I'm using logic to refer to a language about relations on an abstract level, and more specifically, it has to do with implications/inferences of relations.

Not everything in the world is a language about relations, is it? — Terrapin Station

OK, fine. I would call De Morgan's law, modus ponens, modus tollens, or the axioms of first order propositional logic "laws", but if you would rather view them as "relations... (that have) to do with implication/inference relations", that is perfectly fine with me.

I am no less puzzled by your claim that logical possibilities are a subset of metaphysical possibilities.

Again, can you think of just one metaphysical possibility that isn't a logical possibility?

I wouldn't say that there are any laws of formal logic.

There are basically language games that are set up as conventions re different species of logics, where logic, as I had said, is a subjective linguistic reporting of how individuals think about relations on the most abstract level. — Terrapin Station

You are the one who brought up the topic of logical possibility and claimed, contrary to traditional wisdom, that logical possibilities are a subset of metaphysical possibilities. I am usure how this further characterization of logic explains the claim that you made. Maybe it would help if you would give an example of something that, in your view, is a metaphysical possibility even though it is not a logical possibility. That would possibly help me make sense of your strange suggestion.

In my view that explanation is just compounding problems (note that my comments below are in the context of what's functionally going on, with respect to what's coherent or not, re conventional uses if terms):

* "Possible worlds" talk doesn't make sense except as talk about what is metaphysically possible in the actual world (which for a determinist is only one thing for each "branching point" so to speak . . . also, I see logical possibility as a subset of what's metaphysically possible, although determinists would have to say that it's the subset of things that are possible to non-contradictorily imagine; that, however, is still a subset of metaphysical possibility.) — Terrapin Station

What's logically possible is whatever isn't ruled out by the laws of formal logic alone. Whatever isn't self-contradictory is thus logically possible. Metaphysical possibilities, however you construe them, seem also to exclude propositions that are false because they are ruled out by a priori conceptual truths or by the laws of nature. (But see also Kripke on the metaphysical necessity of identity or of origin). It is very strange to say that logical possibility "is a subset of what's metaphysically possible", rather than the other way around.

* I don't know if you intended them to, but alethic or temporal possibility wouldn't refer to anything other than metaphysical possibility (they'd just be limiting the consideration to metaphysical possibility re truth-value judgments or changes that obtain relative to something)

Something is an alethic possibility ('a-possibility') if there is a true subjunctive conditional (e.g. a causal conditional) in which it figures as the consequent.

e.g. I arrived at work late because I forgot to set my alarm clock but it is a-possible that I would have arrived to work in time since if I had (counterfactually) not forgotten to set it then I would have arrived to work in time.

What is the allowed range of counterfactual antecedents that make alethic possibility statements either true or false is of course a pragmatic contextual matter that a semantic model must be sensitive to. (There seems to me to be a flaw with David Lewis's 'counterpart' models and their seemingly context insensitive similarity relations between possible worlds. Saul Kripke seems to me to provide a better informal account in Naming and Necessity.)

A temporal possibility is a possibility for the future that may become actual after the time has come. When the time has come, and the proposition now represents the past, it becomes (temporally) necessarily true or necessarily false. The range of possible worlds under consideration, in this case, is the set of possible worlds that share the same past with the actual world, at a time, and are branching out in the future consistently with the laws of nature. Just in case determinism is true, of course, then there is no branching out. There only is one actual future.

My point only was that epistemic possibilities depend on neither of those two modalities since they are premised on my ignorance regarding the actual world and don't always depend either on whatever is as of yet unsettled (for the future) or on what would have happened in counterfactual circumstances.

* "Possible worlds as semantic models" -- we can note that the semantic models that individuals happen to possess (meaning is subjective and only obtains insofar as individuals actually think it) are actualized possibilities, but this is still metaphysical possibility (and rob a deterministic actualized (or to-be-actualized possibilities) are the only possibilities there are)

No. Semantic models aren't "actualized possibilities". This is nonsense. They are sets of characterizations of possible worlds with only one among them being labelled as the actual world. When something is actual, then it is possible but only one possible world is actual.

* The idea of "subjective probabilities" is just nonsense--if there are probabilities and that's not just an illusion, that's going to be a name that picks out some objective relational feature(s) of the world

If you throw a die that you know to be fair and balanced, then, regardless on the laws of nature being deterministic or indeterministic, after the die has been thrown, but before you are informed of the result, the subjective probability for each one of the six conceivable results, from your own epistemic perspective, is p = 1/6. I am unsure why you would think this is nonsense. It is just part of the normal course of practical deliberation for people to make use of their own subjective probability estimates of the consequences of various possible actions, even when the consequences already are settled conditionally to their choices.

* It's fine to note that we can be ignorant about which possibility obtains, where we believe that prior to something obtaining, there is more than one possibility, but if we're determinists we do not believe this; we believe that there is only one thing that's a possibility prior to each "branching point,' and our ignorance is about which thing was possible. Thus (a) ignorance isn't the same thing as some sense of possibility, and (b) this is not a different sense of possibility than metaphysical possibility; we're merely talking about our ignorance and beliefs re metaphysical possibility.

