This is super interesting - thanks for sharing and writing. I was also wondering what the relationship between an abstraction and an explanation is (as you briefly mentioned). I like how you link the idea of a really good abstraction as being hard to vary.
In The Fabric of Reality, Deutsch starts off by describing better explanations having more "depth" (akin to "hard to vary" in the beginning of infinity, but also alluding to "deeper" explanations as being more fundamental) and more "breadth" (i.e. having "reach" across domains).
Linking this to the way you have described abstractions - I wonder whether using your suggestion of abstractions being decomposed into more fundamental abstractions is related to Deutsch's "depth" of explanations. And in parallel, I suppose that abstractions that apply across multiple domains (i.e. what you describe as "powerful") might relate to their "breadth" of explanation.
Let me know what you think.
Also, would you be able to give an example/clarify what you mean here? Really intriguing point but I'm struggling to figure it out: "Finally I want to point out that we can still have high epistemic confidence in explanation-less knowledge. However I think it is human nature (or at least my human nature!) to seek explanations and not just knowledge."
I've been meaning to go back and reread Deutsch's writing on explanations and think about it all some more.
The last point you asked about. I'm thinking here of the type of knowledge we get from statistics - we know there (from Randomised Controlled Trials say) that is some really strong causal correlation between two things but we really don't know why there is a correlation. Does that make sense?
Ah yes - I understand what you mean now, particularly "human nature" of wanting to seek explanations.
I'm always a little unsatisfied by the studies done in cardiovascular medicine claiming one antiplatelet is better than another, based on the findings of a massive double armed RCT, with no underlying explanation for why that may be the case. Even the classic Framingham studies are no better from an explanatory perspective, despite being quite powerful from a "knowledge" perspective.
That's exactly how I feel about those sort of studies! It's good that people do them, and good can definitely come of them, but that is an unsatisfactoriness to me! I don't think I could do that type of research even though I am glad that other people do.
This is super interesting - thanks for sharing and writing. I was also wondering what the relationship between an abstraction and an explanation is (as you briefly mentioned). I like how you link the idea of a really good abstraction as being hard to vary.
In The Fabric of Reality, Deutsch starts off by describing better explanations having more "depth" (akin to "hard to vary" in the beginning of infinity, but also alluding to "deeper" explanations as being more fundamental) and more "breadth" (i.e. having "reach" across domains).
Linking this to the way you have described abstractions - I wonder whether using your suggestion of abstractions being decomposed into more fundamental abstractions is related to Deutsch's "depth" of explanations. And in parallel, I suppose that abstractions that apply across multiple domains (i.e. what you describe as "powerful") might relate to their "breadth" of explanation.
Let me know what you think.
Also, would you be able to give an example/clarify what you mean here? Really intriguing point but I'm struggling to figure it out: "Finally I want to point out that we can still have high epistemic confidence in explanation-less knowledge. However I think it is human nature (or at least my human nature!) to seek explanations and not just knowledge."
Looking forward to the next part!
Thanks for taking the time to read this!
I've been meaning to go back and reread Deutsch's writing on explanations and think about it all some more.
The last point you asked about. I'm thinking here of the type of knowledge we get from statistics - we know there (from Randomised Controlled Trials say) that is some really strong causal correlation between two things but we really don't know why there is a correlation. Does that make sense?
Ah yes - I understand what you mean now, particularly "human nature" of wanting to seek explanations.
I'm always a little unsatisfied by the studies done in cardiovascular medicine claiming one antiplatelet is better than another, based on the findings of a massive double armed RCT, with no underlying explanation for why that may be the case. Even the classic Framingham studies are no better from an explanatory perspective, despite being quite powerful from a "knowledge" perspective.
That's exactly how I feel about those sort of studies! It's good that people do them, and good can definitely come of them, but that is an unsatisfactoriness to me! I don't think I could do that type of research even though I am glad that other people do.