The Origin of Life

Originally posted by bbarr
Draw a line between what and what? Between a belief based on evidence and a belief based on faith? If your evidence makes it significantly more likely than not that P is true, then you should believe P. That doesn't mean you ought to be so confident in your belief that you would not revise it in light of future evidence. If your evidence doesn't make it su ...[text shortened]... d that weight in regards to future evidence. There is no extra 20% to be made up by anything.

Again, you use the terms "significantly more likely", "substantially more likely" without actually saying what they mean.

Is a 90/10 split substantially more likely? 80/20? 60/40? 51/49? 50.1/49.9?

Originally posted by lucifershammer
Again, you use the terms "significantly more likely", "substantially more likely" without actually saying what they mean.

Is a 90/10 split substantially more likely? 80/20? 60/40? 51/49? 50.1/49.9?

That doesn't matter. The point is that your confidence in a proposition ought to be proportional to the degree to which your evidence makes that claim likely to be true. Didn't I just say this? The claims I made above that utilized the terms 'substantial' and 'signigicant' concerned the minimal conditions for a belief not to be based on faith. Minimally (and this is a sufficient condition), if your evidence doesn't establish that P is any more likely than ~P, then your belief in either P or ~P is faith based. If your evidence suggests that P is 60% probable, then you ought to be just that confident in the truth of P (no more, no less).

Originally posted by lucifershammer
Again, you use the terms "significantly more likely", "substantially more likely" without actually saying what they mean.

Is a 90/10 split substantially more likely? 80/20? 60/40? 51/49? 50.1/49.9?

Normally, in science, for a single testable hypothesis the value of p we use (the probability of the result we have occurring by chance is 5% (i.e. 0.05). This is typically the minimum statistical likelihood that we require.

Theories are built on numerous, independant investigations, thus the p-value of a theory being wrong is p=0.05^n-1, where n is the number of investigations that have taken place that substantiate that theory (of course, we'll abandon or modify a theory if it's wrong just once!). For example, if n=100 (i.e. 100 independant experiments verify the theory) then p=0.05^99. Therefore the chances of the theory being incorrect is somewhere around 1 in 1.5 *10^129.

Originally posted by bbarr
That doesn't matter. The point is that your confidence in a proposition ought to be proportional to the degree to which your evidence makes that claim likely to be true. Didn't I just say this? The claims I made above that utilized the terms 'substantial' and 'signigicant' concerned the minimal conditions for a belief not to be based on faith. Minimally (an ...[text shortened]... probable, then you ought to be just that confident in the truth of P (no more, no less).

What is the relationship between confidence and belief?

If the evidence suggests that P is 60% probable then, based on my reading of what you're saying, my belief in P is not an instance of faith. But you also say I should only have 60% confidence in P. How does that work? Obviously I cannot have a belief that is 60% P and 40% ~P.

So, if I believe P, I would believe that P is true (i.e. 100% probable or certain).

Or is there something I'm missing?

Originally posted by lucifershammer
What is the relationship between confidence and belief?

If the evidence suggests that P is 60% probable then, based on my reading of what you're saying, my belief in P is not an instance of faith. But you also say I should only have 60% confidence in P. How does that work? Obviously I cannot have a belief that is 60% P and 40% ~P.

So, if I beli ...[text shortened]... believe that P is true (i.e. 100% probable or certain).

Or is there something I'm missing?

I didn't say that if the evidence makes P 60% probable then the belief that P isn't faith based. I said that it is sufficient for a belief to be based on faith that that belief be held despite the fact that one's evidence doesn't make P any more probable than ~P.

A belief is just an attitude one has towards a mental representation. The attitude is one of minimal endorsement; of taking it to be the case that the representation is accurate, or that the propositional content of the belief is true. But this endorsement is not an all or nothing affair; one may be only somewhat confident that P, or very confident that P, or...., and this confidence ought to be proportional to the strength of one's evidence. You could just read the claim "I believe that P" as "I believe that it is X% probable that P".

Originally posted by bbarr
You could just read the claim "I believe that P" as "I believe that it is X% probable that P".

I don't think this is accurate. Belief is a binary discretization of probabilites at the 50% (or possibly higher, depending on the desired connotation) threshold. You can't believe that P if you hold that it is 20% probable that P and thus 80% probable that not-P. Thus, it is improper to read the claim as you suggest.

Further, your suggested read of the claim has no base case, for how do you interpret the interpretation if not as "I believe that it is Y% probable that it is X% probable that P."

A better read of the claim would be "From my set of information, it is more probable that P than not-P."

Originally posted by DoctorScribbles
I don't think this is accurate. Belief is a binary discretization of probabilites at the 50% threshold. You can't believe that P if you hold that it is 20% probable that P and thus 80% probable that not-P. Thus, it is improper to read the claim as you suggest.

