Study Guide

Week Eleven: The Problem of Freedom and Determinism

1. Materials Assigned for the Week
2. The Central Point of This Week's Material
3. Other Concepts and Points you are Expected To Master This Week
4. Miscellaneous Comments and Clarifications
   1. The Selection From Hume
   2. The Selection From Sartre
   3. The Selection From Skinner
   4. Causality and Determination

I. Materials Assigned for the Week

Reading:

Lecture 23: Freedom, Determinism, and Causality
Reading: Hume, Of liberty and necessity, pp. 374-388
Sartre, Existentialism, pp. 494-501
Skinner, Determinism rules out freedom, pp. 402-407

Exercises:

Review Questions: pg. 316 esp. #1, 3
Problems for Further Thought: pp. 316-317

II. The Central Point of This Week's Material

The Problem of Freedom is the problem of whether we can properly be held responsible for those of our actions which we choose to do. (Assume we can't be held responsible for brainwashed and compulsive behaviour.) An answer is not as simple as it may seem because the are good reasons to hold both of the following theses:

1. our actions (all of them) are part of a causal network, and
2. our actions (some anyway) we perform of our own free will.

Each of these gets in the other's way. If 1 is true, it seems that 2 must be false. If 2 is true, it seems that 1 must be false. Sober sets out the prima facie force of this conflict between causality and freedom.

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

• What causality is
• Whether causality is a transitive relation or not- the example in the Box on pg. 307 is a nice one
• The difference between "X causes Y" and "X is a sufficient condition for Y"- go back to the Box on pg. 155 for a refresher on the second concept
• The difference between "X causes Y" and "X is a necessary condition for Y"- ditto
• Whether this implies that "X is both a necessary and a sufficient condition for Y" will provide much of a better analysis of "X causes Y"
• The difference between causation and determination- i.e. between such statements as "the occurrence of X causes Y to occur" and "the occurrence of X determines that Y will occur"
• The difference between causation and necessity- i.e. between "the occurrence of X causes the occurrence of Y" and "the occurrence of X necessitates the occurrence of Y"
• What the "Distant Causation" argument against free will is (pp. 306-307, 385-388, cf. 330) - set it out in steps [pg. 330 will show you how to leave out God] and decide whether it is a particularly good one
• What the "Could Not Have Done Otherwise" argument against free will is (pp. 307-308, cf. 331-332)- ditto
• The difference between a free action and an unfree action - you won't have anything like a final answer yet, just make a start and keep track of examples
• The difference between determinism and indeterminism
• The difference between determinism and fatalism
• The difference, or lack of difference, between "free will", "freedom" and "liberty"
• What Hume means by the claim: "Beyond the constant conjunction of similar objects, and the consequent inference from one to the other, we have no notion of any necessity or connexion" (pg. 376)
• What existentialism is
• What Dostoevsky's slogan means: "If God didn't exist, everything would be possible" (pg. 498)
• What reinforcement theory is
• What Skinner means by the claim: "We can ... achieve a sort of control under which the controlled, though following a code much more scrupulously than was ever under the old system ... are doing what they want to do, not what they are forced to do."

IV. Miscellaneous Comments and Clarifications

IV.A The Selection From Hume

Hume is easier reading than he looks.

The central theses which produce a special problem about freedom are as in II above (compare Sober, pg. 309):

1. We have good reason to believe that we and every action we perform is part of a causal network, certainly as much part of the normal run of causes and effects as anything else in nature.
2. We have good reason to believe that some of our actions anyway we perform of our own free will - i.e. that we have a certain amount of liberty to behave as we do or refrain from so behaving.

The problem about the very concept of human freedom is that it appears we can't have it both ways. Either we are like falling rocks (1), or we aren't like falling rocks (2).

In most of the reading, Hume is simply setting out these two theses. The first thesis takes up pp. 375-383: everyone agrees, says Hume, that human behaviour is exactly as much subject to causation (for Hume "constant conjunction") as anything else in nature. The second thesis takes up pg. 383: everyone also agrees, he says, that sometimes we are free ("have the liberty") to behave as we choose. His "Part II" (pp. 384-388) reiterates the important point that the truth of both theses are required for human morality to make sense. It's a simple plan. No need to make your life overly complicated.

Where things do get complicated is where Hume proposes a specific candidate for the concept of "freedom" or "liberty" which, he claims, does make it possible for both of the key theses to be true. We are indeed exactly like falling rocks (1), Hume claims, as is every other part of the cause and effect mesh of the natural world. Nonetheless there is a sense in which we are still unlike falling rocks (2), in having a genuine claim to being free in our actions. In the language of Week Twelve, Hume is a "compatibilist" - there is a respectable concept of "freedom" or "liberty" which is compatible with the fact that all human behaviour is caused. This is his definition (pg. 383):

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

It may not be clear why this should be especially compatible with the fact of universal causation. Indeed, it may not be especially clear what it means at all. That is for next week.

