metal detector. Detecting furniture -- lumping together chairs with stools and sofas for this purpose -- isn't something one is, it is something one does.
Fodor himself replied at length to Cowie, in an essay called Doing Without What's Within.
It is a lengthy essay and difficult to summarize. Here is a link to the whole thing, for this who want to pursue this jousting. Fodor Strikes Back.
Quite early on in this essay, Fodor suggests that the relationship between empiricism and rationalism, or innatism, as classically understood, was symmetrical -- and that Cowie wrongly re-imagined it as asymmetrical so that she could claim that the empiricists had won it. In Fodor's symmetrical rendering, the problem with the old debate was that "nativism is merely the denial of empiricism insofar as we lack a way of saying what `innate’ comes to other than not learned. Likewise, empiricism is merely the denial of nativism insofar as we lack a way of saying what `learned’ comes to other than not innate."
Cowie puts the burden on nativists "to say exactly what doctrine they're endorsing," keeping the empiricists off that hook. Neither when discussing classical philosophers nor when expressing her own views, Fodor charges, does Cowie ever say what she means by learning. She makes, he says, "occasional references to connectionism as possibly an alternative to Chomsky's rationalism," but even so connectionism doesn't come up in relation to Fodor himself, and her occasional references to it fall far short of endorsement.
Connectionism in the relevant sense is the view expounded by Rumelhart and McClelland that looks at learning neurologically, working from the way that (quoting R & M here) artificial simulated neurons can be "linked to each other through weighted connections representing synapses or groups thereof." Cowie cites it briefly in her book as one of the developments in computer science that "promises us machines that can more naturally implement the 'softer' constraints of probabilistic learning processes."
Fodor had had previous runs-ins with connectionists -- he believed it was a trend that was undoing progress that had been made in AI research. So he was clearly unhappy about Cowie's references, however slanting.
Anyway: let us pause there and turn to Cowie's surrebuttal.
She actually concedes something to the point I just cited Fodor as making. She writes,
"I admit that the position I called Enlightened empiricism – the potential alternative to nativism that What’s Within argues had been unfairly ignored by nativists and empiricists alike – was not adequately discussed in the book. This was a mistake, and if I had it to do over again, I'd do it differently. In any case, I now have a chance to say a little more about how Enlightened empiricism is supposed to work, and I trust that by doing so, I can make clear why Fodor's objections to the position in his §2.3.3 are premature."
Here's the link for the whole discussion:
http://host.uniroma3.it/progetti/kant/field/cowiesymp_replytofodor.htm
Her idea of "enlightened empiricism" is the view that growing children learn language through hypothesis testing, that there are constraints at any given time on the range of possible hypothesis, but that these constraints are determined NOT by innate equipment but by the state of her opinion at an earlier time, t-1.
A three year old child inherits preconceptions, then, not from an immortal reborn soul or from the genetically determined wiring of the brain, but from the two year old child. The connectionist model may have something to do with this. It is not the neural equipment we are born with, but how it grows during the years that we're learning language, that is of momentous importance here.
I return now to the book, and now on to its discussion of Chomsky and of his influential arguments to the effect that, (a) all languages conform to the same structural principles, a 'universal grammar,' but that (b) growing children as they acquire their grasp of these languages, do not receive sufficient data from the outside world to allow us to see this universal grammar as learned, thus (c) we ought to see it as innate -- hardwired into our brains. Further, the pertinent hard-wiring is specific to the linguistic domain.
Cowie thinks this is misguided, and to show why, she invokes a culinary example of learning. She writes about curry, that dense use of spices and herbs associated with the Indian subcontinent.
Many humans, perhaps most, eventually learn to recognize curry as a type of cuisine.
How do we learn about curry? Chomsky often asks in effect how does a growing child learn what is an appropriate sentence and what is not? In the same spirit, Cowie wants us to ask: how does anyone (perhaps an adult sheltered in some blander cuisine in his youth, coming to learn about the spicier stuff) come to learn what counts as curry and what does not? One could make the same sort of argument to a mysterious inference, since "it is not the case that everyone can be guaranteed access to the same curry examples," it may be thought mysterious that "all normal people exposed to a curry or two arrive at more or less the correct view about what curries are."
Further, just as there are some cases where qualified speakers of a language, with presumably the same neurological equipment, can argue about whether a sentence is grammatically correct, so there are some dishes that one gourmand might think "curry," another might think not.