Further, your suggested read of the claim has no base case, for how do you interpret ...[text shortened]... e interpretation if not as "I believe that it is Y% probable that it is X% probable that P."

Minimal endorsement is a state where one takes it to be the case that P is more likely than ~P. So, the variable X can only be replaced with values above 50.

I have no idea what your 'base case' question means.

Originally posted by bbarr


I have no idea what your 'base case' question means.

You have defined belief circularly.

You said that "I believe that P" should be read as "I believe that it is X% probable that P."

But how should "I believe that it is X% probable that P" be read?

Well, it is of the form "I believe that P", since "It is X% probable that P" is itself a proposition. So then it must be read as "I believe that it is X% probable that it is X% probable that P."

Given that LH doesn't understand belief to begin with, your definition does not help him, for he must continue this process ad infinitum, never producing a statement that he understands, but always a statement of the form "I believe that P," which he doesn't understand to begin with.

I think you need to eliminate the term belief from the original suggested reading and replace it with a term that LH is comfortable with which represents one's assessments of probabilities.

How about "I believe that P" should be read as "I have assessed that P is at least 50% probable."

Originally posted by lucifershammer

If the evidence suggests that P is 60% probable then, based on my reading of what you're saying, my belief in P is not an instance of faith. But you also say I should only have 60% confidence in P. How does that work?

It works just like the concept of freezing.

Water freezes at temperatures below 0 and doesn't freeze at temperatures above 0.

If the temperature is -10, then water freezes at that temperature.
If the temperate is 10 then water doesn't freeze at that temperature.

If I assess the probability of some proposition P to be .7, I believe it.
If I assess the probability of some proposition P to be .3, I disbelieve it.

It's just a mapping from a set of numbers to a binary characterization of them.
{t < 0} => freezing
{t > 0} => not freezing

{p < .5} => disbelieve
{p > .5} => believe

The mapping to belief/disbelief masks the underlying probabilities, but they are still there, just as you can still take the temperature of a piece of ice or a glass of water after having declared it to be frozen or unfrozen.

Further, just as you can characterize unfrozen water as cold, cool, warm, hot, scalding, based on its temperature, so can you characterize your confidence levels in things that you believe based on their probabilities. Just because water is frozen doesn't mean it is at a suitable temperture to drink - it could be scalding. Just because you believe something doesn't mean you should rely on that belief when making decisions - don't make life altering choices based on beliefs of propositions in which you have 60% confidence.

Originally posted by DoctorScribbles
You have defined belief circularly.

You said that "I believe that P" should be read as "I believe that it is X% probable that P."

But how should "I believe that it is X% probable that P" be read?

Well, it is of the form "I believe that P", since "It is X% probable that P" is itself a proposition. So then it must be read as "I believe that hat P" should be read as "I have assessed that P is at least 50% probable."

Translations are not definitions. If you believe that it is, say, 60% probable that P, then you are claiming that your evidence establishes with 60% probability that P. If you are unsure about just what is included in your evidential set, then you very well may want to say that it is, say, 80% percent probable that it is 60% probable that P. But I'm not sure why you'd want to do that, since you could just do the math.

Originally posted by bbarr
Translations are not definitions.

But your translation is not helpful to somebody who doesn't understand the term believe.

You have left "I believe" intact and have translated "P" into "it is X% probable that P." But "I believe" is the part that is misunderstood.

The "I believe" is what he is asking you to translate. I propose that it be translated as "I have assessed that it is more likely than not."

Originally posted by DoctorScribbles
But your translation is not helpful to somebody who doesn't understand the term believe.

You have left "I believe" intact and have translated "P" into "it is X% probable that P." But "I believe" is the part that is misunderstood.

The "I believe" is what he is asking you to translate. I propose that it be translated as "I have assessed that it is more likely than not."

Somebody who doesn't understand the term 'believe' ought to learn the language. As I said above: To believe a proposition is to stand in the relation of minimal endorsement to the propositional content of a mental representation.

Originally posted by DoctorScribbles
But your translation is not helpful to somebody who doesn't understand the term believe.

You have left "I believe" intact and have translated "P" into "it is X% probable that P." But "I believe" is the part that is misunderstood.

The "I believe" is what he is asking you to translate. I propose that it be translated as "I have assessed that it is more likely than not."

Your translation sucks. It precludes people from believing propositions which result from brute causal forces (where there is no assessment). We often just come to believe things, without ever having assessed their likelihood.

Originally posted by bbarr
But I'm not sure why you'd want to do that, since you could just do the math.

With your provided translation, the reader who does not understand the term believe has no other choice, since he does not understand "I believe that it is 60% probable that P." (If he understood this, he wouldn't misunderstand "I believe that P" in the first place.) And he can't stop at that level - he has to keep applying the translation, layer after layer.

Originally posted by bbarr
Your translation sucks.

Yours sucks a bigger one. At least mine terminates with something the reader understands.