Determinism and Choosing Otherwise

Steve has responded in our ongoing discussion on determinism and choice.

I had said: “If determinism is true, a person can't choose otherwise.”

Steve responded: As usual, Dan is equivocating. A predestined agent can contemplate different hypothetical courses of action. And the hypothetical he chooses to act upon always turns out to be the hypothetical that God decreed to be. Indeed, God decreed the agent to choose that hypothetical option. A predestined agent doesn’t know in advance which hypothetical is a live possibility. But the apparent alternatives influence his choice of the viable alternative. So they serve a purpose. Although they are merely apparent, they are still functional in the deliberative process. Psychologically useful.

There’s nothing unusual about this. Take a card game. Given the cards that are on the table, face up, along with the cards remaining in the deck, a gambler will decide to bet or to fold based on the possible and probable combinations which remain outstanding. At a metaphysical level, only one of these ostensible possibilities is a live possibility. For the cards in the deck are (randomly) arranged in just one sequence at a time. But the gambler doesn’t know which combination is the actual combination. At an epistemic level, several combinations are still possible. Are still in play.

That calculation affects his choice. Even though the possible hands which he contemplates are mostly impossible hands (given the actual, albeit unknown, order of the deck), he is still making a choice based on the apparent alternatives which are available to him.

I don't believe I am equivocating; given determinism, a person can’t choose otherwise and if one is a determinist, they can’t consistently think they can choose otherwise. I don't think the epistemic sense of ‘possible’ reconciles determinism with 'the ability to choose otherwise'.

The card player example relates to the outcome of choices not choices (I said choose otherwise, not do otherwise). The epistemic sense of 'possible' relates to the execution of choices, not choices themselves. In fact, the card player example is twice removed from the choice itself. The card player chooses to take another card; his success or failure in that attempt is getting another card or not (i.e. does he have a heart attack while asking for one, or does a ceiling tile fall on the dealer's head...). Steve's example is about the outcome of the draw, not the draw itself. Further, the card player isn't thinking "is choosing to draw possible?" (i.e. can I make the internal mental resolution?), he's thinking about the outcome of drawing another card, so his use of possible relates to outcomes of choices, not choices themselves.

But even if I had said "do otherwise", a determinist card player still shouldn't think he is able to do otherwise. In an epistemic sense, the player might think 'blackjack is possible' and he might think 'bust is possible'. But he couldn’t reason 'since either blackjack or bust is “otherwise”, therefore "otherwise" is possible'. The sense in which blackjack is possible and bust is possible excludes any relationship with the category "otherwise"; so since "otherwise" isn't in the two major premises, it shouldn't be imported into the conclusion. The category ‘otherwise’ is impossible and the determinist poker player knows it’s impossible (even if he doesn’t know which instance falls within the category).¹ Worse, the fallacy creates the direct contradiction of thinking that doing otherwise is both possible and impossible (even in the epistemic sense of possible - i.e. given what the player knows).

Beyond 'the ability to choose otherwise'; I don't think the epistemic sense of possible could reconcile the determinism with the dictionary definition of 'choose' (i.e. selection between possible alternatives). Someone could use the epistemic sense of possible and still be ruling out determinism. The epistemic sense of possible in conjunction with determinism works on one possibility at a time. Granted, a determinist could talk about one possibility in an epistemic sense, but when he starts talking about two or more possibilities, he undermines determinism. The blackjack player could say "blackjack (21) is possible" and he could say "20 is possible". But if he is a determinist, he can't combine them, because he doesn't believe in twofold possibilities.

This "one possibility at a time" limitation of determinism becomes a problem when it comes to choice, as choosing involves at least two options. A determinist can't say (or think or imply) 'I know 20 is possible', if 20 is possible or 21 is possible and he doesn't know which. That would be a division fallacy like saying "I know 'Dan drank coffee this morning' is true", because he knows 'Dan drank coffee this morning' or 'Dan did not drink coffee this morning' is true. He erects one pillar and the other falls, and if he tries erecting both at the same time, they both fall. While still in an epistemic sense, given determinism, he could also say "I know 20 or 21 is impossible", so 20 and 21 are possible and impossible. He can't say 20 and 21 are possible, because that would be like saying 41 is possible. He can't say "20 or 21 are possibilities" or "they are possible alternatives" or "they are alternative possibilities"; these expressions (even understanding 'possible' in an epistemic sense), undermine determinism.

Further, the epistemic sense of possible isn’t a valid candidate for defining choose. But before we get into the details, I need to take a brief detour into the difference between causal and logical impossibility.

Impossible is impossible; no matter why it’s impossible. But semantically, sometimes we speak of impossibility in relation to why things are impossible; thus we have the distinction between logical and causal impossibility. Logical impossibility arises due to contradiction, causal impossibility arises due to causal forces. 1+1=3 is logically impossible. In the expression X + Y = 3, where X & Y are whole numbers between 0 and 3, X can be 0, 1, 2 or 3 (thus they are logically possible). However, if we know Y is 2, X can only be 1 (all other numbers being logically impossible). Barring miracles, it's causally impossible for fire not to burn paper.