Be careful right now about two words, however. Hume often uses the word necessity in the reading. Do not take it in any of the earlier senses you are used to (e.g. pp. 155, 83, 179). His argument is that when any philosopher proposes that "the occurrence of X is necessarily connected to the occurrence of Y", s/he is blowing hot air. There is really no deeper meaning for "necessity" to have as applied between items in the world (as opposed to applied as a relation of propositions say) than "X causes Y", where "cause" means solely the constant conjunction of X and Y. Likewise, be careful about Hume's use of the phrase "inference of the understanding", when he gives it as a second meaning of "cause" (e.g. pp. 376, 384). He is not thinking about "inference" as we took it up in Week Two - the rational grounds for inferring some conclusion from a set of premises (such as we have in deductive and abductive inference). He means instead the blind habit which the mind gets itself into of expecting Y when we observe X, the habit of simply passing from the thought of X straight to the thought of Y (his criticism, remember, of all inductive inference). The mind gets into this rut, naturally enough, whenever X and Y are constantly conjoined, so it is a natural enough extension of the basic sense of "X causes Y".

IV.B The Selection From Sartre

This isn't so difficult either. Provided you don't get pushed around by the more flamboyant slogans like "existence precedes essence" and "man is the future of man". The examples are where Sartre does the real work.

What do the examples show? Sartre insists they show the terrifying fact that we have no excuses behind which we can hide when it comes to our acts and plans and undertakings.

"If existence really does precede essence, there is no explaining things away by reference to a fixed and given human nature. In other words, there is no determinism, man is free, man is freedom.... [Once] thrown into the world, he is responsible for everything he does." (pg. 498)

And by "everything", Sartre really does mean everything. We may be subject to causal laws about our height and weight and the fact that we haven't got wings. But about every facet of human action - every facet - there is no place to lay responsibility except upon ourselves. Why? Because any attempt to find a causal explanation of why we performed one act rather than some other is really a lie - "bad faith" he calls it elsewhere. We choose what to do, we choose what to refrain from doing, we choose the advisors we go to for advice as to what to do, we choose whether to accept their advice or not, we choose whether to follow our emotions or our head in some matter - no exceptions to absolute freedom anywhere down the line. In effect, then, we basically are always choosing which causes will be operative on us in some situation. Causes don't happen to us, rearing out of some "causal nexus"; causes are always chosen by us, to be effective on us or not.

This impacts on the two theses. First, it implies that there are no exceptions to thesis 2 above. It's not just "some of our actions" which we perform of our own free will. It is all of them, every last one. Second, because of this there are no human actions for which thesis 1 is true. We are allowed to look for causality for everything in the universe, with the single exception of human action. To look for causation there is to look for a hiding place, to attempt to negate our absolute responsibility. As philosophers, then, we shouldn't even try to make freedom and causality "compatible". Causality is a concept applicable to one arena. Freedom is a concept applicable to a very different arena. It is a philosophical mistake - more, it is the only real moral sin - to blur those arenas.

IV.C The Selection From Skinner

Skinner is the direct opposite of Sartre. For Skinner, thesis 1 is true, because everything - every last event and process and entity - is subject to cause and effect. Human behaviour is therefore no exception. Because thesis 1 is true, thesis 2 must be false for every human action. There is no human behaviour which is exempt from the law that every event has a cause. Therefore there are no human actions which can free in any sense which opposes universal causation. Looking for the causes of human action is not merely not hiding behind an existential lie. It is the only scientific, and hence the only ultimately respectable, explanatory game in town. Skinner and Sartre wrote about the same time, but good thing they never met!

IV.C Causality and Determination

The problem of freedom requires a more careful attention to exact expression than most other philosophical problems. (David Hume complains bitterly about inexactitude here, pg. 375 etc.) This is especially true of the two main concepts whose apparent slip-sliding helps generate the problem in the first place, the concept of causation and the concept of determination.

Here are very rough definitions:

Causality

This seems - at first anyway - to be no more than the notion of following an empirical law of nature. Thus:

if X causes Y, then when X happens Y will happen.

Some alternative ways of putting it, perhaps: "If X causes Y, then were X to happen Y would happen". "If X causes Y, there is law of nature which states that where X happens we can reliably predict that Y happens".

Determination

This seems - at first anyway - to be a slightly stronger concept:

if X causes Y, then whenever X happens Y must happen.