Those are just dogmatic assertions that you are making. Determinists and indeterminists alike usually agree about the fixity of the past. Yet, it makes sense to speak about epistemic possibilities regarding past events.

* Mathematics (and logic) are simply languages that report our subjective understanding of contingent relations, as they are thought about on the most abstract level, and

* Truth-value is a judgement about the relation of a proposition to something else (the exact something else being whatever the individual believes to be the pertinent relational consideration (for the context at hand). That could be their perception of the external world, or consistency with their stock of previously adjudged propositions, or usefulness per their judgment, etc.)

Well, yes, sure. That's exactly what epistemic modalities are about.

Again, it's not at all plausible to me that beliefs about possibility are beliefs about something other than whether an event can metaphysically obtain.

At least where one is using the term "possibility" in anything like its conventional sense. — Terrapin Station

The words "possible", "possibly" and "possibility" have a multiplicity of senses and not just one single conventional sense. Just look up the two or three main definitions in any common English dictionary, as well as the various examples of common usage that they give.

It is not at all plausible that epistemic possibilities really are beliefs about the metaphysical (or alethic, or temporal) possibility of events. One good way to get at the difference is to pay attention to the fact that most modalities can be construed as statements regarding some set of possible worlds (semantic models). But epistemic possibilities always are statements about the *actual* world. (See the examples in the Wikipedia article on epistemic possibility and the mention of the actual world at the very beginning.)

If we are waiting for the bus, and we are unsure if we may *possibly* have missed it, and ought to walk rather than wait any longer, we are not pondering over whether there might be possible worlds other than the actual world where the bus has passed ahead of schedule. We rather are pondering over whether it is at all likely (where "likely" expresses subjective probabilities) that it has passed ahead of schedule, and we thereby missed it, in the actual world. This "possibility" is entirely premised on our state of ignorance regarding some features of the actual world.

Another telling example might be this. If I endeavor and seem to have succeeded in proving a mathematical conjecture then I may still wonder if the conclusion might not be false because I have made a logical mistake while constructing the proof. So, I think it is still possible that the conclusion may be false. But the conclusion being a mathematical proposition, if it is true in this world then it is true in all possible worlds. (In other words, the truth of mathematical proposition isn't a contingent truth). Therefore my belief that, for all I know, the conclusion might still be false (because I am still unsure about the validity of my proof) doesn't entail the claim that there might be possible worlds other than the actual world where the statement is false. The possibility of its being false is purely epistemic.

What would be the grounds for a claim that a proposition like "There is a possibility that x" has a structure like "(NP(A)" (or EP(A) or EP(NP)A) or whatever)? — Terrapin Station

I am simply trying to understand *your* suggestion that it may be incoherent to interpret epistemic possibilities to be "about" anything else than nomological possibilities. I provided a simple counterexample regarding my girlfriend possibly being at home (for all I know). It would help if you would explain what you meant with your "about" claim if it's not related to content.

I'll be away from my computer over the next several hours.

It's not as if all beliefs are about possibility. — Terrapin Station

Of course not. But you were the one expressing doubt that epistemic possibilities could coherently be thought to be "about" anything else than nomological possibilities, not me. I had construed that as the claim that epistemic possibilities have the form: 'for all I know, A might be nomologically possible', or, in shorter form, 'EP(NP(A)'. If you meant something else with your suggestion that all epistemic possibilities are "about" nomological possibilities, I don't know what that is. You would need to explain what you mean by "about" if it's not propositional contents.

What in the world does that have to do with what I asked?

I didn't say anything about the only sorts of propositions that might be true. In fact, I didn't say anything about propositions or truth whatsoever.

I didn't say anything about "complex" versus "basic" propositions. — Terrapin Station

What you asked me was this: "Actually this is a better question: why wouldn't epistemic possibility be beliefs about nomological possibility? I'm not at all convinced that it's coherent to say that it's anything other than that."

Are you not familiar with the idea that the contents of beliefs are propositions?

It really sounds like you were doubting that epistemic possibilities can coherently be said to be about anything other than propositions of nomological possibility. That would make those contents complex propositions of form (NP(A), where A is a simple proposition in subject/predicate form, and the belief itself (the epistemic possibility) has the form EP(NP(A). I have no idea why you would believe that, or believe that beliefs of the form EP(A) that aren't themselves "about" nomological possibilities, are necessarily incoherent.

To put it quite simply, when I say that, for all I know, my girlfriend may still be at home, this statement of epistemic possibility is about my girlfriend being at home. It's not about any sort of nomological possibility. I may not even believe that there are laws of nature. (Maybe I am a Humean regularist about laws, and I don't believe in natural necessities at all, say).

