Moral accountability under Compatibilism
Compatibilism attempts to account for free will in a manner consistent with determinism. This stands in contrast to Libertarian Free Will, which denies determinism. Libertarians point to the Principle of Alternative Possibilities (PAP), insisting that one can only be considered responsible for an act if one could possibly have chosen to act differently. If determinism is true (they say), then we could not have chosen differently – and therefore no one is responsible for their behavior. In this post, I will show that accountability is consistent with compatibilism.
By definition, a choice is a behavior in which a person has multiple options before them, and selects one. Choices are plausibly consistent with compatibilism because they are the product of a person's memories, beliefs, dispositions, and impulses. At a point in time, memories, beliefs and dispositions are fixed. Impulses imply a seemingly random element to the choice, but not actually random. The impulse has a basis in one's attitude at the moment (e.g. optimistic, pessimistic, anger, happiness...) or related to an internal or external factor that triggers a transient memory or feeling. These factors plausibly collectively determine the decision - no alternative decision could have been made given the specific set of memories, beliefs dispositions, and impulses that were present at the time of the decision. For purposes of this discussion, we'll assume that there were deterministic causes of the prior memories, beliefs and dispositions as well (in general, a combination of nature and nurture).
Consider this scenario: a parent raises a child with a lack of discipline, essentially letting the child do whatever she pleases while shielding the child from any negative consequences. The child reaches adulthood and behaves irresponsibly, drinking to excess and driving. One day, the adult child is driving drunk and hits a bicyclist; the adult child drives away leaving the bicyclist on the road to die. The drunk driver chose to drive while drunk, and chose to drive off after hitting the bicyclist. These decisions were plausibly the deterministic result of her beliefs and dispositions. Under this assumption, is this adult child responsible, or is the parent who failed to teach the child discipline and responsibility? Does the prior cause (the poor parenting) negate the adult child’s responsibility? The driver’s sister testified that the poor parenting should at least mitigate responsibility somewhat. But I expect we all agree that the adult child is responsible and should be held fully accountable, even if we also bear the mother responsible for her failure as a mother. (This actually happened in Houston a few years ago).
With the above scenario in mind, here's two compatibilist accounts of responsibility:
1.The natural reaction to hearing about the drunk driver killing the bicyclist is a reactive attitude that the driver is guilty. In most cases, a perpetrator has a feeling of guilt after recognizing a consequence of a bad choice (e.g. the girl expressed this to friends, and it was these friends who reported the crime).These morally reactive attitudes are the basis of our moral responsibility practices. They are natural responses, not mere social convention. They are an aspect of interpersonal relations and expectations, and of our natural internal feelings. It is inconceivable that we would stop holding such people morally accountable or stop feeling guilty, even if it were somehow proven that determinism is true. Indeed, the fact that we have these attitudes contributes to our behavior, because we generally prefer to avoid guilt and social approbation, and enjoy pride and respect.
2. Could the drunk driver have done differently? Yes she could have, if she had held the strong belief that the risk of driving drunk was so great that it outweighed her impulse to do so. This could only have occurred had there been something different about the past (formation of that belief), but that's reasonable. Our choices are fully explainable as being the result of our personal beliefs and dispositions. As a thought experiment: In the actual world, you are presented with a choice between X and Y. You deliberate on the options, weighing pros &cons consistent with your background beliefs and dispositions, and you ultimately choose X (possibly influenced by some sudden impulse). Is there a possible world with an identical history to this one, so that you have exactly the same background beliefs, desires and impulses at the point at which the choice is presented - but you instead choose Y? If yes, then your choice is made for no reason (this seems to be what LFW gets you). If no, then your choice has been caused (consistent with determinism).
When she is released from prison, let’s hope the drunk driver will actually have learned this, and doesn’t repeat the risky behavior. Our beliefs and dispositions are part of what we are - we own the results, and this makes us accountable. We can learn new beliefs, and these will influence our behavior.
#1 and #2 are more or less independent, but in tandem they provide not only a coherent account of moral responsibility, they also explain why normal functioning people strive for generally moral behavior. We want to avoid guilt, fit in, and we want to avoid approbation by others. We CAN always do better, but it requires learning things. Social consequences (positive and negative) and internal guilt/pride provide incentives to learn what is needed to behave morally.
By definition, a choice is a behavior in which a person has multiple options before them, and selects one. Choices are plausibly consistent with compatibilism because they are the product of a person's memories, beliefs, dispositions, and impulses. At a point in time, memories, beliefs and dispositions are fixed. Impulses imply a seemingly random element to the choice, but not actually random. The impulse has a basis in one's attitude at the moment (e.g. optimistic, pessimistic, anger, happiness...) or related to an internal or external factor that triggers a transient memory or feeling. These factors plausibly collectively determine the decision - no alternative decision could have been made given the specific set of memories, beliefs dispositions, and impulses that were present at the time of the decision. For purposes of this discussion, we'll assume that there were deterministic causes of the prior memories, beliefs and dispositions as well (in general, a combination of nature and nurture).
