5 ms·

I wish more people had read that book. Anybody with some education in philosophy figures out that utter, logical-proof certainty can't be had. So what does one

I wish more people had read that book. Anybody with some education in philosophy figures out that utter, logical-proof certainty can't be had. So what does one do for epistemology instead? There are two main alternatives: -- Religious-style faith. This is not my preferred choice. -- An aesthetically-tinged approach to epistemology. What I mean by the latter is, for example, generalizing Occam's Razor into usability. The problem with Occam's Razor is that it says, in effect, "In case of doubt go with the simpler answer", without giving a general way to judge what's simpler. Any solution to that problem winds up being an aesthetic kind of judgment.

abritinthebay 9y ago
Except that's not what Occam's Razor says and that problem comes from the awful and inaccurate simplification of it.
- derefr 9y ago
  It's sort of what the Minimum Message Length formulation says. It's hard to communicate the "essential meaning" of Occam's Razor in simple language, because the only good analogies for it involve things like compression algorithms, quantum physics or topology. If you know what the word "axiom" really means, and can picture an address of a thing in a set taking up space (e.g. bytes of a URL) as your query gets more specific, then Occam's Razor is a very "obvious" statement about the probability of reification of mathematical objects in our universe given their size. If you don't have that context, someone can talk for an hour and it won't communicate the point.
  - abritinthebay 9y ago
    > It's hard to communicate the "essential meaning" of Occam's Razor in simple language It's really not. Here's a few that don't make the same mistake. - Plurality must never be posited without necessity. - Among competing hypotheses, the one with the fewest assumptions should be selected. - Entities must not be multiplied beyond necessity. - Whenever possible, substitute constructions out of known entities for inferences to unknown entities. - favor hypotheses that make the fewest unwarranted assumptions about the data from which they are derived. All of these are more correct and communicate the principles better than the "simplest" formulation which is just flat-out wrong without extremely heavy clarification of what "simplest" means (at which point you may as well use one of the above).
    - CurtMonash 9y ago
      See my other comment.
    - sideshowb 9y ago
      Quantum many worlds violates most if not all of those everyday language formulations. Doesn't violate the minimum message length one, though. I've not heard that one before & I like it.
      - andybak 9y ago
        I can't find the source of the quote now but a famous physicist once critiqued The Many Worlds interpretation along the lines of "It might be light on assumptions but it's very heavy on universes"
      - abritinthebay 9y ago
        > Quantum many worlds violates most if not all of those everyday language formulations. This is a common objection to which supporters would say means you're not reviewing the data in enough depth. It's a fair argument tho. Given Occam is supposed to only be a guideline and not some kind of universal law it wouldn't actually matter if QMW did violate it though (even though that's debatable anyhow).
    - Chris2048 9y ago
      > without extremely heavy clarification of what "simplest" means kolmogorov complexity? But then there is no universal measure, so it's relative.
      - abritinthebay 9y ago
        You could formulate a version that used kolmogorov complexity, that's perfectly valid. It would be a narrower, less generally applicable, version... but a completely valid formulation.
- alasdair_ 9y ago
  True. Occam's Razor was an attempt at proving God exists. Hardly the most common usage now, however.
  - abritinthebay 9y ago
    That's... a bit of an over simplification. It was a common statement of a generally justifiable philosophical tool that happened to be commonly used by theologians like Occam (who wasn't that first to use the idea - Aristotle stated something similar). The intertwining of philosophy and theology makes it related but I'd say your characterization was simplistic and quite incorrect overall.
- CurtMonash 9y ago
  The actual Occam's Razor is useless unless adapted. How does one count "assumptions" -- if we take Wikipedia's word for what it states -- or "entities" -- if we take the version I first learned? Most versions that actually work can be lightheartedly equated to a "Principle of Least Foofarah".
  - abritinthebay 9y ago
    Sure, you do you and academics will continue to use formulations that have worked for about 2500 years or so while you claim it's useless. Yeesh.
GavinMcG 9y ago
Why does one need it to be aesthetically tinged? Another approach is virtue epistemology, where we count as true that which trusted sources and processes say is true. That is, the drug dog who has been right in the past is one we listen to in the future. The scientific project that makes predictions that turn out to be accurate is the one we listen to.
- erikpukinskis 9y ago
  > That is, the drug dog who has been right in the past is one we listen to in the future If we're in a situation where we get good data about what's right, we can just do science and it doesn't really matter what authorities or dogs think. The place where you need non-scientific epistemologies is when you don't have good feedback data. And sometimes even when there is data, bad predictors can mutually self reinforce. Racist cops arrest more black kids, more black kids get convicted, that proves racist cops are good predictors of criminality. It's not true, but the data says it's true. It's equivalent to a scientist doing many studies and only publishing the ones that are positive. The stats only work for independent measures but nothing is independent especially when you are using predictions to make policy.
  - GavinMcG 9y ago
    Those are fair criticisms, but they can be addressed without demanding aesthetic judgments as a component of epistemology.
- anigbrowl 9y ago
  Scientism is good for making better predictions, but it won't tell you how to live. You could be a utilitarian and declare that you will live to the end of maximizing your utility (however you define that) but the problem with utilitarianism is that omniscience is elusive.
  - GavinMcG 9y ago
    That means that there are certain things which aren't answerable, at least not easily or precisely. But that isn't a problem with the epistemology you adopt: some things are going to be hard or impossible to know no matter how you define knowing. Or, "how to live" isn't an epistemological question. So demanding that an epistemology accept aesthetic standards to accommodate concerns about how to live is confused.