So the three senses for impossible are 1) absolute (given all relevant logical and causal factors), 2) logical impossibility (things which are impossible due to the rules of logic) and 3) causal impossibility (things which are impossible to the rules of causality). Causal and logical impossibility are subcategories of absolute impossibility (impossibility without qualification) and as we shall see, epistemic possibility is a subcategory of logical impossibility.

If I said “it’s possible a 50 lb weight would crush an unopened coke can”, it may sound as if I am talking about causal impossibility, but I am not. Granted the relationship between the coke can and weight is causal, not logical. But if causal laws are such that the coke can will withstand the pressure, the weight crushing the can is causally impossible. What we are dealing with here is my ignorance (have a rhetorical softball Steve). I don't know the causal relationship between the weight and the can, so given the data I have, I can’t deduct the truth about what would happen. So the “possible” here is logically possible, not causally possible.²

When we are dealing with “ignorance” and the mental projection of outcomes, “possibility” always relates logical possibility, and never causal possibility. We contemplate facts, truths and ideals and we reason through them logically. Thus the epistemic sense of possibility is a subcategory of logical possibility.

Back to Steve... Epistemic possibilities are things we think are possible (which may not in fact be possible). It this logical or causal possibility? It might look like Steve is saying we don't know if we can causally produce something or not (while leads us to think he's taking about causal possibility), but Steve's card player example is tell that he's speaking of logical possibility.

The card player's ignorance isn't related to what outcomes he can causally produce, it's related to what outcomes he can logically deduct. The epistemic sense of 'possible' relates to logical possibility; given this data, 20 is still logically open. Possible, in the sense the poker player uses, is a narrowing down of outcomes based on given facts, it doesn't relate to the poker players' causal power. Using logic, the player tries to project the outcome; but he can't, it's indeterminate from an epistemic standpoint. If he can't determine an outcome (i.e. he doesn't know it to be either necessary or impossible), he might say it's possible.

This issue is similar to the discussion Steve and I had about God's foreknowledge and future contingents; we must distinguish between truth and the basis of truth. The basis of truth is causal. Causal forces outside the player and the player's causal power will determine the outcome. When we say someone can choose, we are making a positive assertion about an agent's causal abilities. We are saying what the agent can cause. Logical possibility relates to ideas, not persons. A truth can't reach out and grab you, constrain you or causally determine what you do, but a person might.

So in the expression "the ability to do X", which makes more sense:

A) the lack of data to logically demonstrate statement's about X are false, or
B) a person's causal ability to produce X?

Similarly, in the dictionary definition of choice: "selection between possible alternatives" which makes more sense:

A) selection between [two or more things we don't have enough information to logically rule out], or
B) selection between [two or more effects we can causally produce].

Defining 'choose' is the primary debate and most of Steve's arguments hinge on it. But he mentioned a few other things I wanted to briefly touch on. He cited a definition of choice determinist could accept. Paul did the same thing. The primary issue isn't the noun, choice, it's the verb, choose. Second, Steve said I was guilty of a illegitimate totality transfer, in that I seek a one size fits all definition for choice. Not so, I am comfy with other definitions in other contexts. It's the context of the mental process of selection that I am interested in. Finally, Steve looks for Greek and Hebrew word studies. I have already pointed out that modern scolorship is unanimously translates bâcha and eklegomai as choose. I don't think anything is being lost in translation because choosing is an experience everyone has regardless of language or era. It's kinda like the word "porcupine". What I mean by porcupine and what my one year old son means when he points and says "ug" are the same thing. Same with anyone from any era and language that had a word for that spiky little critter. Perhaps there are other contexts for 'choice' besides the mental process we all experience. Words that get imported in to an alien discussion can be harder to define; like "sarx" which sometimes requires skinning the context to flesh out the meaning. But when you are naming an everyday experience, you stand on fairly solid ground.

Me: a person can be both classic Arminian and Molinist

Steve: No, he can’t. Classic Arminianism operates with simple foreknowledge rather than middle knowledge.

Middle knowledge can be found in the writings of Arminius, Episcopius, Grotius, Grevencovius, Goodwin, and Bird, to name a few from the early generation.

Dan: God does not choose between possible worlds, He chooses between hypotheticals (sometimes called feasible worlds).

Steve: God is the agent who chooses which world will be the actual world, and God is also the agent who actualizes a possible world.

I have no problem with your specific wording or the quotes you provided, but I still doubt we are on the same page here. One key difference between Molinism and Calvinism on the point of possible worlds is that in Molinism God can't choose all the possible worlds; some of them are possible as a result of the possibilities He gives us. In Calvinism, God can choose any possible world, because all possible worlds are indexed only and directly to His power. But in Molinism, if we would choose chocolate; God can't choose the possible world in which we choose vanilla. That's why I raised the possible/feasible world distinction; a distinction that I hope clarifies one of the differences between Molinism and determinism. I suspect Calvinists will not accept the idea that God can't actualize some possible worlds.

