MAR 15

15: MAR 15

Dennett, Daniel. Chapter 3: "Thinking about Determinism" Freedom Evolves. (300) (63-97)

1. We think determinism limits what is possible. (Austin's putt)
2. We assert that S(0) causes or explains S(t). This misses the way causal inquiry works. (Computer marathon / JFK)
3. Determinism rules out self-directed change, change in character or "life-hopes". (Closed/open futures)

Dennett believes his arguments in showing these errors apply to both det and indet worlds.

Possibility -- Whatever isn't "necessarily not" the case. Roughly, all of the possible differences one might imagine between worlds. (Informal and identification predicates come in here.)

Determinism -- There is at any instant one possible future. In possible worlds talk, "A world is deterministic if it has the property such that, if it shares the same S(0) with any other world at time 0, it will share S (the same state description) at t." Determinism is about causal sufficiency, not necessity.

Causation - How do we assign causes? (Note: Laplace's demon knows state descriptions not causes!)

"Causal necessity" - without A, C would not happen. Had Bill not tripped Arthur, he would not have fallen. In all of the possible worlds in which Bill trips Arthur, he falls.
"Causal sufficiency" - A is sufficient to cause C, but other antecedents might as well. Arthur's fall is an inevitable outcome of being tripped. In any world in which Bill trips Arthur, Arthur falls.

Evaluating counterfactuals (as in the causal necessity example) requires establishing a "comparison set X" of worlds approximately similar to ours in which tripping Arthur leads to his falling. The selection of the comparison set is crucial.

In assigning causes, we also typically assume "independence" and "temporal priority". But here are some cases to show how various considerations are used to assign causation in different contexts:

The Sharpshooter Case -- The sharpshooter has a low probability of hitting the target, but does. We favor 'causal necessity' over sufficiency in this case in saying he cause the death.

The King and the Mayor (overdetermination) - Both issue exile orders for someone. Neither is necessary. Pick one, maybe the king?

Billy and Susie - Billy's rock is sufficient to cause a bottle to break, but Susie's gets there first. We favor temporal priority in assigning the cause.

French Foreign Legion case -- a series of "but for" causes, all of which are sufficient. Which is the cause?

narrow method for choosing comparison set X - worlds identical to Austin's prior to his putt. If you choose the set this way, Austin could not have made the putt (looks like determinism eliminates the possible). But you could choose a slightly different comparison set and in some of those worlds, Austin makes the putt
It follows that even in det world it makes sense to say that he might have made the putt.
Austin seems to choose the narrow method, but equivocates about "further experiments" (which imply changing the antecedent conditions).

Conclusion (77): The truth or falsity of determinism should not affect our belief tha tcertain unrealized events were nevertheless "possible," in an everyday sense of that term. [Well, possible worlds talk isn't exactly everyday. ]

random number generators. To generate variations in the play, we introduce slightly different conditions.
With this random variation, you find the A beats B a thousand times in a row. It would not be explanatory to say A was caused to beat B. You have to go up to the design or intentional level to explain A's behavior.

Could B have castled? You have to do the analysis. Might look at related possible worlds and say it was a fluke B didn't. Or you might find that B would have found the option if he's been coded a bit more efficiently. Point: (82) Philosophers choose the narrow set (Could I have done something different in exactly the same universe as I am in?) when thinking about determinism and free will, but no one seriously investigates possibility and causation that way.

Determinism is about causal sufficiency, not necessity. The actual universe at S(0) was sufficient to lead to JFK's death, but we don't know if it was necessary. Note that we wouldn't say that S(0) caused JFK's death.

coin flips have "no cause" even though they occur in a deterministic universe. Note details. Important thing is to create conditions that make prediction impossible. In a sense the coin flip amplifies micro-variaitons and thereby reduces necessity.

Randomized Control Trials and Randomized Experiments. Use "uncaused events" break the influence of patterns we want to exclude for purposes of the experiment (and to determine causation).

88: Why do we focus on necessity if it confuses us about free will? Our rationality requires it. Example of man falling down elevator shaft. Landing is inevitable, maybe dying isn't. We can change the future because evolution designed us that way. We have search algorithms, we are "anticipator-avoiders" (who look for necessary relationships). The fatalists lose in the the evolutionary competition.

Third Error: Determinism rules out self-directed change, change in character or "life-hopes".

Whether the future is Open or Closed is independent of determinism / indeterminism. Things can be "determined to change" "In some deterministic universes there are things whose natures change over time, so determinism does not imply a fixed nature."