Actually this is a better question: why wouldn't epistemic possibility be beliefs about nomological possibility? I'm not at all convinced that it's coherent to say that it's anything other than that. — Terrapin Station

I'm unsure why anyone would think that the only sort of propositions which, for all one knows, might be true, are complex propositions regarding the nomological possibilities of basic propositions. There are more things under the Sun than just nomological possibilities. Maybe a determinist would believe something like that (that everything that is true is nomologically necessary) since, on her view, everything that will ever become temporally necessary in the future always has been in the past also. But then, that would only be true if we assume the nomological necessity of the 'initial state' of the universe.

Wait, in the situation I'm presenting, what are you claiming the person believes is nomologically possible? — Terrapin Station

There isn't any proposition that she is claiming to be nomologically possible, if I understand you. Rather, she is claiming that one and only one among two propositions, A and B, is nomologically possible but she doesn't know which one it is. (If she would care to draw logical inferences from her own beliefs, though, she could conclude to the nomological possibility of the disjunctive proposition 'A xor B', as I've suggested to Michael, since this is a valid inference from (nomological) modal logic.)

Otherwise, how exactly does that amount to mixing anything up, because all I did was talk about what they do and do not believe? — Terrapin Station

That's because in the case you are describing, as I've displayed through formalizing it, there are two different sorts of modalities involved. It is a case where you avow ignorance regarding which one among two propositions is nomologically (or temporally, maybe) impossible, while the other one is nomologically possible, although you don't (yet) know which one. Since epistemic possibility doesn't entail nomological possibility, there is no validity in inferring the conclusion that you believe both to be nomologically possible. It's just your conflating the those two sorts of 'possibility' that generated the invalid inference and the false conclusion that you thereby believe both to be possible conjointly.

I had said simply said this, which you never addressed — Terrapin Station

Yes, I did address it. You didn't reply to my comment about it. Maybe you missed it.

So when you don't want to deal with some particular individual's beliefs, you just claim that that is not a rational person. — Terrapin Station

Everyone has inconsistent beliefs but there ought to be a consistent core. Who said I wan't prepared to deal with it on a personal level? We were discussing rules of inference that are valid for epistemic logic, presently. For those rules to apply to the belief systems, or epistemic perspectives, of real persons, there must be a minimal presumption of rationality. When a person believes that P, and also that Q isn't logically consistent with P, she isn't normally prepared to accept that, for all she knows, Q. If she is nevertheless prepared to accept Q as an epistemic possibility, then she must thereby either acknowledge that she doesn't really believe that P for sure, or be prepared to revise her belief that P and Q are inconsistent.

If a person shows no tendency to revise some of her beliefs when they are shown to be mutually inconsistent, then there is no saying what it is that, for all she knows, might be true, and epistemic modal logic breaks down as a means for interpreting her. (Look up Donald Davidson on 'radical interpretation', the 'constitutive ideal or rationality' and also the 'principle of charity'.)

Looks like you had this page open for a while before replying. I edited my post a few minutes ago. — Michael

No trouble. I must however retract what I just said. I hadn't actually said "both impossible" originally, but rather "not the case that they're both possible". It is the latter that is the correct paraphrase from the right hand side of my revision of your formula, and it is the proposition that I take to follow from the left hand side thought the correct application of De Morgan's law.

Your own "IMPOSSIBLE(A ∧ B)" also follows indirectly, but it follows from modal logic rather than propositional logic.

From ¬(◇A ∧ ◇B) you can indeed infer ¬◇(A ∧ B) since if

(1) it is not the case that ((there is a possible world where A) and (there is a possible world where B)),

then

(2) there is no possible world where both A and B.

I think "they're both possible" is ambiguous here. Do you mean "A is possible and B is possible" or "either A or B is possible"? I'm saying the second, not the first. — Michael

I was paraphrasing you formula and interpreting '¬◇' as "impossible". I meant "both impossible" also as a paraphrase for "¬◇A ∨ ¬◇B".

¬◇A ∨ ¬◇B. Which, again using De Morgan's theorem, just entails ¬◇(A ∧ B), — Michael

It rather seems to me that applying De Morgan's law to ¬◇A ∨ ¬◇B yields ¬(◇A ∧ ◇B).
If either A or B are impossible, then it's not the case that they're both possible.

But epistemic possibility necessarily has to do with the individual's beliefs re possibility, no? — Terrapin Station

What epistemic possibility "has to do with" is rational people's consistent sets of beliefs, and what they are about is specific propositions in relation to those prior sets of beliefs. The content of those propositions can concern ordinary empirical claims, claims of temporal modalities, alethic modalities, mataphysical modalities, or even other epistemic modalities (or anything else that one can intelligibly believe or disbelieve).

For instance, I may claim that, for all I know, for all Donald Trump knows, he will be remembered at the greatest president ever. This could be formalized thus: EPPN(EPDT(P)), where EPPN is the epistemic possibility operator relativized to my own present epistemic perspective, EPDT is the same operator relativized to Trump's epistemic perspective, and P is the proposition that he will be remembered at the greatest president ever.