If you list all of the causally relevant facts (the "full-Monty" X as it were), then these facts uniquely determine what will happen next. Sober puts this in several equivalent ways: "Y's future isn't left open by X's present state." Given X's present state, "there is only one option as to what will happen next." "A complete description of the system determines what will happen next". "If all the causally relevant facts are set out, these will leave open only one possible future for Y". To say the same thing, if X causes Y, then whenever X happens Y can't not happen. The universe "can't act differently from the way it does" as regards the occurrences of X and Y if X causes Y. If X causes Y and X happens, then it "can't fail" that Y happens. If X causes Y and X happens then "it isn't possible for it to be otherwise than that Y happens". [Incidentally, which of the three senses of "impossibility" from Week Three (pg. 179) is this? Or is it a new one we haven't met yet?]

I don't want to - I couldn't anyway - decide "yes" or "no" here for you. But I do suggest you become sensitive, prickly even, to the issue. So I will lay out here two columns and ask you to think about any differences and similarities between the parallel entries. (Some of the entries are drawn from Sober, others from the readings: Hume, Sartre and Campbell.)

Work away on the following table!

You may find this exercise hard, even impossible, to do the first time you read Lecture 23. But keep coming back to it, because the two concepts merit lots of chipping away at.

Sober is reluctant to put the central problem of freedom in terms of the concept of determination. He himself believes the real problem is freedom's connection with causality. Thus he says:

"... it is possible to have causation without determination. So my hunch is that the real problem isn't to see whether determination and freedom can be reconciled, but to see whether causality and freedom can be reconciled. However [he continues somewhat reluctantly], since the traditional positions about freedom all focus on deterministic causation, I'll do the same." (pg. 314, cf. 316)

Sober presents three reasons for thinking that the concept of causality and the concept of determination should not be collapsed into each other:

1. The implications of quantum theory (pp. 312-313). Here even a complete description of some physical systems involving elementary particles leaves open what its future will be like (it is indeterministic). This even though we might still be reluctant to say that what eventually happens in that system is somehow completely outside the realm of cause and effect (that is, the concept of causation is not put completely by the board - we don't say particle physicists work in the land of miracles for instance).
2. Sober's example of a surgeon who inserts a tiny roulette wheel into my normal deliberation processes (pp. 313-314). Here my deliberations suddenly become indeterministic (the present state of the system leaves open what will happen next). But of course roulette wheels follow causal laws as much as the rest of middling large physical objects, hopefully even in Las Vegas and Auckland, so the concept of causation continues to apply.
3. Dretske and Snyder's rather grim game of shooting the cat (pg. 314). In this example, Sober treats the roulette wheel as an indeterministic system. I spin it; if the ball lands on 00 a gun is caused to fire, killing the cat. Any death of the cat traces back to my spinning the wheel, for all that my spinning the wheel didn't make it inevitable that the cat would die. So I still caused the cat's death and the SPCA should hear of it.

In each of these examples, we seem to be able to cut more or less cleanly between the concept of causation and the concept of determination. One of the concepts applies, but the other doesn't. So they can't be - quite - the very same concept after all. We could add a fourth reason by using some materials from Hume (not all of the materials are available from the reading selection in Sober, however).

4. Hume argues for a specific theory of causation (an analysis of what causality really comes to): "X causes Y" means nothing else but "X is constantly conjoined with Y" (pp. 376, 382). He also argues - elsewhere - that the universe puts no a priori limits on what can be conjoined with what: there is nothing in the intrinsic nature of X that requires it to be conjoined with Y rather than Z - in effect, then, anything can be conjoined with anything else. Finally, as we saw in Week Six, Hume is completely sceptical about induction: the fact that X has always been conjoined with Y in the past provides no rational justification for believing that X will continue to be conjoined with Y in the future - i.e. that X and Y are indeed constantly conjoined; it provides only an occasion for our developing the blind habit of expecting them to be (pg. 382). The upshot of these three theses is this. Even if we agree that X was indeed the cause of Y, when causality and its implications are properly understood, we see that the occurrence of X doesn't cut off the options for what will happen next one teensy bit - much less tell us what must happen. Hence even if it is true that "X causes Y", and X does in fact happen, it makes no sense to claim that the occurrence of X determined the occurrence of Y.

In Hume the concepts of causation and determination can be firmly prised off each other. He does so himself (pg. 383), by challenging all comers to find anything more to their concept of "necessary connection" (i.e. determination) than is found in his concept of "constant conjunction" (i.e. causation).

Contents
Previous Week: Functionalism
Next Week: Compatibilism and Libertarianism

Content © 2000, 2001 Massey University | Design © 2000, 2001 Alun David Bestor | Any questions? Email the webmaster