In the latter situation, there is the mulligatawny soup [pictured above]. As Cowie explains, this is "a delicious concoction of meats, vegetables, and spices with that characteristically subcontinental taste." On the other hand, although 'paradigmatic' curry is a stew, the mulligatawny soup is, as the name implies, a soup. If one draws a firm line between stew and soup, it is NOT a curry. That firm line is not predestined, and we don't say that someone is utterly ignorant of curry simply because he draws the line where we would not.
Anyway: curry competence may well be considered a nice parallel to linguistic competence. But, Cowie says, we don't posit a special "culinary faculty" in the wiring of the brain "as guarantor of our ability to acquire curry competence" and "we should be equally reluctant to accept [an inborn language faculty] as the conclusion of the precisely parallel argument offered by the nativist in the linguistic case."
Imagine somebody DID make an argument for an innate culinary faculty that enables our learning what is or isn't curry. This somebody would likely be making four mistakes:
- he'd probably be underestimating the amount of "negative evidence" around for an empirical inquiry into what is and isn't curry;
- and might be idealizing culinary learning as if it were instantaneous, whereas actual culinary learning is gradual and piecemeal;
- and is probabilistic, not a realm of certainties;
- and, finally, he may be assuming that there is an endpoint where we all agree about what is curry, whereas in fact there will always be room for disagreements.
The second of these sounds odd, and those not familiar with these debates might not know to what she is making reference. But Chomsky has said, in his KNOWLEDGE OF LANGUAGE (1986), that in understanding language we can presume that the "order of presentation of data is irrelevant so that learning is 'as if it were instantaneous.'"
If we accept the analogy with Curry learning, though, we can see the problem. The order of the presentation of the data may be very relevant in my developing conception of curry. If my first dish of curry is relatively bland, it may take me a while to understand that other curry connoisseurs consider the heat of the dish to be of its essence.
Likewise, language is a complicated developmental fact, and the home in which one grows up has a good deal to do with what one comes to understand of language. The is not Cowie's example but I'll offer my own. I began to learn speech in the Hudson Valley, within sight of the old Tappan Zee bridge. During this process my parents moved our household north and east into New England, to with a matter of yards south of the boundary between Connecticut and Massachusetts, on the Connecticut River. My younger siblings ended up speaking like stereotypical New Englanders to a degree that has never been true of my older brother or myself.
You will say, "regional accents are trivial and besides the point. The issues involve grammar and syntax." It is fair to say it is a trivial example, but it will work as well as others to illustrate the substantive point. Cowie quotes a student of the development of language skills saying that during the second and third years of life "the child seems to be proceeding in a bottom-up fashion, acquiring the language system brick by brick." And that IS the point, for bottom-up learning is exactly what one would expect were learning of language NOT innate, but acquired through our general cognitive abilities applied to the relevant circumstances.
You will say, "regional accents are trivial and besides the point. The issues involve grammar and syntax." It is fair to say it is a trivial example, but it will work as well as others to illustrate the substantive point. Cowie quotes a student of the development of language skills saying that during the second and third years of life "the child seems to be proceeding in a bottom-up fashion, acquiring the language system brick by brick." And that IS the point, for bottom-up learning is exactly what one would expect were learning of language NOT innate, but acquired through our general cognitive abilities applied to the relevant circumstances.
Cowie only briefly discusses Chomsky's famous reply to Skinner, and so far as I can tell her discussion accepts the idea that Skinner got outside of the range of his competence by writing VERBAL BEHAVIOR, and that Chomsky was right to rebuke him.
Her view on language development, then, is not "forget all this subsequent nonsense, go back to Skinnerian 'positive conditioning' as a model, and try again." Her view, rather is that if there is to be a tenable empiricism with regard to language it will have to be an "enlightened empiricism," a post-Skinnerian sort.
And she seems to hope such an enlightened empiricism will develop as a school of thought, though she does not explicitly align herself with it. If I understand her, she is saying that a minimalist nativism as to language ability may be necessary to explain the facts, though it will fall far short of Chomsky's maximalist nativism, or of Fodor's analogous views on conceptualization.
She has in mind a near future in which both Skinner and Chomsky are seen as unsustainable extremes as to the development of language, and in which the enlightened empiricists clash with the minimal nativists over a middle ground.