Well, the epistemic sense is about what I believe. So if I believe that it's not in fact possible for A or B to have obtained--I believe that it's only possible for one of them to have obtained and the other was always impossible (even before one obtained), then how does my saying "I don't know" imply a belief that both are possible? I rather explcitly believe that one was never possible; I simply don't know which one was possible. — Terrapin Station

You are freely mixing up metaphysical and epistemic possibility operators in this paragraph, so the question seems ill posed. I already mentioned that epistemic possibilities don't entail other sorts of possibilities. If you believe that only one among two 'possibilities' can be actual, due to the conjunction of them being nomologically impossible, then this could be formalized thus: EP(◇A xor ◇B), where EP is the epistemic possibility operator, '◇' is a different sort of possibility operator.

Getting to the won't answer a simple question point. Nice. — Terrapin Station

But I did answer it, didn't I? You're just complaining that it took more than 10 minutes, due to my being busy answering other posts of yours!

Don't you ever have time available where one wouldn't have to wait ten minutes for a response in a conversation? — Terrapin Station

You can't possibly be serous. Just listen to yourself. I am patiently replying to your rapid fire quibbles over a simple notion (epistemic possibility) that you could have gotten acquainted with easily on your own with a Google search. We didn't come to a prior agreement that this exchange, which has spanned a few days, ought suddenly to become be a no-break-allowed conversation.

Maybe post when you're not so busy, so that it doesn't take so long to get a simple yes or no answer? — Terrapin Station

Really? You can't wait ten minutes? Should I take an appointment with your personal secretary, next time? I am allowed more than one bathroom breaks in a day?

If I believe that I don't know if A or B is the case, does it follow that I believe that it's possible that either A or B are the case? — Terrapin Station

Yes, if you are using 'possible' in the epistemic sense; no, if you are using it differently. If I don't know whether or no my girlfriend is home, that doesn't entail that I believe that she is some sort of a Schroedinger Cat living in an unsettled state of both possibly being or not being at home. Mere epistemic possibilities don't logically entail other sorts of possibilities.

An epistemic sense has to do with individuals' beliefs, right? — Terrapin Station

Of course! I said that from the very beginning. It is relative to the body of beliefs of a person at some point in time.

What??? You didn't seem to understand if we're only. If we're only saying that "I don't know," then we're not saying "that's possible," right? Because "that's possible" is not ONLY "I don't know." — Terrapin Station

If the person isn't badly confused, and she genuinely doesn't know, this tends to imply that she doesn't know or believe something on the basis of which the proposition or its negation can be deduced. (Although this raises issues regarding the logical closure of knowledge, not very relevant here). That means that for all practical purposes, claims of ignorance are equivalent with claims of epistemic possibility. They have logical implications governed by the logic of epistemic modalities.

If I'm only saying I don't know, it doesn't follow that I believe it's possible. — Terrapin Station

If it means that both the proposition and its negation are consistent with everything that you know (or believe to be true), then that's 'possible' in the epistemic sense. Nobody ever suggested that epistemic possibility implies other sorts of possibility. But when people make use of the word "possible" in ordinary contexts, what they mean satisfies the definition of epistemic possibility, and what they say is often true.

This comment suggests that you didn't really read or understand the comment you're responding to. — Terrapin Station

I was just trying to be helpful in providing links to the commonly discussed issues (in the free will literature) that I had mentioned and that you claimed not to be familiar with. There is no obligation for anyone to make use of them. They're just there for the taking in case anyone is interested.

If we're ONLY saying "I don't know if she's home or not" how are we talking about possibility? — Terrapin Station

Because it an extremely common and everyday use of the words "possible" and "impossible" (Did you know that P? I don't know, that's possible, for all I know.) And also because those uses of the words possible and impossible (and necessarily) in an epistemic context obey the very same rules of modal logic as alethic or metaphysical modalities. Which is why, of course, it is natural to use those words in epistemic contexts, since they obey the same logic and hence licence the same forms of inference.

If you believe that I'm saying that the definition has it as an illusion you're misunderstanding my comment. I'm saying that what the definition is describing would be an illusion, if determinism is true. — Terrapin Station

You still seem to be missing the point of the concept of epistemic possibility. If when I am claiming that for all I know my girlfriend may (epistemic 'possibly') still be at home, I am saying nothing more than that I dont know. I don't know because both the proposition and its negation are consistent with everything that I do know or believe falsely. In what sense would this seemingly justified claim of ignorance be illusory, and what is the relevance of determinism to it? Maybe I know that she is home and have temporarily forgotten due to some distraction? So I can be brought to remember that knew it already? That would by one way to interpret your claim that a statement of epistemic possibility can be mistaken. But determinism has nothing to do with it.

