Isn't the Lockean Thesis clearly false?

Liz Jackson writes ("The relationship between belief and credence", Philosophy Compass, 2020): 

Lockean Thesis (traditional): S ought to believe p iff S has a rational high credence in p. 

    On this version of the Lockean thesis, a rational high credence is necessary and sufficient for rational belief. This does justice to the intuition that rational agents are confident in the propositions that they believe. 

    Nonetheless, worries loom. The first is the lottery paradox (Chandler, 2010; Douven, 2012; Douven & Williamson, 2006; Kyburg, 1961; Leitgeb, 2014b; Smith, 2010a; Weintraub, 2001; Weisberg, 2015, Section 5). Suppose you have a lottery ticket; uncontroversially, you should have a high credence your ticket will lose (a fair 100-ticket lottery puts your credence at 0.99, and we can increase the number of tickets). If the threshold for belief is any value short of one, then, according to the traditional Lockean thesis, there are lotteries in which you ought to believe your ticket will lose. But your ticket is not special: by the same reasoning, you should believe, of each ticket, that it will lose. Further, many think that if you should believe p and you should believe q then you should believe p and q; this is an example of a closure principle (Kvanvig, 2006; Luper, 2016). Given certain closure principles, you should believe a large conjunction: ticket 1 will lose and ticket 2 will lose and...ticket n will lose. Nevertheless, you know one ticket will win, and thus you should also believe the negation of this conjunction. But intuitively, you should not believe contradictions. In response, some have suggested that the traditional Lockean thesis is false, and you ought not believe lottery propositions, such as my ticket will lose, even though you have a rational high credence in them (Douven, 2002; Jackson, Forthcoming; Kelp, 2017; Ryan, 1996; Smith, Forthcoming; Staffel, 2015). 

Let's disambiguate the thesis:

 

*Lockean Thesis (traditional): S ought to believe necessarily p iff S has a rational high credence in p.

 

Then I should believe that each ticket will necessarily lose. But that's clearly a false belief. Each ticket will not necessarily lose, only very probably. I shouldn't believe that my ticket will lose, only that it very probably will.

 

**Lockean Thesis (traditional): S ought to believe very probably p iff S has a rational high credence in p.

 

That's better, though a) terms like 'ought' and 'rational' need to be unpacked before I understand what this thesis is saying and b) arguably it's impossible to not believe very probably p while having a high credence in p. 

 

Though that implies one cannot have credences without also having beliefs. It does seem to me that credences are by far the more complex mental state between the two, with belief being a simple and common representational mental state. Anecdotally I see folks admit to believing something (say, God's existence) much quicker than admitting to a specific confidence level in that belief.

 

See for example WLC (https://www.reasonablefaith.org/media/reasonable-faith-podcast/questions-on-greatest-being-theology-certainty-and-the-moral-argument):  

Kevin Harris: Here is another question, Dr. Craig:

Greetings. In your debate with Christopher DiCarlo, he asked if you could estimate your level of confidence in your belief in God, and that if your belief was as likely as a universe with no God, or if it was higher. You said your belief was higher than 50/50 but that you had no way of measuring whether it was highly probable or not. Wouldn’t this seem to be of utmost importance to the question and what you have spent your life debating? How on Earth are we to accept that you find the existence of a God more probable when you have no way of knowing what more probable means?

Dr. Craig: Well, wait a minute. I never said I don’t have any idea of what more probable means. For something to be more probable than not means you have a greater than 50% chance of being true. Right? That’s what it means to be “more probable.” What I’ve refused to do is to quantify it and to say I am 75% certain or it is 83% certain that God exists or 99% certain. I think anybody who does try to put those kind of numbers on it is being disingenuous, and you ought to be very suspicious. The fact is we can talk, I think, in only rough terms about this, and saying “I think it is very probable that God exists” or “more probable than not.” Things of that sort. I think that is quite acceptable.

So Craig is happy to say that he believes in God and is happy with vague language like "It's more likely than not that God exists" or "It is almost certain that God exists", but Craig doesn't like giving specific numbers to it, because of how arbitrary that would be.

Also consider what Craig says here (https://www.reasonablefaith.org/media/reasonable-faith-podcast/debate-with-dr.-kevin-scharp-part-1): 

KEVIN HARRIS: . . . The first thing that kind of struck me that I wanted to run past you is he said, “We should be formulating theism and atheism in terms of confidence levels from zero percent to 100 percent, not in terms of belief.” Kevin really complimented you on building systems.[3] He wants to build a system here it seems – a philosophical system, systematic theology, systematic philosophy. But the argument itself is going to require confidence. Are you confident in the confidence?