Consider this scenario: a parent raises a child with a lack of discipline, essentially letting the child do whatever she pleases while shielding the child from any negative consequences. The child reaches adulthood and behaves irresponsibly, drinking to excess and driving. One day, the adult child is driving drunk and hits a bicyclist; the adult child drives away leaving the bicyclist on the road to die. The drunk driver chose to drive while drunk, and chose to drive off after hitting the bicyclist. These decisions were plausibly the deterministic result of her beliefs and dispositions. Under this assumption, is this adult child responsible, or is the parent who failed to teach the child discipline and responsibility? Does the prior cause (the poor parenting) negate the adult child’s responsibility? The driver’s sister testified that the poor parenting should at least mitigate responsibility somewhat. But I expect we all agree that the adult child is responsible and should be held fully accountable, even if we also bear the mother responsible for her failure as a mother. (This actually happened in Houston a few years ago).
With the above scenario in mind, here's two compatibilist accounts of responsibility:
1.The natural reaction to hearing about the drunk driver killing the bicyclist is a reactive attitude that the driver is guilty. In most cases, a perpetrator has a feeling of guilt after recognizing a consequence of a bad choice (e.g. the girl expressed this to friends, and it was these friends who reported the crime).These morally reactive attitudes are the basis of our moral responsibility practices. They are natural responses, not mere social convention. They are an aspect of interpersonal relations and expectations, and of our natural internal feelings. It is inconceivable that we would stop holding such people morally accountable or stop feeling guilty, even if it were somehow proven that determinism is true. Indeed, the fact that we have these attitudes contributes to our behavior, because we generally prefer to avoid guilt and social approbation, and enjoy pride and respect.
2. Could the drunk driver have done differently? Yes she could have, if she had held the strong belief that the risk of driving drunk was so great that it outweighed her impulse to do so. This could only have occurred had there been something different about the past (formation of that belief), but that's reasonable. Our choices are fully explainable as being the result of our personal beliefs and dispositions. As a thought experiment: In the actual world, you are presented with a choice between X and Y. You deliberate on the options, weighing pros &cons consistent with your background beliefs and dispositions, and you ultimately choose X (possibly influenced by some sudden impulse). Is there a possible world with an identical history to this one, so that you have exactly the same background beliefs, desires and impulses at the point at which the choice is presented - but you instead choose Y? If yes, then your choice is made for no reason (this seems to be what LFW gets you). If no, then your choice has been caused (consistent with determinism).
When she is released from prison, let’s hope the drunk driver will actually have learned this, and doesn’t repeat the risky behavior. Our beliefs and dispositions are part of what we are - we own the results, and this makes us accountable. We can learn new beliefs, and these will influence our behavior.
#1 and #2 are more or less independent, but in tandem they provide not only a coherent account of moral responsibility, they also explain why normal functioning people strive for generally moral behavior. We want to avoid guilt, fit in, and we want to avoid approbation by others. We CAN always do better, but it requires learning things. Social consequences (positive and negative) and internal guilt/pride provide incentives to learn what is needed to behave morally.
Comments (78)
The intoxication argument is an interesting one in general. You decide to get intoxicated, although sometimes it happens by accident, you never intended to have more than a safe amount to drive home on. Your consequent decision to drive after having a skin full because it was a mates birthday at the pub and you'd forgotten was actually made under the influence of alcohol. So there is a question of causal influence, alcohol influence which makes it even more complicated..
Right, IF her past had been different, she would have been raised in such a way as to potentially care and emphasize moral values. However, in the scenario you provide, she was not, in fact, raised that way. It is possible that someone having experienced such neglect as a child might be incapable as a practical moral-psychological matter of forming the beliefs and convictions about appropriate behavior necessary to avoid the crash, unless one assumes some intervening corrective event to mitigate the neglect. Since there was no such corrective event in the hypo given, as a result, she was never reasonably in a position to develop such beliefs and convictions and, therefore, should not be held responsible for her actions stemming from not having them.
Quoting Relativist
I think you would have to say more about how you are imaging possible worlds in order for this hold. It doesn't seem to me given that the mere existence of an alternative timeline that differs only with respect to its final event (choice X vs. choice Y) implies that there could be no reason for either choice. You don't even really need possible worlds here. Suppose a person has equally good reasons to make some decision, X, as they do to make some decision, Y. Since the reasons don't preponderate in either direction, they make an "executive decision" to do Y. The decision is arbitrary to some extent, but the fact that it's arbitrary doesn't imply a person has no good reason to do it; they have all the same reasons they had when they were weighing those reasons against the countervailing reasons.
The conclusions of the thought experiment seem to be committed to the view that the only appropriate choice or decision is the one that has a preponderance of reason in its favor. Clearly that will be the best choice, but it's not clear that it's the only rational choice. In other words, the implication is that the only rational choice is the best choice, and I'm not sure that's necessarily true.
I was giving one example of a difference in the past that might have made a difference. For example, a near miss where she almost kills someone or herself.
Quoting MentalusionNo. I mentioned that the choice was a product of beliefs, disposition, and impulse. In impulsive choice is not rational, but the impulse is the reason for it.
Regarding an "executive decision" - I suggest this is due to nonlinear logic/pattern recognition/insight. This is the process where we recognize a solution to a problem without having arrived at it deductively. Depending on the circumstance, it can be subconsciously influenced by desires, false beliefs (like a racist jumping to conclusions about something done by a member of the hated race), or by a positive or negative attitude. Our judgment can be impacted by any number of things - becoming angry due to a perceived slight (getting cut off in traffic, arguing with a spouse about money, grief...). For this reason, I used possible world semantics to conceptually wind back the clock to the point of decision, keeping every possible factor identical.