Re the other comments, I'm not that familiar with Van Inwagen's argument (and not that fond of the fact that it seems to be made strictly as a formal logical argument, due to my belief about what logic is, etc.). I'd have to go more into detail just what the issue would be there. Re the "luck" and "intelligibility" issues, for one, will seems to bias probabilities. There's no reason to believe that different possibilities are equiprobable. And the bias can be near 100% in some cases. — Terrapin Station

Getting acquainted at least with an informal statement of van Inwagnen's consequence argument (also credited to Carl Ginet) is useful because it has been central to the debate about free will and determinism for many decades now, and it brings into focus many of the incompatible commitments that ground the accounts of the libertarians, the compatibilists and the hard determinists.

The "intelligibility problem" for for libertarian free will is a very old objection that has been raised for it and that has been much discussed by one of the most prominent contemporary libertarian: Robert Kane. See page 23 in Four Views on Free Will, which you can preview for free on Google Books.

The "luck objection" is closely related to the intelligibility problem but it is most often raised in the context of the libertarian accounts of the "possibility to do otherwise", i.e. the libertarian way to simultaneously satisfy the principle of alternative possibilities (PAP) and secure "agent control" over the choice actually being made. Maybe you can just Google "agent", "control", "luck" and "PAP". It is also mentioned in the SEP article mentioned above (search the word "luck" in the page).

Yes it is on my view. The definition of epistemic possibility that you give describes an illusion--something that one is mistaken about re how the world really is (or a false belief). Namely that it's possible that C can obtain consequent to A, in the scenario where there is only one possibility that can obtain consequent to A, B. — Terrapin Station

You are misunderstanding the definition. It's not an illusion, it's a claim of ignorance. Saying that P is an epistemic possibility (always relative to the consistent body of beliefs of an individual or some community consensus) just means that it isn't inconsistent with this prior body of beliefs. If I am coming back home and I don't know if my girlfriend is home already, that means that either (1) her being home already or (2) her not being home already are propositions that both are consistent with everything that I already know or believe (truly or falsely). This is what it means that both propositions are epistemic possibilities from my perspective.

??? That's not my view. There is no "epistemic possibility" where there is no ontic possibility. An "epistemic possibility" with no ontic possibility is otherwise known as a false belief. — Terrapin Station

I was using the term "epistemic possibility" in the way it is commonly used in philosophy. From Wikipedia: "In philosophy and modal logic, epistemic possibility relates a statement under consideration to the current state of our knowledge about the actual world: a statement is said to be:
epistemically possible if it may be true, for all we know..." (my emphasis)

So, a claim of epistemic possibility regarding a proposition that is (unbeknownst to one) metaphysically, historically or nomologically impossible isn't the expression of a false belief. It is more akin to an avowal of ignorance.

My own account of free will doesn't rely at all on epistemic modalities. They are rather irrelevant to it. The reason why the issue came up is because you (and also John) were charging me with conflating epistemic and ontological issues when I was attempting to distinguish two different sorts of possibilities for the future, neither one of which is epistemic on my view.

However, it seems to me, the tendency to conflate those two sorts of 'necessities' (the duals of the corresponding 'possibilities') often leads philosophers into a dilemma. Compatibilists who deny the principle of alternative possibilities are embracing one horn of this dilemma (and face problems in dealing with van Inwagen's consequence argument), whereas many libertarians (including traditional agent-causal theorists) are embracing the other horn (and face difficulties in dealing with the luck and intelligibility objections). My suggestion is that when the proper distinction is made between the two sorts of modalities that are relevant to the metaphysics of agency (neither one of which is merely epistemic) then there appears a third path between the two horns of the traditional compatibilist/libertarian dilemma.

Thanks for expressing your view regarding my 'room-escape' thought experiment. I'll comment further on your response later on.

Paper : http://www.academia.edu/4309985/Epicurus_refutation_of_determinism — FreeEmotion

Thanks for that. It would be interesting to compare this exegesis/reconstruction of Epicurus's account to Chisholm's, Clarke's or Kane's more recent libertarian accounts of free will and see if it suffers from the same limitations (such as the problems from luck and intelligibility).

What is the importance of the free will debate, in your eyes? In mine it is one of responsibility- the metaphysics only interest me insofar as they inform the notion of responsibility. If responsibility is the main focal point, then one can be a compatibilist even if determinism is false because the free will the compatibilist is concerned with is one of responsibility, not metaphysics. — Chany

I agree with you that the topic of responsibility is centrally important to the 'free will and determinism' debate.

Some philosophers, such as Helen Steward, defend the thesis of agency incompatibilism, according to which determinism isn't compatible with animal agency in general, and not just incompatible with free rational agency. (See her A Metaphysics for Freedom). This may be true, and if it is true it would entail that free will isn't compatible with determinism since free will characterizes just one particular sort of animal agency.

