Study Guide

Week Twelve: Compatibilism and Libertarianism

Materials Assigned for the Week
The Central Point of This Week's Material
Other Concepts and Points you are Expected To Master This Week
Miscellaneous Comments and Clarifications
1. A Checklist of "Isms"
2. Freedom and the Concept "S Could Have Done Otherwise"

I. Materials Assigned for the Week

Reading:

Lecture 24: A Menu of Positions on Free Will
Lecture 25: Compatibilism
Reading: Hume, Of liberty and necessity, pp. 383-385 just these three pages a second time
Campbell, Has the self "free will"?, pp. 389-402

Exercises:

Review Questions: pp. 325 esp. #5, 335 esp. #3
Problems for Further Thought: pp. 325, 335

II. The Central Point of This Week's Material

Compatibilism is the theory that human actions are both caused and also free. That is, being caused and being free are compatible with each other. Such a view requires, of course, a clear statement of what it means to say "Our behaviour (all of it) is caused". Usually the concept of causation is simply taken over from whatever definition science is using at the time. But the view requires an even more cunning statement of what "Our behaviour (some of it anyway) is free" means. For this job, philosophers usually cut off a sub-set of ordinary causes (e.g. ones "inside" the agent rather than "outside", or ones in which the epistemological mechanisms are functioning normally rather than abnormally); then they argue that behaviour caused by causes from within that sub-set of causes has good reason to be called "free", while behaviour caused by all other causes is "unfree". Sober discusses three compatibilist ways of marking off this special sub-set of causes.

III. Other Concepts And Arguments You Are Expected To Master This Week

The difference between hard determinism and soft determinism
The difference between the view called "libertarianism" in political philosophy and the view called "libertarianism" in the problem of freewill and determinism
The difference between "first-order beliefs and desires" and "second-order beliefs and desires"
The difference between causation and responsibility- i.e. between "X caused Y" and "X was responsible for Y"
What a function is- and a malfunction
What a belief-generating device is - do we have any, do rabbits, do bacteria?
What a desire-generating device is - ditto
What coercion is- and be prepared to explain the reasons for sorting examples and non-examples the way you do
What the "Direct causation" argument against free will is - again
What the "Could not have done otherwise" argument is - ditto

IV. Miscellaneous Comments and Clarifications

IV.A A Checklist of "Isms"

Sober gives a checklist of the off-putting number of "isms" populating this territory on pp. 318-319. It may help to whittle down the number of serious players, and to put things strictly in terms of Thesis One and Thesis Two from Week Eleven (pg. 309).

Thesis One: We are part of the causal network of nature. Every action we perform is caused exactly as much as every other event in the natural world is caused.
Thesis Two: We are free beings. Some of the actions we perform we perform freely, out of our own free will.

Here are the only permutations worth memorising:

The truth of Thesis One rules out the truth of Thesis Two, One is true, therefore Two must be false - "Hard Determinism".
The truth of Thesis One rules out the truth of Thesis Two, Two is true, therefore One must be false - "Libertarianism".
There is no special conflict between the truth of Thesis One and the truth of Thesis Two, One in fact is true, Two is also in fact true - "Compatibilism" (alternatively called "Soft Determinism").

IV.B Freedom and the Concept "S Could Have Done Otherwise"

This is the key component of every variety of compatibilist position. There are many different compatibilist positions (Sober lays out three and Campbell has his whack at the concept of "S could have done otherwise" too, but as a libertarian not a compatibilist). So obviously there are many different analyses of what this rather opaque phrase means. Here are some of the less opaque ones.

Hume's analysis of "Could have done otherwise"

Here is Hume's definition again, from Week Eleven (pg. 383):

"S is free" =def "S did X but S could have done otherwise than X had S wanted to"

To my reading of the Hume selection (pp. 384-385), that comes more or less to this:

"S is free" =def "S did do X all right, and X was caused all right, but the causes of X were internal to S ("determinations of his will") rather than external to S ("not a prisoner and in chains")

Sober reads Hume's proposal similarly (pp. 321-323):

"S is free" =def "S did X, and X was indeed caused, but the cause of X was S's beliefs and desires as opposed to S's genes and environment"

Notice that Hume's aim in this part of the reading is quite limited. It is simply to define "freedom" in such a way that it doesn't conflict with universal causation - so that when we say S performs some voluntary action, it will be correct to say that S causes it to happen in some fairly straight-forward sense of "causes" (namely "constant conjunction"). Hume's procedure is typically ingenious:

He takes all the sorts of causes of events there exist in the world.
He isolates a special sub-set of such causes.
He assigns every cause of every event as belonging either to that special sub-set or as belonging with the rest.
He then insists that if some action was caused by a cause belonging to the special sub-set, we have a good reason to call it "free".
If it was caused by a cause not belonging to that sub-set, we have good reason to call it "unfree".