Understand that I'm not suggesting Libertarian Free Will is impossible or incoherent. I'm just showing that compatibilism is also coherent and plausible, it just depends on a different conceptual framework.
One incompatibilist argument is to claim that if LFW is false, then the bicyclists death was entailed by the state of the universe when the big bang occurred: everything that follows is deterministic, so there were no true choices and thus no one to blame. This overlooks the role of the direct cause, and I'm explaining that we naturally cast blame to the direct cause, and that it is reasonable to do so.
Of course, keep in mind that I'm of the view that compatibilism can't be made coherent in the first place.
Which is quite an extreme view really when you think that incompatibilism has been largely abandoned.
I still think that argumentum ad populums are fallacies.
Thanks for the comment. It reminded me that I failed to include a preliminary description of a compatibilist choice. I have added that - it's now the second paragraph of the Op, and I'll repeat it here:
By definition, a choice is a behavior in which a person has multiple options before them, and selects one. Choices are plausibly consistent with compatibilism because they are the product of a person's memories, beliefs, dispositions, and impulses. At a point in time, memories, beliefs and dispositions are fixed. Impulses imply a seemingly random element to the choice, but not actually random. The impulse has a basis in one's attitude at the moment (e.g. optimistic, pessimistic, anger, happiness...) or related to an internal or external factor that triggers a transient memory or feeling. These factors plausibly collectively determine the decision - no alternative decision could have been made given the specific set of memories, beliefs dispositions, and impulses that were present at the time of the decision. For purposes of this discussion, we'll assume that there were deterministic causes of the prior memories, beliefs and dispositions as well (in general, a combination of nature and nurture).
The drunk driver chose to drive while drunk, and chose to drive off after hitting the bicyclist. These decisions were plausibly the deterministic result of her beliefs, dispositions, and impulses of the moment. The question I was addressing in #1 and #2 pertained to whether or not a plausible account of moral accountability could be provided.
Quoting Terrapin Station
I'm not making an argumentum ad populum. I'm noting that each of us has a natural reaction to such deeds as I've described, and it is these natural reactions that are the basis for assigning responsibility.
Then it's not actually a choice and not compatibilist. There's no actual (ontological) freedom involved.
The argumentum ad populum comment was in response to Jamesk saying "Which is quite an extreme view really when you think that incompatibilism has been largely abandoned." In other words, supposing it's been largely abandoned only matters if you're swayed by the crowd for its own sake.
Then it's not actually a choice and not compatibilist. There's no actual (ontological) freedom involved. [/quote]
It is a choice by definition (a choice is a behavior in which a person has multiple options before them, and selects one), and I explained how it is consistent with determinism, so it is compatibilist.
You don't have multiple options if it's the case that "no alternative decision could have been made."
Multiple options are available, and one is chosen. This is the case irrespective of whether or not compatibilism is true. This is indisputable.
I described what is involved in the selection process (i.e. it is a product of a person's memories, beliefs, dispositions, and impulses), and this is also irrespective of whether or not compatibilism is true. Do you disagree? Am I omitting something?
Yes, but some deny there is moral accountability if choices are the product of determinism. I was addresssing that in my Op.
Not if "no alternative decision could have been made."
What situation could you envisage that would offer no alternative courses of action? Unless you have unwillingly and unknowingly been 'possessed' or 'taken over' in some way then there are always alternatives. An unattractive alternative is still an alternative.
"no alternative decision could have been made" was in quotation marks because I'm quoting Relativist.
But doing a bad thing is never an act of free will.
Alternative courses of action exist, but only one is selected. Under my compatibilist account, the one selected is determined by the chooser's memories, beliefs, dispositions and impulses. Given those specific memories, beliefs, etc - no other decision could have been made.
It seems reasonable to think those factors determine our choices. If these don't determine our choices, then we're making choices for no reason. But hypothetically, we could make a different choice if we had a different belief, memory, disposition, or impulse.
The thesis here being defended by @Relativist is purportedly compatibilist while resting on the rejection of the principle of alternative possibilities (PAP). If you construe a possible "alternative decision" as an "alternative possibility", which is consistent, that is, with the past state of the universe and with the laws of nature, then, there are indeed no "alternative decisions" that are genuinely open to an agent if determinism is true.
I take it, however, that when Relativist speaks of alternatives, s/he is speaking of a range of options that merely appear open to the agent, for all she knows; since a deliberating agent never (or very seldom) is in an epistemic position where she would know in advance what decision she is being predetermined to make.
Hence, Relativists's position could properly be termed a form of semi-compatibilism. John Martin Fischer argues for such a position, claiming that determinism is compatible with moral responsibility although it precludes alternative possibilities. Depending on whether you define "free will" as requiring the truth of PAP, or as resting only on the criteria of moral responsibility, you might say that semi-compatibilism doesn't or does, respectively, allow for the compatibility of determinism and "free will". This issue becomes a matter of mere semantics.
Yeah, i would say that's just determinism, then. I don't see it as a semantic issue, really. I don't think it matters what we call anything. I just don't see how we can have both ontological freedom and ontological determinism at the same time (whatever we call them).
You dont seem to be disagreeing with the thesis Relativist presents in the original post, then. S/he is arguing that personal and moral responsibility are compatible with determinism. This is not the sort of compatibilism that you are disagreeing with.
I mildly object to saying a decision is predetermined. Saying the decision was "predetermined" can be interpreted to mean the same decision would be made irrespective of the cognitive processes the agent engages in. I stress that the agent's specific cognitive processes were necessary to the reaching of the decision, even though no other decision could have been made given the full set of characteristics of the agent. This is relevant to avoiding fatalism. An agent's role is an active one.
Yes, I quite concur with your mild objection ;-) But if it is right, it also threatens the cogency of the doctrine of determinism, on my view. On your view, on the assumption that determinism is true, the agent's decision is being jointly determined by the external constraints that the agent is being subjected to and, also, by the intrinsic characteristics of the agent. Hence, since the agent's own character and cognitive processes are involved in the determination of her behavior, her personal responsibility for her actions (and hence, also, the propriety of reactive attitudes towards her, such as gratitude and resentment) aren't necessarily threatened by the (alleged) falsity of the principle of alternative possibilities.
But that means, also, that what is usually being construed by hard determinists and by libertarians alike as "the past" isn't entirely removed from the scope of the responsibility of the agent. It is, in a way, through the ongoing process of practical reasoning (and the present operation of all the cognitive processes enabling the power of practical reasoning) that "the past" gets funneled into intentional action and hence that choices are being made by the agent between her various (seemingly open) opportunities. This process would be deterministic (and hence only one option would by genuinely open at any given time) only if there were deterministic laws of nature governing what intelligible actions follow from an agent's 'total' circumstances (i.e. her external circumstances and opportunities, and her intrinsic cognitive characteristics). I am prepared to argue that there can't possibly be any such laws. There can't be any such laws irrespective of there being, or there not being, deterministic laws that govern the evolution of sub-personal neurophysiological processes. (My view of the causality of rational agency is a combination of substance causation and of rational causation, rather similar to those defended by E. J. Lowe and by Eric Marcus, broadly following Aristotle and Kant, respectively.)
If we assume that the agent's decisions necessarily have some rational connection to the cognitive processes the agent engages in (I don't think this is at all the case, but we can pretend that it is), and we say that the decisions follow from those cognitive processes in some deterministic way (stemming from that "ideal" rationailsm), then the question is simply pushed back to whether the agent's cognitive processes are determined or not. If they are, then effectively, any decision is predetermined and we're not talking about compatibilism. If they're not, then determinism isn't the case and again we don't have compatibilism.
Yeah, I don't agree that ontological freedom/determinism has any necessary implication for anyone's view of culpability (aside from "if determinism is the case, then S necessarily has x view re culpability"). That issue is simply semantic in the sense that it simply depends on the concept of culpability that an individual holds.
Sure - and they ARE effectively predetermined. I'm drawing the distinction between entailment and causation. Per determinism, the decision is a truth that is entailed by the truths at the big bang. The logic parallels the causation: Big bang truth ->entails a logical chain of truths->entails the truth of the decision. The transitive property applies to the logic, so it's valid to say: big bang truth ->entails truth of the decision. Although this is valid logic, causation unfolds in a temporal sequence and each step in the sequence is necessary to the next (i.e. the transitive property does not apply to the causal sequence). This means we are warranted in considering the necessary role of the immediate cause of the decision.
But then that's not compatibilism, because you have no ontological freedom in your ontology. You're just saying something about culpability.
Could the agent have decided differently? Yes, if there were some difference in the factors contributing to his decision. This is sufficiently free to be classified as "free will," and sufficiently free to be held morally culpable.
Is that not free enough for you? Do you insist that true freedom entails being sufficiently free to make a different choice given exactly the same set of deciding factors? That seems absurd - because it implies a freedom to make choices for no reason at all.
I was just about to post a similar response/challenge to @Terrapin Station. On the one hand, I am myself a libertarian incompatibilist, just like TS. But my position is distinctive from what I like to call "rollback libertarianism'. Rollback libertarianism rests on one particular construal (which I dont endorse) of the principle of alternative possibilities (which I endorse!), but which is endorsed by many libertarian philosophers such as Robert Kane (with some caveats). Under this construal, it must be possible for a free agent who actually did A that she could have done something else (or merely abstained from doing A, or could have done A differently) in the exact same 'circumstances' in which she actually did A, where those 'circumstances' (so called) are construed as including her own actual mental states and proclivities up to the moment of decision. I think you are absolutely right that this sort 'contra causal' criterion for freedom seems to threaten the intelligibility of rational action and hence to undermine what it seeks to salvage.
But that's just changing what we're referring to in the conversation. No one in the debate was using "free" to refer to whether a choice is a product of the agent's mental processes or not.
So it's not compatibilism, it's "redefining what the words are referring to so that we can use both of them in conjunction with each other."
Quoting Relativist
That's like saying, "Could nuclear bombs produce flowers instead? Yes, if atoms behaved differently."
Quoting Relativist
But that's what ontological freedom is.
I don't see the point as trying to "salvage" anything. We're simply wondering whether ontological freedom obtains in relation to "will phenomena," so that more than one option is a possible consequent state given identical antecedent states. It seems to some of us that such ontological freedom does obtain. We're not campaigning for anything in this, not issuing value judgments about anything, etc.
That's why I like to use for my decision examples options that I choose between a la flipping a coin. It seems as if I can make those sorts of choices.
I wonder who you take to be the proper authority for the definition of the concept of freedom. Most people who are party to the conversation (which I take to include philosophers engaged in the debate about free will and determinism) acknowledge various conceptual and/or constitutive connections between the concepts of freedom, desire, rationality, belief, intention, praise and blame, responsibility, etc.
You're barking up the wrong tree re asking me who I'd take to be a "proper authority" re something so broadly discussed.
I'm simply referring to the conventional conversation re focusing on the ontological question re whether one sort of phenomena or another (can) obtain in conjunction with will.
So you would call an act independent of and maybe even contrary to any conscious decisions an act of free will?
No. It has to involve will (which is conscious), or we're just talking about ontological freedom in general.
Very well. But then very many philosophers might inquire whether such a very weak concept of 'ontological freedom' (which seemingly amounts to nothing more than the the mere indeterminism or randomness secured by quantum mechanics) has any connection at all with the traditional philosophical questions regarding rational agency and freedom of the will.
What would be a "robust concept" of ontological freedom (versus determinism)?
"Ontological freedom" is your term. @Relativist and I had issued a challenge for your own conception of what it might be, since it appears to us to amount to nothing over and above mere indeterministic randomness in the production of bodily movements, while severing their connections with the will construed as a psychological faculty.
I'm not sure I can follow this distinction. Even a quark jumping around wildly and a-causal cannot be called free without raping the concept. Trancendental freedom is defined in terms of subjectivity. Alledged or not.
It's not a term I invented, but "weak conception" was your characterization, so presumably you had some idea of a non-weak conception of ontological freedom.
If we're talking about free will it's obviously not severed from will phenomena. But yes, the issue (freedom a la free will vs detereminism) is only coherent as wondering about whether it's possible for at least two different consequent states to follow the same antecedent state.
What in the world? Obviously you'd have to explain "cannot be called free without 'raping the concept'" and "transcendental freedom (defined in terms of subjectivity)"
At the moment those just look to me like random words thrown together.
Fair enough. But then, it would appear to be part of the conventional conversation about the will (conceived as the faculty that issues in intentional actions) that it involves some sorts of intelligible connections between antecedent mental states of agents and their subsequent actions. It is true that those connections generally are believed not to be fully determinative. But then, the challenge for the libertarian incompatibilist is to explain how the mere lack of full determination of actions by antecedent psychological states (and its replacement by mere randomness) is helpful at all to the ordinary conception of personal freedom.
Ignoring whether there's some set of utterances, where we've empirically established the commonality of the same, that we're calling the "ordinary conception of personal freedom," the idea of any of this needing to be "helpful" towards it is exactly what I was talking about NOT doing re "not campaigning for anything, not suggesting value judgments or normatives" etc.
I think it is obviously coherent. For instance, both the radioactive decay or the absence of radioactive decay of an atom are consistent with the same antecedent (non-decayed) state of this atom. Likewise, regardless of the indeterminacy of fundamental physical laws, the modelization of cognitive processes as coarse-grained outcomes of a chaotic dynamical system make different actions consistent with the very same (coarse-grainedly) defined antecedent mental states. Both those possibilities, which are intelligible from the physical point of view, seem not to secure the kind of freedom that we intuitively ascribe to the will.
What different sort of freedom would you say we intuitively ascribe to the will? (Do I do this if I don't know the answer to it)?
The common intuitive conception of free action, which seems to me to be broadly correct, rests on a form of agent causation rather than event causation. Agent causation tends to give Humean philosophers headaches, while Aristotelian philosophers account for it more easily. Free actions are actualizations of the powers of practical rationality possessed by mature rational animals. Hence, the causal antecedents of free actions are substances (e.g. rational animals) and their freely (and responsibly) endorsed reasons rather than fixed antecedent circumstances and blind laws of nature.
Okay, that makes some sense, I suppose, although "perceiving" seems like a weird word to use, and "conscious control over yourself" seems redundant in a way that I don't think makes much sense.
Without getting into the many issues I have with that, it doesn't seem to be specifying a different sort of freedom, with a focus on what exactly "free" is referring to ontologically in different cases. It seems to be explaining all sorts of things but the "free" part.
For example, free action via agent causation versus free action versus event causation. That's fine as a distinction re the source of the free action (ignoring analysis of how we got there in each case, at least), but it's not describing a different kind of freedom in each case, is it? If so, what's the difference in what's going on in the freedom part?
For example, "Playing a CMa7 chord on a keyboard via a human pressing the keys versus playing a CMa7 chord on a keyboard via a sequencer triggering the sounds." There's a difference there, but not in the "playing a CMa7 chord" part.
You may construe it as a recasting of the philosophical concept of freedom. It is indeed the very same intuitive concept of freedom of agency that grounds the reactive attitudes of ordinary people, but it is being recast away from its usual distortions by scientistic, crypto-Humean and crypro-Cartesian prejudices. It does not require rollback indeterminism (that is, the possibility that one may have done otherwise in the exact same antecedent 'circumstances'), but it doesn't accommodate determinism either since the antecedent 'conditions' (often construed as a set of atomic Humean 'events') of the agent become irrelevant to the determination of the agent's action by herself, in accordance only with her reasons, good or bad, for doing whatever she is doing.