But while we recognize non-rational animals such as cows, dogs and finches to be agents, we don't ascribe then free will. They are, in a sense, irresponsible, since they aren't able to take ownership of their own natural and/or conditioned tendencies. So, even if it is the case that the bare metaphysical inquiry into possibilities for the future, as it applies to the natural powers of substances and of non-rational animals, is undoubtedly relevant to the free will issue, it remains the case than an inquiry into the topic of specifically rational agency (which includes questions about the philosophy of 'moral psychology') remains centrally relevant.

Our understanding is irrelevant to whether possibilities or free will obtain.

And our understanding of what we are intentionally doing can be part of the cause of what we are doing where either (a) only one possibility exists in any given situation and there is no free will, or (b) at least two possibilities exist in at least some situations and there is free will. — Terrapin Station

On your view, in a world where microphysical determinism obtains in such a way at to enable an ideal Laplacean predictor to foresee all future events (on the basis of his knowledge of all the present physical conditions and of the deterministic laws of physics), there are two senses in which a future occurrence can meaningfully be said to be 'possible' -- epistemic possibility or ontological possibility -- where the latter doesn't depend on an agent's perspective at all.

I think this leads you to paper over an important metaphysical distinction regarding possibilities for the future (though you hardly are alone in doing this). Imagine the following two cases.

Case 1: Suppose an agent is placed in a room with only two doors leading out. The agent is informed that the room will be flooded in ten minutes. At that time, both doors will lock down automatically. The agent must thus exist the room within ten minutes in order not to drown. Furthermore, the agent is informed that there is a wild tiger behind the first door and that there are a dozen venomous snakes laying on the ground behind the second door. The agent is, as of yet, indecisive about what to do, except that she is quite certain that she must exit soon. Until such a time when she will have made up her mind what kind of beast(s) might be more manageable or least dangerous, she doesn't know what door it is that she will open. (But the Laplacean predictor knows what her eventual choice will be already).

Case 2: Consider now a variation on the previous scenario where one (and only one) of the two doors is locked from the very beginning, and the agent is informed of this also, though she isn't told which one it is. However, if the agent would first try to open the door that happens to already be locked, there will be no penalty. She would then be allowed to walk to the other door and open it (provided the ten minutes time limit hasn't elapsed already). Now, the agent still doesn't initially know what door it is that she will eventually manage to open and what sort of beast(s) it is that she will confront. But there is no point in her losing time pondering over the issue since in any case only one of the two doors is unlocked. She might as well try out one door (chosen at random) and thereby find out as fast as possible which one is unlocked, and then confront the tiger, or snakes, on the other side, whatever the case may be.

On my view, in the first case, before the ten minutes have elapsed, and before the agent has made up her mind which door it is that she will try to open first, *both* options are open to her (and hence represent possible future outcomes) from her own practical perspective, and this not merely a matter of epistemic limitation, or so I would argue.

In the second case, only one option is open, and it is indeed merely a matter of epistemic ignorance which one is open from the practical perspective of the agent. The agent only has one exit option but doesn't yet know which one it is.

On your view, though, there is only one 'ontological possibility' in the Case 1 scenario. And the fact that the agent doesn't yet know (prior to making up her mind) which door it is that she will open is a case of epistemic possibility, just as it is in the Case 2 scenario.

My question to you, then, is this: Why is it that the agent, in the first scenario, wouldn't be justified to just try one door at random and forego any prior deliberation regarding the potential threats (tiger versus snakes), just as it would make sense to forego such pointless deliberations in the the first scenario when she knows that only one door is unlocked anyway. Why is there any practical point in her prior deliberating what choice to make when, on your view, there actually just in one real (ontological) possibility that already has been set by the past state of the universe and the laws of physics; and her 'feeling' that there are two options really (ontologically) open to her reflects nothing more than mere 'epistemic possibilities'?

It's not the case that it's not nonsense ontologically just because a lot of people are talking about it. — Terrapin Station

No, of course not. But you were suggesting that the idea of top-down causation doesn't have any "correlate" in the real world. So, I thought you rather seemed unfamiliar with the very idea of top-down causation, which refers to an abundance of phenomena in both the natural and the social sciences. But, in any case, I offered an argument why the idea isn't nonsensical at all, can be defined with a scientifically sound criterion for the existence of causal relations. You've ignored my explanations.

It's not that I'm stumping for "low level causality" as you seem to believe. The whole "level" idea is nonsense. It's simply an artifact of how people are choosing to think about this stuff.

This is like saying that the very idea of "rabbits" is nonsense. "Rabbits" just is an artifact of how biologists are choosing to think about lumps of biological stuff.

What makes it an epistemological issue is that you're talking about explanations, understanding, what our physical laws can do, etc. — Terrapin Station

That's because in the case of rational agency, our understanding of what we are intentionally doing is normally part of the cause of our doing it. Just because epistemically relevant concepts are mentioned in the context of the discussion of human cognitive powers hardly means that no real causation can be involved. As I mentioned in an earlier post, there are two radically different manners to know that something will happen. The first one is to learn about a sufficient causal antecedent. The second one is to choose to make it happen (when it is within our powers to do make it happen). The second one is a reliable means for us to know many things that will happen that doesn't depends on us our knowing, or ignoring, any other causal antecedents of those happenings.

