Saturday, October 15, 2016

Can Words Like Belief, Desire, Hope, Fear Be Scientific?

Review and thoughts triggered by “Skinner Skinned” by Daniel Dennett (1978)

Many hate Skinner because his behaviourism implies that people are not free, dignified, morally responsible agents. He does this by reducing all behaviour to probabilities based on past regimes of stimulus and response. Since our responses are conditioned by forces outside of our control when young then our responses in later life are automatic, unconscious, not free.

Nevertheless, Skinner's was driven by the admirable goal of explaining behaviour scientifically. So, when we evaluate Skinner we also evaluate our attitudes to science and the role of scientific world view in our whole life. If we reject Skinner then there is an accompanying danger of rejecting a scientific world view, of throwing the baby out with the bathwater. Alternatively, a better pathway would be to develop a more robust and broader view of what it means to be scientific. This is what Daniel Dennett is attempting here.

This preamble serves to highlight the importance of accurately identifying the real mistakes of Skinner. In the past I was satisfied with Chomsky's critique. I now feel that Chomsky's critique is inadequate and leads people to irrationally reject all of behaviourism.

Skinner's enemy is mentalism, that talking about people's behaviour using terms such as beliefs, desires, ideas, hopes, fears, feelings, emotions is not allowed because it is not scientific.

Dennett, however, sees a positive role for these mentalist idioms in explaining behaviour / psychology / mind. So this essay, while acknowledging there is some point to Skinner's objections to mentalism, is about clarifying where mentalism is useful and where it stops being useful and becomes unscientific.

Skinner gives many reasons for disqualifying mentalism:
  1. mental things must be made of non physical stuff (dualism objection)
  2. the mental is private (whereas behaviour is public and objectively measurable)
  3. mentalism appeals to events that can only be inferred
  4. mental events are internal
Chomsky takes (3) as Skinner's prime objection against mentalist psychology but Dennett points out that Skinner is not against inference as such in other passages in his writing. Chomsky Reference: “The Case Against B.F. Skinner”, New York Review of Books (Dec 30, 1971)

However, Dennett points out, through a close reading of Skinner's “Beyond Freedom and Dignity”, that although Skinner uses these arguments occasionally he also qualifies them and contradicts them. So, there must be something else to Skinner's objection to mentalism.

The something else is the virtus dormitiva, the dormitive virtue, defining an effect as its own cause. From Moliere's Le Malade Imaginaire, The Imaginary Invalid. What is it in opium that puts people to sleep? Why, it is its sleep producing powers of course!

The key argument of Skinner's objection to mentalism is a little man in the machine, a homunculus.
“The function of the inner man is to provide an explanation which will not be explained in turn” (quoted from Skinner's Beyond Freedom and Dignity, p. 14)
We must abolish “the autonomous man – the inner man, the homunculus, the possessing demon, the man defended by the literature of freedom and dignity” (Beyond Freedom and Dignity, p. 200)

Dennett is exasperated by Skinner's array of different objections to the little man in the brain: moral agent = little man in the brain = demons. Skinner then sees superstition behind any and every claim of moral responsibility. “Mental” means “internal” means “inferred” means “unobservable” means “private” means “virtus dormitiva” means “demons” means “superstition”.

Nevertheless, Skinner has a point even though he exaggerates his case. There is a real danger in presupposing intelligence when we try to explain intelligence.

When you use a certain vocabulary ( the mentalist words which refer to beliefs, desires, ideas, hopes, fears, feelings, emotions) then does that presuppose intelligence or rationality?

Yes it does. Dennett agrees with Skinner here.

These mentalist terms are called intentional idioms by philosophers.

Quine also is opposed to intentional idioms in psychology. But his objection is different to Skinners. He does not argue that intentional idioms presuppose rationality or offer no explanation. His argument is that we can't translate sentences containing intentional idioms into sentences lacking them, they can't be reduced to the sentences of the physical sciences.

Quine Reference. Word and Object, #45 The Double Standard, pp. 198-203

The issue becomes more complicated because Skinner, unlike Quine, believes, at least in some of his writings, that intentional idioms can be translated into the language of physical science. Dennett argues that Skinner is inconsistent (“sloppy”) in his arguments on this point. More on this later.

Dennett's advice is that we can agree with Skinner that no satisfactory psychological theory can rest on any use of intentional idioms, for their use presupposes rationality, which is what psychology is supposed to explain.

However, Dennett advises us to disagree with Skinner when he takes the further step that intentional idioms therefore have no legitimate place in psychological theory at all.

Dennett argues that we can use intentional idioms as a starting point of explanation provided we are aware of the dangers (of virtus dormitiva, defining an effect as its own cause or the little man in the machine). It is ok to speculate first and explain later in more scientific language.

Returning to Skinner's allegedly sloppy arguments / inconsistency about intentional idioms. Dennett demonstrates through quotations from Skinner that sometimes he eschews intentional language such as “beliefs” and instead argues in terms of changing probabilities. But at other times Skinner argues that intentional words such as belief can be translated into behaviourist (“scientific” in Skinner's view) terminology. But, on balance, Skinner concludes that “scientific', which for him means behaviourist or probabilistic explanations can't co-exist with intentional explanations which cite beliefs, desires etc.

The reason many don't like or hate Skinner is that he wants to take away our personal moral sense, that we, as individuals, possess freedom, responsibility and dignity. By reducing all behaviour to probabilities based on past regimes of stimulus and response this is what Skinner ends up doing.

In contrast, Dennett argues that intentional idioms and scientific language, include stimulus response language, can co-exist in our thinking and description of psychology. Specifically, Dennett is saying that intentional idioms, expressions of belief, desire etc. can be reduced or translated into other more scientific terms, as a consequence of study, research, theorising etc.

Dennett's claim here, as mentioned in this essay is disputed by Quine, who says translation from the intentional to the scientific is not possible. It is also disputed, I believe, by Hilary Putnam, although I need to study more of Putnam before I could accurately state his position.

In the rest of the essay Dennett provides some examples of situations where intentional explanations can co-exist with more reductive scientific explanations. One of Dennett's favourites is the chess playing computer. … “... we know that there is a purely mechanistic explanation of the chess playing computer, and yet it is not false to say that the computer figures out or recognises the best move, or that it concludes that its opponent cannot make a certain move, any more than it is false to say that a computer adds or multiplies.” (p. 64)

This and other examples provided by Dennett (such as someone handing over their wallet when threatened) demonstrate that the only type of explanation allowed by Skinner of always having to provide back to basics, stimulus-response probabilistic reasons is taking a good idea and stretching it too far.

This issue of the limits and usefulness of intentional idioms is one I need to study more. As noted above Quine and Putnam put a more sophisticated argument against than Skinner.

No comments: