Saturday, March 17, 2018

Post-EAAN Debate Reflections (p. 3)

Post-EAAN Debate Reflections
< Prev   |   1   |   |   3   |   Next >

The Debate



Property Dualism



Epiphenomenalism is the view that states that consciousness and other mental states are a mere accessory to the neurophysiological processes whose presence or absence makes no difference. Semantic epiphenomenalism (SE) says that beliefs have two properties: syntax (the neurophysiological or NP properties) and semantic (e.g. the belief that p for some proposition p, e.g. the belief that Snow is white). A minor point, but at around 50:33 to 50:58 Spencer disagreed with me that SE makes this dualist property claim, and that I was going against how Plantinga described it. From pages 6 through 7 of Naturalism Defeated?
A second possibility is semantic epiphenomenalism: it could be that beliefs have causal efficacy with respect to their behavior, but not by virtue of their content. Put in currently fashionable jargon, this would be the suggestion that beliefs are indeed causally efficacious, but by virtue of their syntax, not by virtue of their semantics. On a naturalist or at least a materialist way of thinking, a belief could be something like a long-term pattern of neural activity, a long-term neuronal event. This event will have properties of at least two different kinds. On the one hand, there are its neurophysiological or electrochemical properties: the number of neurons involved in the belief, the connections between them, their firing thresholds, the rate and strength at which they fire, the way in which these change over time and in response to other neural activity, and so on. Call these syntactical properties of the belief. On the other hand, however, if the belief is really a belief, it will be the belief that p for some proposition p. Perhaps it is the belief that there once was a brewery where the Metropolitan Opera House now stands. This proposition, we might say, is the content of the belief in question.[2]
So, I was right.

Semantic Externalism



In about 20:33 to 24:01 of the debate my interlocutor (Spencer Hawkins) tried to use semantic externalism to attack premise (1) of this argument:
  1. (On naturalism) If a belief’s associated NP properties had any different semantic content associated with it instead, the same outcome in the physical world (e.g. one’s behavior) would result.
  2. If (1) is true, then a belief’s semantic content is causally irrelevant (on naturalism).
  3. Therefore (On naturalism), a belief’s semantic content is causally irrelevant.
What is semantic externalism? To explain I’ll use the famous Hilary Putnam’s famous Twin Earth scenario. His scenario goes something like this (I’ve modified it only slightly): suppose there is a Twin Earth that is identical to ours prior to the discovery of water being H2O except what its inhabitants call “water” is actually a complex chemical compound XYZ that looks and behaves just like water (rain has XYZ instead of water, lakes have XYZ instead of water, etc.). Let’s go back to 1750 before we learned the chemical composition of “water” is H2O. Two individuals, one on the actual Earth and one on Twin Earth, have identical mental states when they think, “That over there is water.” Nonetheless, in a way the sentences “That over there is water” have different semantic content between Earth and Twin Earth because they each refer to different chemical substances. While a belief’s semantic content is still partly dependent on one’s mental states, in at least some cases it is also dependent on one’s external environment. So, premise (1) is false; it’s possible to have the same NP properties but different semantic content and the outcome be different due to the external component of semantic content (and thus some other part of the world) be different. (Steven uses a different example involving treasure maps, but I prefer the classic Twin Earth scenario.)

The problem with this objection is that it largely misses the point. Suppose it’s true that in at least some cases, semantic has both an external and an internal (i.e. part of the mental state) component just as it does in the Twin Earth scenario. By “semantic content” in premise (1) I had in mind only that part of the semantic content that is part of one’s mental states, and if that semantic content could be literally anything at all without affecting behavior—even garbage beliefs—then we still get the same problem particularly when garbage mental states (akin to the mental states one has in dreams) vastly outnumber those that align with one’s external environment. I tried to explain to Spencer that by “semantic content” I had in mind only that part of the semantic content that is part of one’s mental state, but he apparently had some difficulty grasping the concept, so I switched the argument to be like this, replacing “semantic content” with “mental state.”
  1. (On naturalism) If a mental state’s associated NP properties had any different mental state associated with it instead, the same outcome in the physical world (e.g. one’s behavior) would result.
  2. If (1) is true, then mental states are causally irrelevant (on naturalism).
  3. Therefore (On naturalism), mental states are causally irrelevant.
Mental states include sense experience, memories, and the semantic content of beliefs (at least, that part of semantic content that is part of one’s mental state). This aspect of the debate largely inspired my article Mental States are Causally Irrelevant on Naturalism.