The sub-set of causes which Hume hits upon as relevant for decisions about free will - and remember he insists that there are causes, perfectly ordinary ones - are those causes which we can locate "inside of" him rather than "outside of" him. What is inside Hume are the likes of his beliefs and desires, everything Descartes argued belonged to the mind (Hume was a dualist) and every event and process and state a Materialism will locate inside the brain. What is outside Hume are the likes of his genetic heritage - inside his epidermis admittedly but hardly a brain event - his upbringing, his present exterior environment (whether he is chained to the floor, locked up in prison, or someone right now is holding a gun to his head), the opinions of the Almighty about his latest book, and so on.

Of course it will be hard to draw a very precise line between all the states and events and processes in the universe which belong "inside" Hume and all the states and events and processes in the universe which belong "outside" Hume.

Think of examples!

clear-cut "inside" items

clear-cut "outside" items

borderline cases it is hard to decide

However, so long as some kind of more or less usable line can be drawn, then that will automatically draw a more or less usable line between two different kinds of causes for "X causes Y" statements as well. There will be events, states, processes which are causes - perfectly ordinary causes remember - residing inside Hume. There will be other events, states, processes which are causes - more perfectly ordinary causes - residing outside Hume. Now for the punchline:

When S is caused to do X because of a cause that resides outside of S, then that is good grounds for saying that S was not in fact free to do otherwise - "Sure I was late for the meeting, but the bus broke down 10K away, so what could I do?"

When S is caused to do X because of a cause that resides inside S, then that is grounds for saying S was indeed free to do otherwise - "Sure I overspent this month, but that stereo system was such a bargain that I decided not to put it off."

Take your examples one more step!

For the clear-cut "inside" causes, is it clear we could have done otherwise?

For the clear-cut "outside" causes, is it clear we couldn't have done otherwise?

For the borderline causes, is it borderline whether we could or couldn't have done otherwise?

There is one more benefit. Wherever we can draw a line between inside and outside causes, and therefore between actions where we could and couldn't have done otherwise, the very same line seems to cut off those actions for which we can be held morally responsible and those for which we can't. For example, I can excuse myself from any blame for being late to the meeting; it's wasn't my fault; it was the fault of the bus; the cause of being late lay outside of me; I wasn't responsible for that; don't put this one down to moral flaw. But I will have to take moral responsibility for overspending my paycheque this month; I can make all kinds of excuses to be sure ("What a bargain", "The salesman said it was the last one", "I couldn't resist"); but when all is said and done that purchase and its consequences is still down to me; no one else is morally to blame for the fact that we are now broke (not even the all-too persuasive salesman).

Take your examples the final step!

For the clear-cut "inside" causes, is it clear we would (or should) accept moral responsibility?

For the clear-cut "outside" causes, is it clear we would (or could) refuse moral responsibility?

For the borderline causes, is it borderline whether we would or wouldn't accept moral responsibility?

I'm not saying this long story is entirely correct in all its details. (Very rarely does a philosophical story turn out to be entirely correct.) But Hume's concept of free will and its ramifications is certainly more powerful and consistent than it may have looked to you at first.

cause of X is "inside" of S	cause of X is "outside" of S
S could have done otherwise than X	S couldn't have done otherwise than X
S is morally responsible for X	S is not morally responsible for X

It all hangs together nicely. And note for a final time that it is a thoroughly causal story, so there is no conflict whatsoever between Thesis One (everything is caused) and Thesis Two (some of our actions are free). Freedom is indeed compatible with universal causation.

Dworkin and Frankfurt's Analysis of "Could have done otherwise"

Sober introduces this refinement (pg. 324) to handle some difficulties (with compulsive behaviour especially) which Hume's compatibilism doesn't seem to be able to treat consistently. The refinement is to distinguish "first-order desires" from "second-order desires" - we can desire to have or not to have certain other desires.

"S is free" =def "S did X because S has the [first-order] desire D to do X and S does not have any [second-order] desire to drop D from S's list of first-order desires"

What does this mean? As usual, the best starting place is to look at the problem which the refinement is supposed to solve. We are reluctant to say that kleptomaniacs steal "of their own free will". But at the same time the causes of a kleptomaniac's stealing do seem to reside "inside" rather than "outside" the kleptomaniac - no gun to the head and so on. The causes belong in the sub-set of causes, "beliefs and desires", as opposed to the subset of causes, "genes and environment".

Dworkin and Frankfurt solve this puzzle by putting two sorts of beliefs and desires "inside" S, rather than Hume's one sort. "Inside" each of us (including the kleptomaniac) are ordinary beliefs and desires, and then as well beliefs and desires about those beliefs and desires. The first are "first-order" and the second are "second-order". Both sorts are just as much causes of whatever behaviour gets caused as any other causes in the universe. Thesis One is never in the least doubt. But now we have available a subtle refinement on Thesis Two. We are free not purely when the causes of our behaviour do reside "inside" us, as opposed to "outside" us. Instead, it is only when the first-order and second-order beliefs and desires which reside "inside" us and cause our behaviour are related to each other in a certain way. Specifically, when one of our first-order beliefs and desires is the cause of the behaviour and when the second-order belief or desire we have about that first-order belief or desire is neutral.