DR. CRAIG: I'm not! And I made that point over and over again. I have no confidence whatsoever in his claim that in order to believe in God I need to have a probability significantly higher than 51%. He gave no argument for that other than just an example. He gave an illustration that he thinks it is 51% probable that Hillary Clinton will win the presidential election. But that is not enough for him to believe that she will win. Well, fine, that is just an illustration. But in many other cases, 51% might be enough for a person to believe in something, especially (as William James pointed out) if this is a pressing conclusion that imposes itself upon you and is a matter of great urgency and importance, William James argued that in a case like that you are perfectly rational to believe in something even without evidence. This is James' famous essay, “The Will to Believe.” This idea that in order to believe something you have to be significantly more than 51% confident in it, I think is person-relative and subjective. For some people that might be true, for other people it might not. For some beliefs it might be true, but for other beliefs (as James said) it may not. 

Rejecting the Lockean Thesis has the consequence that even if something is incredibly likely to happen, say a 99% chance of rain, you can't believe that it will rain. But isn't that crazy? You totally can plan as if it will rain, act as if it will rain (toss an umbrella into your car before you drive off, etc.).

 

But of course it's silly to believe both that there is a less than 100% chance of something happening and that it will necessarily happen. Instead, you believe that there is a 99% chance of rain or that it's very likely to rain. And such probabilistic beliefs perfectly justify the same actions, or at least extremely similar actions, as the actions justified by full belief (belief in the necessity of or certainty of). 

 

I'm not sure Craig is correct when he says the threshold for belief is context-dependent. But I think acting as if something will happen need not require belief. Craig mentions William James' idea of the leap of faith. But the person who leaps the gap to escape the avalanche does not believe that they will make the jump. All they need to believe to make sense of jumping is that jumping is better than doing nothing. It's not that the bar of belief is lowered based on context, it's just that what you believe changes based on context. The desperate skier who makes the jump doesn't jump because he or she believes they will make it or will very likely make it (like in ordinary circumstances), but, again, only that jumping is better than doing nothing.

 

A standard definition of belief is given by Liz Jackson in the article:

 

"To believe something is to regard it as true or take it to be the case . . ."

 

But this can sound like: "To regard something as true is to regard it as true" or "To believe something to be the case is to believe it to be the case", both tautologies. The point being that defining 'believe' in terms of 'regarding' or 'taking' (or 'accepting' or 'assenting to') doesn't seem helpful. (I'm not saying Liz Jackson's definition of belief is not helpful; I'm sure she could give a much better gloss on 'belief'; it's the standard definition that is not helpful.) 

 

So I lean toward the view that 1) Propositions are representational (i.e. propositions claim a match between external properties and "internal" properties; if there is a match then the proposition is true, if there is a mismatch then the proposition is false); 

 

2) Non-propositional beliefs are representational via mental states (i.e. the internal properties are internal to a mental state, not to a proposition);

 

3) Propositional beliefs are representational via propositional attitude (the internal properties are internal to the proposition that corresponds to the belief; OR the internal properties are internal to a mental state which corresponds [contains? represents? constitutes?] to a proposition);

 

4) Truth itself is representational, connecting truth and belief;

 

5) Beliefs do not come in degrees in the sense that you can partially believe something, but beliefs do come in degrees in the sense that you can believe in stronger or weaker claims;

 

6) In the case of believing p, not only do we believe p in the sense that our mental states represent (or attempt to represent) reality as described by p (details to be specified), but we also believe whether p is (epistemically) necessarily true or probably true (or lack such a belief, or only believe dispositionally that p is necessary, etc). Credences are levels of confidence, and are beliefs about how likely p is true. Beliefs and credences appear to come apart because the belief that p and the belief (or dispositional belief) that p is necessary (or 10% probable or highly probable, etc.) come apart, and because the additional modal belief or probability belief is not necessary to hold the belief (which is why Craig can believe in God without believing that God's existence is X% likely; granted, Craig probably believes it's not epistemically necessary that God exists).

 

Jordan Peterson infamously struggles to answer the question of whether God exists, and complains about vagueness over terms 'God' and 'exists'. (Which is fine as long as you proceed to offer some help in defining those terms, as otherwise you announce that you don't have anything thoughtful to say.) I suspect Craig similarly struggles to give God's existence an explicit percentage because of problems of vagueness. Precise attitudes don't fit imprecise propositions. There's also the problem of the complexity of the question. The more complicated a debate is, the harder it is to feel confident not only that one position is right, but it's harder to feel confident that you even understand what the debate really amounts to (is the debate itself based on false assumptions? e.g. like when 'nature' is falsely pitted against 'nurture' for explaining human behavior) and what it would mean to take a stand on one position versus another.