Plantinga’s Objective Probability



Spencer raised the objection that applies for theists who believe God is metaphysically necessary. On the sort of probability that Plantinga (the originator of the EAAN) normally uses which he’s called “objective probability” a claim being metaphysically impossible means it has a probability of zero. So for the theist who believes God is metaphysically necessary, naturalism is false in all possible worlds, which means N&E is false in all possible worlds, which (on the typical probability axioms) Pr(R|N&E) is undefined because it would mean dividing by zero. I agreed with Spencer that the theist who believes God is metaphysically necessary can’t reasonably affirm the Probability Thesis when using Plantinga’s type of objective probability.

But it’s not a very serious problem. First, this objection won’t help the naturalist all that much because on this objection the Pr(R|N&E) is undefined by God existing in all possible worlds, and if the naturalist is to concede that God exists in all possible worlds (including the actual one), the falsity of the Probability Thesis by this means is a very Cadmean victory for naturalism. Second, other types of probability are available. In the debate I used the fact that mathematics can give us a probabilistic primality test (the Miller-Rabin primality test in particular) and quantify a high probability (high but less than 100%) that a certain number is prime. Assuming math theorems give us objective truths, that probability is in a real sense objective, it’s probable in a way that’s independent of human opinion. But that sort of objective probability wouldn’t be Plantinga’s objective probability as he defined it, since on Plantinga’s sense of objective probability the objective probability of the number being prime is either 0% or 100% (a number is either prime in all possible worlds or composite in all possible worlds). We can call this type of probability, one of objective evidential support relations, as objective epistemic probability. It’s true when Plantinga originally introduced EAAN he was using his variety of objective probability, but he also claimed the argument can work with epistemic probability, and I agree; the justification I gave for the Probability Thesis does seem to work for objective epistemic probability.

I found it odd that Spencer seemed to think his objection posed a serious problem after how I explained why it wasn’t (e.g. 1:55:51 to 1:56:41). Spencer (in 1:56:19 to 1:56:41) was apparently unconvinced that my justification for the premises that argued for the Probability Thesis worked for epistemic probability, but was extremely vague about why anyone should doubt that such justification works for epistemic probability.

Supervenience Thesis



To illustrate what supervenience means, mental states supervening on brain states means that there cannot be a difference in mental states without there being a difference in brain states. On naturalism is true, I think it is very likely that mental states supervene on brain states due to some set of physical laws; i.e. that mental states supervene on brain states as a matter of physical necessity. What philosophers call strong supervenience is a supervenience relation that holds in all possible worlds; e.g. mental states strongly supervening on brain states means that there is no possible world where a different mental state is associated with the same brain state. The supervenience thesis (as Spencer used the term) claims that mental states strongly supervene on mental states. Even if I were a naturalist, I wouldn’t find the supervenience thesis plausible; electrochemical reactions generating a different mental state seems too easily conceivable to me, with no good reason (it seems to me) to think that the relationship couldn’t be different in any other possible world, any more than thinking that the physical laws themselves couldn’t have turned out differently.

Still, let’s leave that aside and suppose the supervenience thesis is true. Spencer claimed that the supervenience thesis, if true, would render premise (1) false of this argument (1:36:22):
  1. (On naturalism) if any different mental states were associated with the same physical conditions, the same outcome in the physical world (e.g. one’s behavior) would result.
  2. If (1) is true, then mental states are causally irrelevant (on naturalism).
  3. Therefore (on naturalism), mental states are causally irrelevant.
The context here is that we were discussing panpsychism (the view that all matter has some degree of consciousness—if that view sounds wacky to you, you’re not alone) and I was basically asking what would happen if we had all the same physical conditions including all the same laws but different mental states were associated with those conditions—would we on naturalism get the same outcome (1:33:58 to 1:34:40)? My answer: yes, of course we would, because we have the same initial conditions, same physical laws (by definition, physical laws correctly say what will occur given certain physical initial conditions in the absence of any supernatural intervention), and when naturalism is true there’s no supernatural intervention to change what would occur. Thus, premise (1) seems obviously true.

Spencer’s objection to premise (1) was the supervenience thesis.