Aren't you the user Streetlight under a different name by the way?

Certainly not. I only have one single account and my handle is my real name.

"Top down" and "bottom up" are nonsensical ontologically. They might make sense re how some people think about causal relations, but they'd have no correlate in the external world. — Terrapin Station

They are multitudes of correlates in the real world. You should keep up with recent literature on the philosophy of science, the philosophy of biology, the philosophy of chemistry and the philosophy of physics. (Search "top-down causation on PhilPapers). Nobody seems to deny that externally triggered changes in high-level features of organized systems seem to reliably produce systematic effect on low-level features (and this matches up with traditional interventionist accounts of causation: if manipulation of A produces systematic change is B, then there is a causal relationship between A and B). So, the only issue that separate strong emergentists from skeptics on that issue is the question of the potential reducibility (or 'eliminative' analysis) of those prima facie real cases of inter-level causal efficacy into low-level constituent causal processes. The chances for that seem bleak and only to be motivated by die hard reductionist or physicalist tendencies.

Which means that you're not a compatibilist. — Terrapin Station

Indeed, I am not. But I do share with compatibilists the commitment to the idea that free will and responsibility are compatible with the causal closure of the physical domain. As for the question of the determinism of fundamental physical laws, I also share with compatibilists the belief that it is irrelevant to the question of free will. So, my position is intermediate between traditional compatibilism and traditional incompatibilist libertarianism.

What you're addressing in this passage is an epistemic issue. That's why you're conflating an epistemic and an ontological issue. Whether possibilities and free will obtain has nothing to do with our ability to explain anything, our understanding of physical laws, what physical laws can say about anything, etc. — Terrapin Station

Causal explanations for sure address our explanatory needs, but to single then out for that reason doesn't turn them into merely epistemological issues. There is a tight conceptual connection between causation and explanation (as there is, also, between causation and natural laws and/or natural powers, and between causation and counterfactual dependence). So, when I'm suggesting that, the fact that some bodily motions happen to realize, in some specific practical contexts, some specific sorts on intentional actions, is explained by the intelligible source of the intention of the agent (i.e. her reasons for acting in that way precisely in contexts of that sort), I am not merely pointing to the fact that she may believe, truly or wrongly, that this fact is causally relevant to the production of her action. I am suggesting that it is explanatory by virtue of its being causally relevant.

The alternatives to the reality of those real instances of rational-causation are either causal-overdetermination or something like pre-stablished harmony (à la Mallebranche or Leibniz). But the only alternative explanation of the fact that those bodily motions come to constitute the specific form of intentional action that they do constitute, the allegedly determinative neurophysiological (or microphysical) causal entecedent of the bodily motions, doesn't actually explain the fact at all. It is a very poor candidate cause. On the other hand, the rational-causal explanation makes perfect sense in light of the fact that human brains have have evolved over phylogenetic time frames, and then continue to adapt through learning over ontogenetic time frames, precisely to sustain the practical rational powers of mature human beings. But it is enough to account for the dependence of rational abilities on underlying neurophysiological structure to cast the latter as enabling-causes of cognitive function, and not as determinative causes of our specific exercises of practical and theoretical rational powers.

OK, so it seems to me you are saying that microphysical processes, including I would assume, cellular processes, do not exhaustively determine (at least some) macro phenomena, including human decision and thought. — John

Sure, I am saying that but this is something traditional incompatibilists proponents of agent-causation have been saying for a long time. But what they have been saying also always had seemed incredulous to the vast majority of philosophers because traditional agent-causalists (such as Roderick Chisholm and Randolph Clarke) thought that since free agents control their own bodies, then the only way for agent-causation to be irreducible to neural-causation is for agent to be able to control their own neural processes, and thus, in some mysterious way, initiate new causal chains in their own bodies at the neurophysiological level that are not themselves initiated by earlier neurophysiological events. Hence, those intentional "acts of volition" are "contra-causal" in the sense that they are "uncaused causes" of neurophysiological events.

This is indeed incredible and it is not at all what I am saying.

If this is correct then we have no argument and there is also no need for you to hold a compatibilist position it would seem, because you are rejecting determinism, at least as I understand it.

Indeed, I am rejecting determinism. But unlike traditional libertarians, and unlike traditional agnent-causalists, I am not denying causal determinism and causal closure at the level of neurophysiology and "raw" (physicalistically described) bodily motions. Rather, I am adducing results in the philosophy of biology and the philosophy of emergence to explain how the specific internal causal structure of 'higher-level' intentional actions, and their integration with our mental lives, make them dependent on their neurophysiological underpinnings in a way such that those underlying processes are merely enabling of our rational powers of agency and not determinative of our responsible decisions.

I still doubt that it is possible to offer any coherent explication of the 'relationship' between so-called "bottom up' and 'top-down' causalities, though.

