Brainstorms: Philosophical Essays on Mind and Psychology

.

Daniel Dennet's second book is a collection of essays on consciousness, free will, the possibility of artificial intelligence and other subjects. Good as always.

Dennett is my favourite contemporary philosopher. And, of course, in a case like this, it is hard to say whether I like him because I agree with him on a large number of issues or whether I actually formed my opinions because of him. (See his How to change your mind.)

He is witty, knows the subject and is sure to come up with fresh ideas and new metaphors. On the other hand I understand that many people think he is a bit too witty. And surely he lost a lot of credit by publishing a book called Consciousness Explained that people quite generally thought did no such thing. (I also do not like that book but for different reasons.)

I had never read this particular book before. It is from 1978 and it shows. (Use of pronouns). Some of the essays seem really dated and he even manages to present some really boring stuff here. E.g. Towards a Cognitive Theory of Consciousness

On the other hand you will find plenty of great stuff, sometimes hidden in footnotes. Like the quip that the expression c-fibers has lost its empirical anchor in neuroscience and is used as a philosopher’s wild-card for whatever event turns out to be identical with pain. (193)

Or this one: “It has seemed important to many people to claim that computers cannot in principle duplicate various human feats, activities happenings. Such aprioristic claims, we have learned, have an embarrassing history of subsequent falsification.” (190)

This is as true today as it was then.

And I just love this one (a propos the concept of pain): Real Chateau Latour has to have been made in a particular place in a particular way by particular people.... (Real vodka, on the other hand, can be made from just about anything, anywhere, by anybody.)” – You can have a complete seminar on this.

And talking about actual content. Sabine Hossenfelder should maybe read On Giving Libertarians What They Say They Want to get an idea how determinism and free will might be compatible.

But I want to say a few words on his essay Skinner Skinned (... too witty) to give an example of where I think Dennett was on the wrong track.

Skinner (and behaviourism) is today more or less forgotten. But in 1978 he was obviously thought of being worth fencing with although, as Dennett says, “psychologists just don’t find behaviorism very reinforcing these days“. (53)
Skinner was of the opinion, that mentalistic explanations were not helpful, were unscientific.

As Dennett wonderfully summarizes: “mental” means “internal” means “inferred” means “onobservable” means “private” means “virtus dormitiva” means “demons” means “superstition”. (58)

And he quotes Skinner (from Beyond Freedom and Dignity) with this beautiful statement: ‘“Science does not dehumanize man; it de-homuncalizes him...”’ (58)

The basic argument of Dennett is that intentionalism is capable of explaining things that behaviorism cannot. It is an application of his “stances“ theory. There are different ways to explain the same thing. You can argue from the design, from physics from intention. And he is right with this, of course. You can say that a chess playing computer “wants” to avoid chess mate and “therefore” moves its king to a different field. (And if you not prepared to say so you cannot say that a computer can multiply, because it can only add, it cannot add because it can only open and close tiny switches etc.)

But this, I think, does not affect Skinners basic argument, that mentalism is not an explanation.

Here is the thought experiment by Dennett: Imagine we have two “self-learning” chess-programs. Now, suppose, he says, we let them play against each other without recording the games. And then we discover that their game has improved. This is a remarkable prediction of machine learning, but then he goes wrong, because he says: “It must be possible in principle to determine mathematically what these evolved programs are now...” (68) He thinks we could, in principle, take the computers apart, figuring out what their programs were in that way. No. We could not.

Now, he says, the intentionalist can predict the next move of the improved computer. (Right, but a prediction is not an explanation.) Skinner on the other hand, must deduce that the “computer during the night must have been ‘reinforced’...”. (69)

This is right, but it is not an argument against behaviourism but one in favor of behaviourism. And the recent advance in AI just confirms Skinner. The AI learns by reinforcement. (It had never occurred to me before that in AI they do use the behaviourist’s term reinforcement!) And there is no mentalistic explanation on top of this.

Another argument Dennett uses is this one. Suppose you were held up and asked to hand over your wallet. This has never happened before and cannot have been “reinforced”. And still, you hand over the wallet. (On the contrary, Dennett says, the reinforced thing to do, would be to apologize.) Dennett says that only the mentalistic approach works. I want to save my skin, I believe I am threatened. Etc.

But here is the test. I asked ChatGPT: If you were a human and were held up and asked for your wallet, what would you do?

Here is (part of) the answer:

1. Stay calm: It's natural to feel scared or anxious in such a situation, but try to stay as calm as possible. Avoid making any sudden movements or showing signs of resistance that may escalate the situation.

2. Comply with the demands: If the perpetrator asks for your wallet or other belongings, it's generally best to comply. Material possessions can be replaced, but your life and safety are paramount.


So, where is the wanting here? Where is the believing? It is all the rational reaction of a language model without mind, trained by reinforcement. And there is no principal reason why the reaction of a human should have any other grounds.

I can't remember much of it but have since come across so many references to Dennett that I have to re-read it.

Some hard papers by Dennett here. I love Dennett's writing style and his critical thoughts. He does have a tendency for run-on sentences and paragraphs, but the thoughts he presents are nuanced and well-considered. I wonder how much time he took to write this collection. This helped me with writing my Philosophy IA.

After reading Thomas Nagel's classic essay, "What is it like to be a bat?", this was the book that turned me onto philosophy of mind. Dennett's a brilliant philosopher, well-read in science, and an engaging writer. "Brainstorms" (perfect title!) is a very good introduction to Dennett's wide-ranging and creative work in philosophy of mind and cognitive science. The last essay in the book ("Where am I?") is a delicious dessert! I heard him present it in person at a philosophy conference at UNC Chapel Hill in 1980 or '81. It ended with a standing ovation from a room full of philosophers (presumably not an everyday occurrence).

Collection of papers from the 60s/70s, shows some progression from "Content & Consciousness" and deals with new theories from Fodor and others. Also reviews then-fading influence of Skinner and Ryle.

The late Daniel Dennett is justly famous. This popular collection of essays puts to shame the joyless writing of so many professional philosophers. And his ability to draw on various scientific disciplines remains the hallmark of good practice in Cognitive Science.

To appreciate all this, you can start with the essay on Behaviourism’s most famous proponent (“Skinner skinned”), and see for yourself just how balanced and thoughtful Dennett’s criticisms are. Chp-7, likewise, does a great job of demonstrating his penetrating observations about Psychology and related fields. And the final chapter (“Where Am I?”) was a delight to read, even if it does not advance any substantive thesis.

Of course, Dennett benefited from a privileged education. And his many public appearances, in my opinion, lack the force of his writings. But if the cognitive sciences programme is to progress at all, we have a solid guiding track in Daniel Dennett’s œuvre, with his characteristic intellectual honesty and his naturalistic take on problems in philosophy.

Hopefully, this will help usher in a more mature field, with mutually constraining discoveries and hypotheses from the multiple sub-disciplines — and take us beyond the current age of fruitless but every-growing theories and models.


©  Creative Commons CC BY-NC 4.0