Steve: The Dan in a merely possible world is not a real person. He’s just a divine idea. And since he’s not a real person, he’s not a real agent. He makes not actual choices. Rather, God thinks of Dan making a choice, which is hardly the same thing as Dan making a choice. A possible agent doesn’t do anything–except in terms of imaginary action. It’s akin to the relationship between a novelist and the fictitious characters he conceives in his own mind.

Probably a closer analogy would be us hypothesizing what someone we know well would do in a given circumstance, but even that's imperfect. Ultimately, it's a unique ability God has.

Just because it isn't actual me, doesn't mean it's not hypothetical me. While hypothetical Dan isn't a real agent, doesn't mean he's not a hypothetical one with an exact logical correspondence to me. This of course is dependent on God's unique power of hypothesis.

While only an actual agent can actually choose, a hypothetical agent can hypothetically choose, which is enough. Indeed hypothetical Dan is quite busy, running around doing this and that and just about everything. However, for the reasons you point out, I suspect God chooses to run the hypothesis (rather than it just being something He naturally knows). There is more than one mind (God's) involved; God's is the only actual one, but there is a hypothetical mind that is not God and is based either in possible us or actual us. So it seems, in some sense based in creation (or at least what would be a creation) as opposed to just God's mind. In the author example you give, there's just the author's mind involved. For this reason, some say a general decree to create the individuals He has and will create comes before He has MK about them. Others say He can know what a person would do, based just on the possibility of creating the person. So there is some question as to if God knows what Santa Claus (a possible, but not actual person) would choose in various circumstances; some Molinists saying He does, others that He doesn't.

Steve: Moreover, actual Dan doesn’t have access to possible worlds, for actual Dan only obtains in the actual world, and the actual world reflects the divine actualization of one possible world to the exclusion of other possible worlds. That’s a fundamental difference between a possible world and an actual world. Actual Dan can’t choose contrary to the actual world. Rather, the actual world exemplifies one particular choice. Actual Dan can’t undo the actual world by opting for another.

Actualization takes place over time, not from the beginning with God's choice. It's actualized for the reasons and causes (predetermined or free) God saw would be the reasons and causes. God's choice doesn't directly determine contingencies, nor does He start a deterministic sequence of events. Other worlds can, but will not ( and would not) be actualized.

We can choose worlds contrary to God's chosen world, but His foreknowledge and MK cannot be deceived. Hypothetically, if we did choose the other worlds, God would have known them. Thus actualizing other worlds is possible, but the combination of us choosing a different world and it resulting in us deceiving God is incompossible.

Steve: Furthermore, Dan seems to envision a situation where, in each possible ice cream parlor, possible Dan contemplates all three alternatives, but chooses a different option in each case. That’s the only sense I can make of his statement that possible Dan has access to the other two possibilities–which he rejects. Possible Dan is considering all three possibilities at once.

God's scenario of us in the ice cream parlor removes the us not in the ice cream parlor scenario. (We are down to 2 not 3). So long as we are choosing between chocolate & vanilla, I suppose that does imply us contemplating those alternatives. But the possible worlds relate to the different choices, not the contemplation of the alternatives. I think we agree on this aspect of possible worlds.

I will take a rain check on discussing God and time. Not that it's over my head. It is over my head, but that's not why I am taking a rain check. I am taking a rain check because Steve and I agree on the point that started the discussion:

Me: Further, so long as the decree logically precedes the act, alternative possibilities have still lapsed. Given God's decree, there are no possible alternatives.

Steve: Yes, given God’s decree.

-------------------------------

¹ A natural question at this point would be "otherwise than what"? Normally it's otherwise than what we will do (when it's future or what we did if it's past), or for a determinist, otherwise than what we have been predetermined to do. In an epistemic sense of possible, perhaps one could say "otherwise than what I think is possible". All of these run across the problem mentioned above.
²Beware two equivocations. We use logic to projecting outcomes and causality often factors into projection. So while it’s causally impossible for me to jump over the moon, I can logically deduct that given the causal forces at play, its logical impossible for me to succeed in jumping over the moon. Although the process at arriving at that conclusion was logical, the factors considered were causal. So there is a marked difference between deriving, given causality, it’s logically impossible to jump over the moon and say deriving that it’s logically impossible to succeed in drawing a square circle. The first is a logical impossibility based on the rules of causality, the second is a logical possibility based on the rules of logic. But both are derived based on logic, so both fall in the catigory of "logical impossibility". This is one of the reasons “possibility” is susceptible to equivocation.

Here’s another way in which “possibility” is susceptible to equivocation. In Steve’s card player example, let’s say a blackjack player has counted all cards and the dealer is down to his last card. The player knows he’s up to 20 and he knows the dealer will give him a 2. Logically, he will bust if he asks for the card. But is it illogical to ask for the card? Yes and no. It’s illogical for him to win by asking for the card, so that’s an illogical way to play (given his goal is winning), but if he does ask for the card, there’s no logical contradiction. Thus “illogical” is susceptible to equivocation regarding the outcome and the rational for the outcome.