abritinthebay 9y ago

Except that's not what Occam's Razor says and that problem comes from the awful and inaccurate simplification of it.

derefr 9y ago
It's sort of what the Minimum Message Length formulation says. It's hard to communicate the "essential meaning" of Occam's Razor in simple language, because the only good analogies for it involve things like compression algorithms, quantum physics or topology. If you know what the word "axiom" really means, and can picture an address of a thing in a set taking up space (e.g. bytes of a URL) as your query gets more specific, then Occam's Razor is a very "obvious" statement about the probability of reification of mathematical objects in our universe given their size. If you don't have that context, someone can talk for an hour and it won't communicate the point.
- abritinthebay 9y ago
  > It's hard to communicate the "essential meaning" of Occam's Razor in simple language It's really not. Here's a few that don't make the same mistake. - Plurality must never be posited without necessity. - Among competing hypotheses, the one with the fewest assumptions should be selected. - Entities must not be multiplied beyond necessity. - Whenever possible, substitute constructions out of known entities for inferences to unknown entities. - favor hypotheses that make the fewest unwarranted assumptions about the data from which they are derived. All of these are more correct and communicate the principles better than the "simplest" formulation which is just flat-out wrong without extremely heavy clarification of what "simplest" means (at which point you may as well use one of the above).
  - CurtMonash 9y ago
    See my other comment.
  - sideshowb 9y ago
    Quantum many worlds violates most if not all of those everyday language formulations. Doesn't violate the minimum message length one, though. I've not heard that one before & I like it.
    - andybak 9y ago
      I can't find the source of the quote now but a famous physicist once critiqued The Many Worlds interpretation along the lines of "It might be light on assumptions but it's very heavy on universes"
    - abritinthebay 9y ago
      > Quantum many worlds violates most if not all of those everyday language formulations. This is a common objection to which supporters would say means you're not reviewing the data in enough depth. It's a fair argument tho. Given Occam is supposed to only be a guideline and not some kind of universal law it wouldn't actually matter if QMW did violate it though (even though that's debatable anyhow).
  - Chris2048 9y ago
    > without extremely heavy clarification of what "simplest" means kolmogorov complexity? But then there is no universal measure, so it's relative.
    - abritinthebay 9y ago
      You could formulate a version that used kolmogorov complexity, that's perfectly valid. It would be a narrower, less generally applicable, version... but a completely valid formulation.
alasdair_ 9y ago
True. Occam's Razor was an attempt at proving God exists. Hardly the most common usage now, however.
- abritinthebay 9y ago
  That's... a bit of an over simplification. It was a common statement of a generally justifiable philosophical tool that happened to be commonly used by theologians like Occam (who wasn't that first to use the idea - Aristotle stated something similar). The intertwining of philosophy and theology makes it related but I'd say your characterization was simplistic and quite incorrect overall.
CurtMonash 9y ago
The actual Occam's Razor is useless unless adapted. How does one count "assumptions" -- if we take Wikipedia's word for what it states -- or "entities" -- if we take the version I first learned? Most versions that actually work can be lightheartedly equated to a "Principle of Least Foofarah".
- abritinthebay 9y ago
  Sure, you do you and academics will continue to use formulations that have worked for about 2500 years or so while you claim it's useless. Yeesh.

GavinMcG 9y ago

Why does one need it to be aesthetically tinged? Another approach is virtue epistemology, where we count as true that which trusted sources and processes say is true. That is, the drug dog who has been right in the past is one we listen to in the future. The scientific project that makes predictions that turn out to be accurate is the one we listen to.

erikpukinskis 9y ago
> That is, the drug dog who has been right in the past is one we listen to in the future If we're in a situation where we get good data about what's right, we can just do science and it doesn't really matter what authorities or dogs think. The place where you need non-scientific epistemologies is when you don't have good feedback data. And sometimes even when there is data, bad predictors can mutually self reinforce. Racist cops arrest more black kids, more black kids get convicted, that proves racist cops are good predictors of criminality. It's not true, but the data says it's true. It's equivalent to a scientist doing many studies and only publishing the ones that are positive. The stats only work for independent measures but nothing is independent especially when you are using predictions to make policy.
- GavinMcG 9y ago
  Those are fair criticisms, but they can be addressed without demanding aesthetic judgments as a component of epistemology.
anigbrowl 9y ago
Scientism is good for making better predictions, but it won't tell you how to live. You could be a utilitarian and declare that you will live to the end of maximizing your utility (however you define that) but the problem with utilitarianism is that omniscience is elusive.
- GavinMcG 9y ago
  That means that there are certain things which aren't answerable, at least not easily or precisely. But that isn't a problem with the epistemology you adopt: some things are going to be hard or impossible to know no matter how you define knowing. Or, "how to live" isn't an epistemological question. So demanding that an epistemology accept aesthetic standards to accommodate concerns about how to live is confused.