What does Daniel Dennett mean by "intentionality"?

Question

In the examples of "The Wandering Two-Bitser, Twin Earth, and the Giant Robot" from the Intuition Pumps book, also found here, what is the discussion actually about?

It seems always like a mixture of:

• Meaning of symbols
• Thoughts/concepts
• Goals/intentions
• Mental states being about something

though "intentionality" traditionally only refers to the latest point.

And mixing in "goals" (or intentions) is especially vexing here.

So does he apply some own idiocratic definition? Or does he want to show that all of those above basically are really just the same?

Answer 1

For understanding some philosophers, it IS critical to get a precise definition of terms. That often isn't the case with Dennett, he generally uses words a bit imprecisely, more like a normal conversation, than like a math proof. That there isn't a single, precise meaning to "intentionality" is actually readily observed from the way he applies it across multiple authors, who are writing with very different objectives and in different modes. Dretske discussing the nature of "error", Kripke on "rule following", Millikan's discussions of function, the intentionality assumed in biochemistry, the intentionality assumed in evolutionary biology, Searle discussing the critical differences between syntax and semantics -- this diverse subject set CLEARLY cannot all be operating off one common meaning of "intentionality!!!

Hence your question -- what is Dennett REALLY talking about? One key to understand this is to recall that US philosophy of mind was dominated by behaviorists for about a half century -- so that when one tried to discuss the mental -- the vast majority of philosophers insisted that one could not actually reference anything INTERNAL. There were a few decades following the behaviorist domination that were similarly dominated by functionalism, where one could reference internal events, but only "functionally". Dennett was writing while functionalism was still in domination, so philosophers actually trying to discuss consciousness are going to often use code-words, or else make an effort to pretend that function == consciousness. One such term often used was "intentionality" or "an intentional stance". Under modern reading, one should generally translate this into a debate as to whether we are conscious or not.

Dennett effectively argues "not", by embracing functionalism wholesale. IF one functionalizes "consciousness", into "intentionality", then it is actually pretty clear that Dennett has the right side on this. Complex enough Robots can do ANY function! And there is not any indication of a critical threshold of complexity of function (some TBD emergence principle) that causes "magic" to happen with only some complex functions, such that they are "conscious" and other functions are not. And it is certainly true that "intentionality", if it is defined purely functionally, can be assigned to quite simple functions, such as that performed by a "two-bitzer". So, Dennett is pointing out that strict functionalists about consciousness, which most of his contemporaries were at the time, should either ascribe consciousness to all functions, or to none.

Functionalism has since gone into partial decline, as many advocates of it were not willing to accept the conclusions Dennett argued for here. It is also now acceptable to admit that we can and do have internal evidence of consciousness -- IE that we have qualia, and perceptions, and we actually do have personal and privileged access to this information. The plurality of philosophers of mind today have fused functionalism and qualia with an acceptance of a TBD emergence process by which complex functions somehow generate consciousness, so that they remain "physicalist" at least in the origin of consciousness. By accepting emergence, this would then be a non-reductive physicalism.

Dennett, who is committed to reductionism, refused to accept this compromise between physicalism and real internal experiences (in other writings is becomes more clear why -- he thinks all such models end up being dualist). Emergence was mostly frowned on at the time of this essay, but it has since become pretty widely accepted. Most of the rivals he was attacking in this essay went on to adopt an emergentist approach to consciousness, and become explicitly non-reductive materialists.

Answer 2

The concept of intentionality is subtle and intrinsically difficult. In the philosophical literature, the term refers to a capacity to make a mental representation of something, a representation which has content or meaning. In other words, say I write a phrase like: "There is a cow standing in a field". I have intentionality because I can make that mental representation as I write it down; you have intentionality because you gather or create that mental representation as you it; the text itself has a kind of intentionality by extension, because I use it to convey that mental representation to you. Intentionality is meant as a bridge between the subjective and objective worlds; if you'd like a rubric, think that if you apply attention to something outside yourself, you create intention in the form of a mental representation.

As I noted, the text I wrote has a different kind of intentionality called derived intentionality. The text does not know what it means; the text does not have a mental image on its own. The text merely transfers my mental image in a way that you can create an equivalent mental image. What you and I have is called original intentionality in the sense that we actually attach some sense and meaning. Having produced that phrase, both you and I could run out and find a field with a cow in it, and say "Yes, that's what that phrase means." We make a correspondence between what's in our heads and what's out there in the world, and doing that is an intrinsic part of our nature. By contrast, a cat that looked at that phrase would not have a mental representation of a cow in a field: cat's can't read, and so cannot use that tool to transfer a mental representation. It's a question of some debate, in fact, whether cats and other animals have intentionality at all: whether they are capable of making mental representations of that sort.

I happen to be in the camp that grants animals a degree of intentionality, but Dennett goes in the other direction and actively disputes that there is such a thing as original intentionality at all. Instead, he sets out to collapse the concept of original intentionality into the concept of derived intentionality, and he does this by reinterpreting intentionality as a capacity to make decisions. He does this in three steps:

1. He starts with a simple machine: a vending machine with a device that 'decides' whether an inserted coin is a valid quarter. This is derived intentionality, of course. The device has no mental representation of a valid quarter, but has been designed to weigh and measure according to some human's specifications.
2. He draws out more and more complicated machines until he gets to a computer programmed to make complex decisions in ambiguous circumstances — a true AI — but asserts that this AI still has only derived intentionality. It does what it was programmed to do by some human; it has no mental representations of its own; it attaches no meanings or content to its processes or actions.
3. He invoke the theory of evolution to suggest that human beings are merely biological machinery, no different in kind than the sophisticated AI discussed in #2, and thus can only have derived intentionality. The chemical processes on the level of DNA and RNA do not have 'original' intentionality — they don't have mental representations of any sort, and are not teleological — and there is no point in the increasing complexity of cells, organs, nerves, or brains in which can be seen to develop 'original' intentionality.

Of course, the immediate problem with this line of reasoning as that points 2 and 3 are presumptive. Since he denies the existence of original intentionality he doesn't seem to bother examining the concept, and he glosses over problematic questions like "What biological, electro-chemical processes caused my fingers to move in such a way that they typed out 'there is a cow in a field' on a keyboard?" That particular action hardly seems to increase my survival fitness (and likely decreases my chances at successful mating), so...

This is really the Problem with Dennett's approach. Intentionality isn't just about 'making a decision.' Intentionality is a matter of 'making meaning' within which decisions appear. If we comes to a fork in the road, we have to first establish the meaning of it — that there are now two paths instead of one — and we have to pull in other meanings, other mental representations of where we want to go and what we happen to be doing, etc. If we couldn't establish that meaning we'd be like a dumb robot that crashes straight into the trees between the two paths. And yes, we can program robots to make more sophisticate decisions at junctures like this, but what isn't clear is whether we can program a robot to make meaning of these junctures. Dennett boldly and blithely asserts that we cannot, and that therefore we ourselves do not have original intentionality.

This is merely speculation or belief on Dennett's part (though it would anger him to hear me say that). Biology and neurology are nowhere near developed enough to create more than the simplest links between objective biology and subjective experience, and computer science hasn't come close to creating a full-fledged AI. Dennett is jumping at this conclusion because he wants to close the door on any possible religious or metaphysical ideation (such as the idea that 'intentionality' is a product of the human soul), and thus has turned a blind eye on the limitations of the sciences and of the consequences of his own beliefs. C'est la vie...