The problem, as you might suspect, is that the truth or falsity of the supervenience thesis is irrelevant to the truth of premise (1). It’s not as if premise (1) is saying that it’s metaphysically possible for different mental states to be associated with the same physical conditions. It’s only saying that if different mental states were associated the same physical conditions, on naturalism we’d get the same outcome. Even if it were metaphysically impossible (somehow) for different mental states to be associated with the same physical conditions, it’s certainly conceivable that e.g. some different mental state be associated with the same brain state; there’s no self-contradiction there. And we can conceive of what would happen if, on naturalism, the same physical conditions had different mental states associated with them.

Note that while it is conceivable for different mental states to be associated with the same physical conditions on naturalism, it is not conceivable to get a different outcome on naturalism. Why? By definition, physical laws correctly say what will occur given certain physical initial conditions in the absence of any supernatural intervention. Since, as I specified in the thought experiment, we were dealing with the same physical laws, it would be logically contradictory to have naturalism be true and have the same physical conditions (including the same laws) and getting a different outcome. [3]

I had a difficult time getting Spencer to understand that the supervenience thesis is irrelevant. Fortunately, the Capturing Christianity moderator Cameron was (somehow) able to break through Spencer’s mental fog. If you want to see me struggle and fail while the moderator succeeds, see 1:33:58 to 1:42:04 of the debate.

Nonpropositional Evidence



Sometimes you know your beliefs to be true because they are justified on the evidential basis of other beliefs (e.g. facts presented in a criminal trial), and in that case you’re dealing with propositional evidence. Nonpropositional evidence is evidence that comes from something other than the inferential basis of other beliefs; e.g. the experience of being appeared to redly (roughly, “It seems to me that I’m having a sensation of redness”) justifies your belief that you are being appeared to redly. The experience of me remembering I had breakfast this morning justifies my belief that I had breakfast this morning. I don’t infer my memory from another belief; I just intuitively experience it. Spencer claimed nonpropositional evidence could save R from defeat (1:57:22), but was fuzzy on the details.

I anticipated that Spencer would bring up nonpropositional evidence in his closing statement, so in 1:54:40 I explain that N&E is like ingesting drug XX and that even if I get nonpropositional evidence for R sometime after ingesting the drug (whatever that might be—a strong feeling that R is true for me perhaps?), R is still defeated for me. I can very strongly feel that I have accurate mental states but as long as that happened long after I ingested drug XX (or the equivalent; e.g. the XX-mutation) I still have an undefeated defeater for my mental states aligning with how the external world is.

So I basically rebutted his point before he made it in his closing statement and he didn’t really respond to my rebuttal, nor did he give any details about what this nonpropositional evidence might be or how on earth it could save R from defeat. It seems very unlikely that there’s any viable way to save R from defeat here.

Conclusion



All things considered I think the debate went rather well for me. One weakness is that my rusty memory failed to recall what neutral monism is (he brought it up in the debate) through I did remember panpsychism as well as nearly every other philosophical tidbit he brought up, including having a more accurate memory of semantic epiphenomenalism. It was interesting how at one point he eventually agreed with me that the Defeater Thesis is true after the Capturing Christianity moderator explained it to him (1:16:15 to 1:21:55) then we went back to argue the Probability Thesis until he apparently couldn’t think of any good objection against my argument for it (1:52:03 to 1:53:00) and then disputed the Defeater Thesis (appealing to vague nonpropositional evidence)! For the most part, my position seemed to be on stronger intellectual ground and I had the means to justify both the Probability Thesis and Defeater Thesis fairly convincingly.

I was genuinely impressed with the moderator’s skills at moderating; in addition to keeping things on track, a couple times he was able to explain things to Spencer when I was unable to! The moderator was also good at trying to fight bias against his own side by giving my interlocutor the last word. If I ever have a debate on my own YouTube channel I’d want him as the moderator.



[2]Beilby, James (editor). Naturalism Defeated? (Ithaca, New York: Cornell University Press, 2002), p. 7-8.

[3] What would “outcome” mean in a naturalistic and truly random universe (where different states of affairs are possible with the same initial physical conditions)? Suppose the laws say that 50% of the time the resulting state of affairs will randomly be state of affairs S1 and the other 50% of the time the resulting state of affairs will randomly be state of affairs S2. In that case, the “outcome” is 50% of the time it will be S1 by chance, and 50% of the time it will be S2 by chance, and that outcome will not be rendered different by having a different mental state associated with the same conditions (by which I would mean the “on average” frequencies would not be rendered differently). In retrospect I perhaps should have used the word “consequences” rather than “outcomes.”

< Prev   |   1   |   |   3   |   Next >

No comments:

Post a Comment