1. Determinism
i. Initial conditions IC at play even before one's birth
ii. Laws DL which, given IC, always determine the same outcome
Example: law of falling body vs. probabilistic law.
NOTES:
(ii) says that events depend on and are determined by previous IC and  DL, not the converse, i.e. that if IC and/or DL had been different, the outcome would have been different.
Don't confuse Determinism with Fatalism
Although the laws of General Relativity are deterministic, those of quantum physics aren't.

2. Importance of issue of free will for ethics: praise and blame involve  responsibility, which involves free will.
Example: punishment as retribution presupposes responsibility, and hence free will.

3. Hard Determinism

1.  Determinism true
2.  Free will entails the ability to do otherwise
3.  Ability to do otherwise incompatible with determinism.
4.  Hence, determinism and free will incompatible.
5.  Hence, no free will.

NOTE: Lack of free will seems to lead to the denial of moral responsibility.  Hence, no retribution.  Incapacitation, deterrence and amelioration are the only components of punishment left.

A. Weak version:  Free will and determinism are merely compatible

1. determinism is true
2. free will is acting as one wants

NOTE: hence, real issue is not whether actions are caused, but whether they are caused by the agent's choice or by external causes.
3. hence determinism and free will compatible.
4. Objection: but if choice determined, then no free will.  Choice must be such that it could have been otherwise.

1. if determinism false, then action not caused by one's choice, i.e. the result of the process of deliberation.
2. hence, action uncaused, and consequently random.
3. I'm not responsible for a random action (e.g., arm's jerk due to link with a geiger counter).
4. hence, responsibility presupposes determinism.

Objection to (1):
determined deliberation is illusory.  It may have the phenomenology of deliberation, but it is not deliberation.
Hence, if determinism, then no choice.
Objection to transition from (1) to (2):
it's true that if action uncaused, then random, and then lack of responsibility.  However, action not caused by choice may be caused by the agent, the substance which constitutes the person.
NOTE.: link to substantial view of personal identity.

A. Attack on compatibilism and hard determinism:

1.  free will involves being able to do otherwise
2.  being able to do otherwise incompatible with determinism
3.  hence determinism and free will incompatible
4.  but we have free will because of evidence from introspection and our capacity to choose in cases of indifference of equilibrium
5.  hence, determinism is false.

Objection to (2):
Two senses of "could" in "could have done otherwise":

1. "could" refers to a causal impossibility, e.g. the outcome being different even if all antecedent conditions be identical.  This sense incompatible with determinism.
2. "could" refers to a causal possibility.  In this sense, "I could have done otherwise" must be understood as elliptic for "I would   have done otherwise if I had so chosen." This sense compatible with determinism.

B. Positive account of libertarianism:
First Argument

1. people are morally responsible for what they do
2. if determinism, then no free will, and hence no responsibility.
3. if indeterminism, then no responsibility.
4. but either determinism, or indeterminism, or agent causation is true
5. hence, agent causation is true

NOTE: agent causation is neither determinism, because action caused not by previous mental events, but by the substance constituting the person, nor indeterminism because action is not random (uncaused), but caused by substance.
Problem: If substance caused to cause action, then determinism; if substance   uncaused cause, then indeterminism and/or unintelligible.
Reply: desires, fears, hopes, deliberations, etc. reasons but not causes (analogy of the absolute ruler who can always disregard the advice and influence of his ministers).

Third Argument

1. We have a capacity to choose in cases of indifference of equilibrium.
2. But if determinism were true, then such choices would be impossible
3. Our choices are not random
4. Hence, agent causation is true

Problems:
• the only evidence we can choose in a state of equilibrium is introspection, which is not reliable
• even if we could choose in a state of equilibrium, the mechanism of choice may be determined (e.g. a subroutine which is called up when we are stuck)

Chisholm: Human Freedom and the Self

A) The metaphysical problem of human freedom is that it seems incompatible both with event-determinism and geiger-counter indeterminism because free will involves the capacity to do otherwise (COD).

B) Compatibilists attempt  the following reduction:

“he could have done otherwise” (P) is elliptical for “he would have done otherwise if he had so chosen.” (Q)

Problems:

• If he could not have chosen  otherwise, (Q) might be true and (P)  false.  Hence, (P) and (Q) don't mean the same thing.
• Compatibilists confuse the real issue, which deals with the actus elicitus, with a spurious one about the actus imperatus.

1. Transeunt causation, i.e. event causation, in which events cause events
2. Immanent causation , i.e. agent  causation, in which agents cause events

D) A taxonomy of action:

• Facere esse = to bring it about
• Non facere esse = to abstain from bringing it about
• Facere non esse = to prevent from having it come about
• Non facere non esse = to abstain from preventing from having it come   about.

Dennett: I could not have done otherwise: so what?

Three types of "could have done otherwise" CDO:
 i.  metaphysical (useless and unknowable)
 ii.  local fatality (a misnomer), e.g., he couldn't have done otherwise because he was locked in the room (useful and knowable)
 iii. character improvement, e.g., he misbehaved, and what's bad is that he could have done otherwise (useful and knowable)

Libertarians invoke CDO (metaphysical version) in context of responsibility, but without good reasons because:

1. The way we consider issues of responsibility in everyday life makes no appeal to CDO (metaphysical). The only form of CDO relevant to moral responsibility arises in the context of “local  fatalism”, as when we say: ”He could not have done otherwise no matter what he had tried or wanted to do because he was in chains."  But local fatalism is neutral with respect to determinism or lack of it (we don't ask physicists or neuroscientists to determine whether someone is morally responsible)
NOTE: This is the standard compatibilist position.
Objection: if determinism, then no chance or opportunity of doing otherwise.
Reply: Not so.  Chances of winning or losing at lottery are the same whether the extraction takes place before or after the sale of tickets.
Duplication:  Hard to see the relevance of the reply; determinism deprives me of CDO not because the cause is before the effect (retrocausation, if possible, would do the same), but because the cause is taken to be complete, i.e. sufficient to bring about the effect.

2. “Here I stand; I can do no other;”  “I could not possibly torture an innocent person for fun”
These are avowals, not disavowals, of responsibility; they indicate that given the person I am (my character), I couldn't do otherwise.
Objection: then you're like a zombie; people can act out of character, as it were.
Reply: Not so; I can see both sides, but the reasons for the course of action I choose are overwhelming. Flexibility requires only that I recognize that in different circumstances I could act otherwise (e.g. torture to save the world).
NOTE: here Dennett close to Hume: responsibility requires determinism

3. If metaphysical CDO were relevant to moral responsibility, we couldn't know whether I'm responsible or not because of ignorance  of whether my actions are macroscopic effects of quantum-level indeterminacy or determined effects because quantum indeterminacies cancel each other out.

4. The significance of knowing CDO (metaphysical) about a subject would be nil because:

• if the subject is determined and therefore could not have done otherwise (metaphysical), I don't know how he will react in same situation because it's very unlikely he will be in  same psychological state ever again.
• if the subject is not determined and therefore could have done otherwise (metaphysical), I don't know that an analogous determined subject won't show as much flexibility, sensitiveness to nuances etc. as undetermined subject.

Problem: True given the finite capacities of the brain? (Sphex example)
Reply: in practical terms, probably yes