The Statistical Worldview

[integral]

The problem I have with this worldview is that it doesn't work in practice.

Go show me all the statistics on why cholesterol causes heart attacks, and then look at it again to understand all of it is wrong.

You can prove everything with statistics.

Edited 7 hours ago by integral

[zurew]

Some of you seem extremely confused about this.

The view is just something that represents credence in probabilities and affirms that there can be multiple contributing causal factors to things and it takes into account alternative hypotheses for any given thing.

 

You can have two people who have this view and they can still disagree on epistemic norms, epistemic approaches and they can disagree on the assigned probabilities to things.

The core idea is just to have a meta-frame that allows you to capture the nuance of the world. It doesnt gurantee anything about forming true beliefs about things in the world, its just given that it makes you entertain alternative hypotheses for any given thing, it gives you the possibility to explore more options (rather than just ignoring alternative options).

it makes you ditch the law of excluded middle and you represent most (if not all of your beliefs and hypotheses) in probabilities or credence talk (likely, not likely etc) rather than in a regular (true-false) dichotomy.

Edited 6 hours ago by zurew

[Joseph Maynor]

1 hour ago, zurew said:

Some of you seem extremely confused about this.

The view is just something that represents credence in probabilities and affirms that there can be multiple contributing causal factors to things and it takes into account alternative hypotheses for any given thing.

 

You can have two people who have this view and they can still disagree on epistemic norms, epistemic approaches and they can disagree on the assigned probabilities to things.

The core idea is just to have a meta-frame that allows you to capture the nuance of the world. It doesnt gurantee anything about forming true beliefs about things in the world, its just given that it makes you entertain alternative hypotheses for any given thing, it gives you the possibility to explore more options (rather than just ignoring alternative options).

it makes you ditch the law of excluded middle and you represent most (if not all of your beliefs and hypotheses) in probabilities or credence talk (likely, not likely etc) rather than in a regular (true-false) dichotomy.

Interesting.  But the statement X probability is reasonable to accept still obeys the true vs. false dichotomy.

[zurew]

53 minutes ago, Joseph Maynor said:

Interesting.  But the statement X probability is reasonable to accept still obeys the true vs. false dichotomy.

Thats just doing the move where you put the probability or the credence talk inside the proposition itself - and sure you can do that, and if you want I can make it more precise by saying  - if you include that kind of language inside the expressed proposition then those propositions will be true or false.

But my understanding is that ditching the excluded middle doesnt commit you to  saying that there isn't any single proposition that is true or false, its just saying "truth is not automatically bivalent for every proposition".  

But this is besides the point, the point is just to capture nuance, where there is nuance.

[Joseph Maynor]

6 minutes ago, zurew said:

Thats just doing the move where you put the probability or the credence talk inside the proposition itself - and sure you can do that, and if you want I can make it more precise by saying  - if you include that kind of language inside the expressed proposition then those propositions will be true or false.

But my understanding is that ditching the excluded middle doesnt commit you to  saying that there isn't any single proposition that is true or false, its just saying "truth is not automatically bivalent for every proposition".  

But this is besides the point, the point is just to capture nuance, where there is nuance.

I agree with the capturing nuance better proposition.  I'm not sure how probabilities are sourced.  It all depends on not lying with statistics.  That alone raises an epistemic problem.  I took statistics in college, but it comes down to how the number is generated, and whether or not has an epistemic justification.  To me this solidifies in the proposition X number is reasonably generated.  But that is a firm claim, which is either true or false, or at least people who allege this claim want it to be.  Thus my claim that even this kind of probabilistic reasoning presupposed the dichotomous true v. false idea.