Let's try it this way.
On your account, we have, in temporal order
(a) the antecedent conditions of the agent
(b) the agent's reasons for doing x
(c) the decision based on (b)
Now, was (c) determined by (b), or was freedom involved somehow between (b) and (c), and was (b) determined by (a), or was freedom involved somehow between (b) and (a)?
(By the way, (b) is actually what I'm calling the "antecendent" and (c) is the consequent, but I'm guessing you know that and there's a reason you're inserting an extra step, which is fine)
Let me respond to this first, for the sake of clarity. When I'm speaking of "antecedent circumstances" I only mean to signify a temporal relationship between material states of affairs (or the so called 'state of the universe' at a time) and the agent's subsequent decision or action. The latter is being causally determined by the former, according to determinism.
Quoting Terrapin Station
On my view both (a) and (b) can figure in the explanation of the agent's decision (or intentional action) albeit in different ways and for complementary explanatory purposes. (b) is, however, ineliminable as part of the explanation of the action as an intelligible occurrence in the life of a rational agent. It is also not something that can be construed as an event in time. It is rather a rational consideration that the agent can adduce in order to justify her action to herself or to others.
Now, regarding the locus of freedom: (b) is fully determinative of (c), as long as (c) is an intentional action (or the formation of an intention). But since this determination occurs as the actualization by the agent of her own powers of practical reasoning, it isn't being determined by (a). There nevertheless are causal relations between (a) and (c) but they aren't determinative. On my view, the causal relations between (a) and (c) are best construed as enabling conditions. Some of those causal antecedents account for the agent having acquired practical rational abilities (or a free rational will) in the first place. Other causal antecedents account for the various opportunities and powers that the agent has prior to the moment of decision. But what it is that determines what the agent does, in those circumstances, is the agent herself on the basis of rational considerations (b), which may be good or bad, and hence make the agent liable to be praised or blamed (or proud or ashamed, or happy or regretful) for her decision.
?? On my view the idea of there being anything divorced from time (and space/location for that matter) is incoherent.
So on my view it's incoherent to say that (b) isn't particular events in time, with a location in the world, and more specifically, that (b) isn't dynamic processes of material stuff (namely the agent's brain). It doesn't seem like you're talking about this, though. It sounds like on your view, the agent's rationality is some mysterious who-knows-what that's not part of the material world and that can somehow operate independently of it?
You need not entirely divorce rational considerations (or numbers, say) qua abstract objects from space and time in order to place them in a different metaphysical category than either concrete material substances or material events. Rational considerations clearly relate to time, in a sense, since they can be entertained and evaluated by agents at specific times and not others. But that doesn't make them suitable to being themselves located in particular places or times (any more than numbers or premises of arguments can be). That would be a category error.
I am acknowledging the existence (and necessity) of some such dynamical processes as enabling conditions (and hence part of (a)) for an agent grasping (b).
Rational abilities aren't mysterious who-knows-what. They are abilities to appreciate and be motivated by rational considerations. They are abilities that are being inculcated though normal education just like, say the ability to count, to answer to one's given name or to play chess. But they are distinguishable from the material processes that implement them since they are defined abstractly by reference to what it is that they are abilities for and, furthermore, they are multiply realizable.
Which is Hume's argument against the Libertarian definition of freewill. An undetermined will is a random will and if all of our acts are random then how can you be morally responsible?
You don't seem to be understanding that I don't agree that it's coherent to say that there is anything not located in particular places and times. That includes numbers and premises of arguments. The only category error there arises from a misunderstanding, or simply a lack of concern with, what numbers and premises of arguments are ontologically.
Quoting Pierre-Normand
I'm not sure what that has to do with the comment of mine it's following. "Enabling conditions" doesn't make it coherent to say that it somehow occurs non-temporally or not in a particular location.
Quoting Pierre-Normand
No, they are not. And abstractions are mental processes.
It may be inconsistent with some strong doctrine of nominalism (of spatiotemporal particulars), but is it incoherent? Where is the number five located, in your view, and when did it go there? In any case, I already granted you that, in a sense, an agent's reasons for acting have some sort of a relationship to time since they can weigh with that agent's process of practical reasoning at some time and not at other times. For all that, an agent's reason for acting (which may be the same as another agent's reason for acting on a different occasion) isn't identical to the process whereby the agent grasps it and acts on its ground. And hence, those two things fulfill different causal-explanatory purposes.
What I am relying on, in order to distinguish conceptually between your (a)-items and your (b)-items is the irreducibility of the latter to the former, and, in parallel to that, the irreducibility of (b*) rationalizing explanations of behavior to (a*) nomological-causal explanations of behavior in terms or 'psychological' laws (or neurophysiological laws). This irreducibility claim doesn't commit me to weird ontologies of abstract objects. It may even be rendered consistent with some reasonable form of nominalism, if you would insist on that.
I obviously think it's incoherent.
Numbers are located at persons' brains when the person is thinking of the number in question.
Quoting Pierre-Normand
It's not literally the same (and yes, I am a nominalist).
I have very little idea what you're saying either here:
Quoting Pierre-Normand
Or here:
Quoting Pierre-Normand
When the agent is at (b), that's a series of events in the agent's brain, during a particular range of time.