I've explained how I construe top-down causation (which is somewhat different from the standard interpretations in the literature on emergence) in a previous post. Bottom-up causation, on the other hand, mainly is a manner of the functional organization of our brains being such as to enable our rational powers of theoretical and practical reasoning. How and what those powers are actualized, is another matter entirely, and it isn't determined from the bottom-up. Processes of top-down and bottom-up causation, though, both are irreducible to deterministic causal explanations that only look down at the causally closed level of neurophysiology and 'raw' bodily motions.

One apparent difficulty with my approach concerns the reconciliation of the possibility of top-down causation with the causal closure of the lower level. But this only seems impossible when we assume the token-identity, and hence a one-to-one mappings, of mental-events with neurophysiological events. But I am denying the existence of any such psycho-physical parallelism. Brain events aren't mental events at all. Brain structures and processes enable basic cognitive functions, but our mental lives are a matter of the dynamical actualization of those functions in a rich context of social embedding and scaffolding of our behaviors. This social context consists in a formative preexisting linguistic community and a meaningful environment rich in rational-behavioral affordances.

So do you think you could manage an executive summary of the argument? I spotted bits of it here & there in your posts, but I couldn't possibly reconstruct it.

I think it might help make your position clearer. A change of scenery. Anti-reductionism is more or less a lemma for you, so maybe if you just presented the lemma separately, that would be tidier. (I was going to say something about people not having entrenched views about meteorology, but ...) — Srap Tasmaner

I think anti-reductionism may be more of a default position for me rather than a lemma. It is actually a position that is shared by many of the philosophers that I am arguing against. When it was the main topic of the thread, I was happy to straddled myself with the burden of defending anti-reductionism. I did it in this thread, starting on the third page, in a protracted discussion with Frederick KOH. (I may have made some comments, along the way, about the irreducibility of rational-causation and/or mental-causation, and how they operate quite naturally and non-mysteriously).

However, the relationship between reductionism and the principle of alternative possibilities (PAP) is extremely complex. My main beef with the manner in which most compatibilists, libertarians and hard determinists alike seem to dismiss the compatibility between PAP and micro-physical determinism is their main motivation for dismissing the possibility of this compatibility, and this seems to be their inchoate acceptance of actualism. (The 'new dispositionalists' who I mentioned mostly are compatibilists, but they explicitly argue against actualism: the thesis that only what is actual is possible, even in a deterministic universe). The rejection of actualism requires a radical shift from a Humean metaphysics of event-causation to a metaphysics of substances and powers. Since substances have causal powers that are irreducible to the causal powers of their material constituents, that is one step in the argument where reductionism is at issue.

What complicates matters is that, as already mentioned, most contemporary compatibilists reject PAP -- i.e. maintain that free will and responsibility don't require genuine alternative possibilities for the agent to chose from -- and yet they do so (rejecting PAP) on the basis of their uncritical acceptance of actualism in spite of the fact that they aren't committed to nomological reductionism at all.

Indeed, Jaegwon Kim's argument for causal exclusion only explicitly relies on the thesis of the supervenience of the domain of mental events over the domain of physical events. Yet, Kim is an self-avowed non-reductive physicalist who even endorses a weak form of emergence. The reason why, though, he is led to a belief in the causal exclusion of the mental, on the basis of the alleged causal sufficiency of the the causal efficacy of the (broadly deterministic) physical supervenience base, is the lack of attention that he pays to the irreducible principles of individuation of mental phenomena (something multiple-realizability is relevant to) and also to the fact that the causal efficacy of the mental isn't at all a matter of event-event causation. I've also already commented on Davidson's endorsement of the principle of the nomological character of causation. Davidson't also believes mental events to by anomalous (qua belonging to mental types) in spite of them being token identical with physical events. I don't think this is perspicuous at all, but it's not mainly reductionism that is the source of the mischief, since Davidson, just like Kim, isn't a reductionist. Rather, the problem again is the reliance on a narrow Humean conception of event-event nomological causation.

So, the core of my argument consists in showing how an overly narrow conception of event-event nomological causation leads to actualism and why the replacement of this metaphysics with a more Aristotelian metaphysics of substances and their natural powers (and of peoples' 'second-natural' powers of rational deliberation) enables the principle of alternative possibilities to be endorsed consistently with determinism and causal closure both having complete reign at the level of their atomic material constituents. Much of this argument, though, consists in reminders about commonly known facts of ordinary life (and of ordinary scientific practice) and the debunking of commonly endorsed metaphysical prejudices mistakenly believed to be obligatory components of the modern scientific picture of the world. One of the most important reminders is that we don't truthfully withhold ascriptions of powers to ordinary objects, even just temporarily, on the mere ground that the conditions of exercise of those powers aren't currently realized. But I'll have to say more about that because the temptation for believing so in the case of human agency (especially in the context of the free will and determinism debate) seems irresistible to many.