What marks poor kleptomaniacs off is that their second-order desire is to be rid of the first-order desire which causes them to steal. That is why we say they are not free to do otherwise: their inner set of causes is at loggerheads. For the rest of us, for most of the first-order beliefs and desires which cause our behaviour, we have no second-order beliefs or desires at all, much less negative one. That is why we say we are free to do otherwise: our entire set of "internal" causes is in a certain state of harmony.

Sober's Own Analysis of "Could have done otherwise"

Sober's version of compatibilism is Lecture 25. The definition of "free" is easy, but "could have done otherwise" is a bit more opaque. Start with the weather vane (where "being free" is a contrast with "stuck", however, not an equivalent for "possessing free will"):

"WV is free" =def "The behaviour of the weather vane is caused, and the weather vane is presently causally sensitive to the factors in its environment which it was designed to be sensitive to"

In short:

"WV is free" =def "The behaviour of the weather vane is caused and the weather vane is not malfunctioning"

This is the idea to be carried over to human agents about whom we have serious freedom and determination questions:

"S is free" =def "S did X, and S was caused to do X, and the devices which cause S's behaviour (S's so-called "belief-generating devices" and "desire-generating devices") are not malfunctioning"

(Compare Sober, pp. 327-329.)

Anything which does perform a function, such as a belief-generating device or a desire-generating device, is perfectly capable of functioning well or poorly, or even stopping from functioning at all. We say that S could not have done otherwise when its relevant devices are not functioning up to scratch - like the weathervane rusted solid so that no matter what inputs it receives from the environment, the device outputs exactly the same thing, pointing north. We say that S could have done otherwise when its relevant devices are all functioning well. That is, when they are sensitive in the manner Mother nature intended them to be. That is, when they provide suitably different outputs whenever they are given suitably different inputs. In such a case, while the output actually produced by the relevant device or devices is X say, given a different set of inputs the output would not be X but Y say. Here then, while S did do X, S could have done Y instead - i.e. S would have done Y were the circumstances different. (Notice Campbell's criticism of all such suggestions, pp. 391-393.)

Campbell's analysis of "Could have done otherwise"

Campbell is not a compatibilist but a libertarian. That is, he does not think freedom is compatible with causality. Rather, when we have finally hit upon the proper concept of freedom, we will find ourselves doing something very different than merely sorting out causes from one sub-set of causes rather than another sub-set. We will have found a relation between person S and action X that belongs completely outside of all such sub-sets of causes. This provides a very different kind of meaning to the concept "S could have done otherwise than X". Here is the key quotation:

"X is just as morally praiseworthy as Y or Z if he exerts an equivalent moral effort, even though he may not thereby achieve an equal success in conforming his will to the "concrete" demands of duty. And this implies, again, Common Sense's belief that in moral effort we have something for which a man is responsible without qualification, something that is not affected by heredity and environment but depends solely upon the self itself. Now in my opinion Common Sense has here, in principle, hit upon the one and only defensible answer." (pg. 394)

The actions Campbell argues are the really important ones for moral responsibility - which is the real reason we are interested in free will in the first place - are restricted to a relatively narrow field of operation, basically the field in which desire conflicts with duty. There is one course of behaviour X which we believe we ought to follow and another course of behaviour Y which we strongly desire to follow. The interesting sort of free will is exclusively "the freedom to put forth or refrain from putting forth the moral effort required to resist the pressure of desire and do what he thinks he ought to do" (pg. 397). Now, of course, external actions are consequences of such inner acts and obviously subject to interference from the environment and the like (so they can hardly depend solely on S alone). So it's those inner acts of putting forth or withholding moral effort that Campbell sticks to. Here his thesis that we are free is rather easier to make.

"S is free" =def "S did X and S's moral effort is the sole determinate of X's getting done, and S is the sole author of exerting that moral effort"

What's ingenious in this definition is that none of the concepts used on the right seems particularly causal. I leave it to you to show whether Campbell actually has got hold of a live alternative to ordinary causality here.

Exercise!

Defend or criticise the following claims:

"S's moral effort is the sole determinate of X" /= "S causes X"

"It is solely S who makes or withholds moral effort" /= "S causes Y".

"S is the whole author of any act of moral decision" /= "S causes X".

That will do for the concept of being "free" perhaps. But what does this all get us for the concept "could have done otherwise"? Campbell goes back to his basic invention here, the concept of "sole authorship". This concept, Campbell argues, delivers much more than the "if the circumstances or beliefs were different S would do otherwise" sense of "could do otherwise" used by compatibilists from Hume to Sober. Since no one else is involved, the exerting or withholding of moral effort is obviously an act our performance of which could have been otherwise. And this not merely if the circumstances and so forth where different from the way they now are. Let the circumstances etc. be exactly the same. Indeed, focus down on a single occasion only, my present one. Concerning the exerting or withholding of my putting some moral effort into something, I have the actual power right at the very point where I decide to do one thing to do another thing entirely. If I decide to make a decision to exert moral effort, I have the genuine present power, not just the "if circumstances were different" power, to decide to withhold moral effort (pg. 395). Being the sole author of the one, I am automatically the sole author of the other. This is "could have done otherwise" like Mother used to make.

Contents
Previous Week: The Problem of Freedom and Determinism
Next Week: Review Week