Suppose I show the written statement of a mathematical theorem to two different mathematicians, Sue and Pam. Both of them convince themselves that the mathematical proposition indeed is a theorem on the basis of a simple proof that they easily come up with. The events that occur in their brains may be very different but the occurrence of those events, in both cases, enable (or implement, if you will) the valid inference of the truth of the proposition on the basis of agreed upon axioms. We may say, then, that the fact that the proposition follows from the axioms explains why Sue and Pam hold it to be true. This fact isn't something that obtains in either Sam's or Pam's brains although it's something that they both grasp thanks to whatever occurs in their brains being in good order (that is, thanks to its being such as to enable correct mathematical reasoning in accordance with commonly endorsed mathematical-logical standards).
The problem with that is that on my view, propositions, meaning and truth only are particular events in particular persons' brains (at particular times, etc.)
Why would that be a problem? What is it a problem for? It is a problem for the ascription of free will to rational agents? How?
You were attempting to explain the earlier comments that I didn't understand, right?--with respect to (b)? But your explanation is positing stuff that I think is wrong and that really doesn't make much sense. (Even if it's a fairly common belief.)
Why it is a problem? Is there no possible explanation, on your view, why Sue and Pam are agreeing (non-accidentally) on the truth of the proposition? If the explanation of Sam's belief merely refers to contingent processes occurring in her brain, and likewise in the case of Pam, then since those processes are different, and no appeal can be made (on your view) to the shared rational principles that govern them both, it would appear that their agreement is merely coincidental.
It's not coincidental--coincidental means they're effectively "random" with respect to each other. That's not the case here. People interact, they influence reasoning, they influence expression, etc.
Yes. Hence, maybe, a hard determinist might argue that Pam and Sue are compelled to believe in the truth of the proposition owing to contingent cultural forces that they both are being passively subjected to (on the model of material nomological 'causal antecedents'). However, such an account, although popular in some circles, which stresses institution over constitution, seems to me to be blind to the existence of the autonomous abilities which rational individuals have to rationally criticize shared conventions and to convince their peers that they merit being overturned on the ground merely of the cogency of their criticisms. (And what makes it the case that a criticism of shared norms is cogent isn't something that is being nomologically determined by prior events).
I'm not saying anything about hard determinism (I buy free will--remember) or being compelled to believe something. It's not a coincidence because we're not talking about apparently "random," unconnected occurrences that have nothing to do with one another. None of that takes any of this outside of particular actions/events that have spatial and temporal locations.
So when we're talking about particular actions/events with spatial and temporal locations, (a) is either connected to (b) (and (b) (C)) in a causally deterministic way or it is not. They're all a series of actions/events with spatial and temporal locations. So it's a matter of whether ontological freedom is possible anywhere in the system or not.
It is the attempt to insert (b) in between (a) and (c) in a linear chain of nomological event-causation that I am objecting to. It rests on a category error since the (b)-items don't have the proper logical form to figure as causal relata in event-event chains of nomological causation. This insertion is an attempt to collapse (or reduce) formal causation into 'efficient' event-event causation. This amounts to an oversimplification of causal explanation. When complex systems such as living things, animals, and human beings, are functionally organized, then, many features of their behavior can be explained by appeal to their specific powers, which are emergent (and multiply realizable) formal features that they have in virtue of the way in which they are internally organized. Looking at the aggregate of 'events' that occurred in the past, prior to them intentionally acting (or behaving), loses the important distinction between (1) those features of their past that generate external constraints on their behavior and (2) those features that enable their functional capabilities to channel their circumstances and opportunities into autonomously generated (and/or rationally intended) outcomes.
But I'm not attempting to insert it in a linear chain of nomological event-causation. That latter part is up to you. I'm just saying that it's there in a linear chain of spatio-temporal events, because anything else is incoherent. Whether those events are deterministic is up to you--that's what I'm asking you.
Quoting Pierre-Normand
I haven't the faintest idea what this is saying, because I have no idea what the "proper logical form" would be in your view, or even why you'd think that we'd be talking about "logical form" per se. (Maybe it's not per se, though--I don't know.)
Quoting Pierre-Normand
You have a tendency to drift towards terms, phrases, sentences, etc. where I haven't much of an idea what you're talking about. Here, for example, what's the difference between "functionally organized" and "organized" or not even bringing this up at all, since living things necessarily have some structure (so we wouldn't have to point that out as if some things do not have it)? That loses me, because I'm not sure what you're saying.
Quoting Pierre-Normand
Explained as in "what's going on ontologically" or explained as in "here's a set of words that some people take to sufficiently satisfy their wondering about this phenomena"?
Re multiple realizability, I don't buy it in any literal sense. Again, I'm a nominalist.
I also don't buy the idea of emergence really.
Quoting Pierre-Normand
I'd need to know what in the world the above is saying to care about "losing the distinction." The first problem is again the word "functional." I just don't know what it's doing there, what it's adding to the word "capabilities."
The next problem is "channel their circumstances."
And then re "autonomously generated," in terms of a free will discussion, that term would probably only make sense if you're claiming that freedom obtains.
That is not in dispute.
Quoting Relativist
This is the very point in question. Personally, I hold no one responsible for actions in which they played no determining role. So, based on my contrary conception, factually, it is simply not inconceivable. As I recall, Clarence Darrow convinced one or more juries to acquit by convincing them that his clients were determined to act as they did. So, this claim is false.
Quoting Relativist
No one is disputing that feelings of responsibility help guide our behavior. The question is are such feelings well-founded. The argument fails to show that they are not.
Quoting Relativist
As it stands, 2 is not a compatibilist account of responsibility, but an argument for why a drunk driver should not be held responsible. One might decide to send her to jail to change their behavior, but that does not mean that she is responsible for what she did, only that we might, by this crude means, re-program her.
No one is denying the role of experience or of beliefs in the decision making process. Practically everyone knows, intellectually, that drunk driving involves grave risks. The question is not about acquired knowledge, but about how the agent weighs the incompatible factors that motivate driving drunk or not. There is no numerical trade-off between the relevant factors, so despite utilitarian objections, no algorithmic maximization can determine the decision. I think we can agree, further, that the decision is made in light of a subjective weighting process -- one that is neither algorithmic nor syllogistically conclusive.
Can't we also agree that how a person weighs such factors is not merely backward looking, not merely a matter of past experience and belief, but also forward looking -- a matter of what kind of person the agent wishes to be? And, if that is so, then the past is not fully determinative. We know, as a matter of experience, of cases of metanoia, of changes in past beliefs and life styles. While this does not disprove determination by the past, it makes it very questionable.
As for being "random," that depends on how you define the term. If you mean not predictable, not fully immanent in the prior state, free acts are random in that sense. But, if you take "random" to mean "mindless," no account of well-considered decisions can hold they are random in that sense. Personal beliefs, dispositions, and impulses all enter proairesis, but they alone cannot be determinative because they are intrinsically incommensurate. They are materials awaiting the impress of form. It is not what we consider, but the weight we give to what we are consider, that is determinative. And, we give that weight, not in view of the past alone, but in view of the kind of person we want to emerge in shaping our identity.
Quoting Relativist
I don't think the arguments given do this. They begin by noting that we feel responsible, and show how this plays a role in our behavior -- none of which is in dispute. The question of why would we have a false belief in responsibility if we are not responsible is simply not addressed. Why couldn't we reprogram the drunk driver with prison or a scarlet "D" because reprogramming works (if it does), and not because of an irrelevant responsibility narrative?
To clarify a bit, I think 'deterministic' is a predicate that is most suitably applied to systems rather than sets of events. That's because material systems governed by laws are defined with respect to sets of intrinsic properties of their constituents (such a physical predicates, like masses, positions and momenta) while other relational or systemic features of those constituents (and of the whole) are being abstracted away. Hence, looking at the behavior of a rabbit, say, construed purely as a physicochemical 'system'; its 'behavior' (viz. the set of the motions of its parts) may truthfully be said to be deterministic. But that's just because the 'events' that we are looking at are restricted to physical and chemical events, and those events indeed may be governed by deterministic laws (modulo quantum indeterminacies). But this abstract way of looking at the rabbit loses features of its biological organisation and is blind to those 'events' (viz. intentional behaviors and functional physiological processes) that physics and chemistry have nothing to say about.
Some philosophers such as Jeagwon Kim have mustered arguments, such as the causal exclusion argument, in order to infer determinism at the supervenient level of description (such as the description of the rabbit in functional physiological and/or behavioral terms) from the determinism of the system being supervened upon (the set of the rabbit's inanimate material parts). I think those arguments are flawed, but Kim at least acknowledges the need for such an argument whereas you seem to take its conclusion for granted or just believe the denial of this conclusion to be incoherent.
Minus the fact that I don't actually agree that determinism is the case, yes, I think there is no need for such an argument, because there's no good reason to believe otherwise, no good argument for an alternate position.The fact that it's easier to talk about "functional" physiological and behavioral stuff from a different conceptual and linguistic perspective certainly isn't a good argument in support of their being some sort of ontological distinction. That would amount to very naively reifying language/the way we find it easiest to think about something.
What is it, exactly, that you are claiming not to be in need of an argument? Is it the inference that if something is made up of material parts (and/or events) that are being governed by deterministic laws, then, in that case, this something necessarily also is being governed by deterministic laws?
So, you had just written:
"Some philosophers such as Jeagwon Kim have mustered arguments, such as the causal exclusion argument, in order to (1) infer determinism at the supervenient level of description (such as the description of the rabbit in functional physiological and/or behavioral terms) from (2) the determinism of the system being supervened upon (the set of the rabbit's inanimate material parts)."
I don't believe there's any need for an argument from (2) to (1), because there's no good reason to believe there's any ontological distinction to be made between (2) and (1).
The only thing I'm disagreeing with is that I'm not actually a determinist and I'm not a realist about laws--I believe there can be nondeterministic phenomena at (2) (and thus at (1)).
But I thought you were objecting to the possibility of there being nondeterministic phenomena at (2) (i.e. at the level of life, physiology and intentions) in case there wouldn't be any at (1) (i.e. at the level of the physical behavior of the inanimate material constituents). This is what a compatibilist who endorses some version of the principle of alternative possibilities would claim to be possible, and that it had seemed to me you were objecting to as being incoherent (although you didn't say why you deemed it to be incoherent, appart from asserting a lack of ontological distinction between the entities at both levels.)