Leaving aside any putative True Theory of Everything which we don’t know yet, the laws we actually know and use today are definitely Humean. We should know, we made them that way.
A true theory of everything is by definition never wrong. In which case there’s no observable difference between Humeanism and non-Humeanism, and it makes no sense to talk about the theory “determining” events or merely “describing” them.
Define: theory of everything: maximally compressed, true and complete description of the physical evolution of the universe over time.
This was a point of some confusion to me. But “The laws of nature” to me means the fundamental laws of the universe, not the models we come up with. I dismissed my confusion as “oh, this must be another obnoxious thing that mainstream philosophy thinks.”
This wording suggests that a major distinction between Humeans and non-Humeans is that the former consider laws of nature ‘descriptions’ (i.e., linguistic/conceptual entities) while the latter consider laws of nature mind-independent. But we could frame the distinction either linguistically or metaphysically. Metaphysically, Humeans think that Nature’s patterns are ultimately just coincidental recurrences, while non-Humeans think that observed patterns have some cause or explanation in a deeper unifying structure.
I suspect that a lot of the support for Humeanism in the above poll can be explained by people pattern-matching ‘laws of nature are… descriptions’ to ‘positivism / metaphysical humility,’ whereas in reality Humeanism is just as substantive and speculative a metaphysical thesis as is non-Humeanism.
As an aside, it should be noted that Hume himself was not a ‘Humean’ in the above sense. So the modern terminology is historically misleading here.
Metaphysically, Humeans think that Nature’s patterns are ultimately just coincidental recurrences, while non-Humeans think that observed patterns have some cause or explanation in a deeper unifying structure.
I think contemporary Humeans would disagree with your characterization. They distinguish between accidentally true generalizations (like the fact that no human is over 10 feet tall) and law-like generalizations. Not all patterns count as laws, hence the use of “salient” in my definition. So a law is not merely a coincidental recurrence. Many Humeans believe that any particular instance of a salient pattern of recurrences is explained by the existence and salience of the pattern, so laws are explanatory.
Also, a number of Humeans do cite metaphysical humility as an advantage of their account of laws. Unlike non-Humeans, who need an additional fundamental metaphysical kind besides spatio-temporally distributed properties (i.e. laws of nature), Humeans purport to make do with just the properties themselves, reducing the laws to the properties. If this works, then it does seem like a more metaphysically parsimonious account, at least according to some accounts of parsimony. Of course, the Humean account is still a metaphysical account, so it wouldn’t be congenial to the strict positivist, I suppose.
Out of curiosity, do you have on hand any good articles on how Humeans (metaphysically, epistemically, etc.) distinguish ‘mere patterns’ from ‘capital-p Patterns’? (Aside from anthropocentric concepts like explanatory reducibility.)
Many Humeans believe that any particular instance of a salient pattern of recurrences is explained by the existence and salience of the pattern, so laws are explanatory.
‘Salience’ is usually an anthropocentric / psychological concept. What is meant by it here? I ask because we have to be careful not to allow non-Humean ‘salience-making’ properties. If salience is something metaphysical that makes certain patterns objectively ‘special’ and ‘lawful’ and ‘regular’, then I’d normally think of salience as non-Humean. On the other hand, if ‘salience’ is a non-unifying brute fact (‘brute correlation,’ ‘unexplainable recurrence’, ‘irreducible coincidence’, etc.), or is somehow relative to our human projects and interests and standards, then I’d normally think of it as Humean. Does this jibe with your experience?
Also, a number of Humeans do cite metaphysical humility as an advantage of their account of laws.
Yes. And ‘metaphysical humility’ here means ‘metaphysical commitment to a smaller domain of entities or kinds’. My point was just to hammer home that it doesn’t mean ‘avoiding doing difficult speculative metaphysics’ or ‘skeptically withholding judgment’ or ‘treating lawfulness as a human construct’ or ‘positivistically denying that non-Humean laws are intelligible’; the latter four options preclude Humeanism as much as they preclude anti-Humeanism, and my worry was that some positivism-friendly LessWrongers would misconstrue Humeanism as somehow unmetaphysical, rather than just as metaphysically sparse.
(Though, again, this is complicated; non-Humeans claim that Humeanism is metaphysically profligate, say, because brute correlations that continue to recur, and recur, and recur are more startling than unitarily ‘law-governed’ recurrences. Cf. Eliezer’s concerns with treating our universe as metaphysically first-order.)
A side-note, since it hasn’t come up: A lot of people in the Humeanism debate unpack the dispute modally. On this conception, Humeans consider the laws and regularities of nature contingent, while non-Humeans believe that there is some sort of metaphysical necessity behind our world’s regularities. I don’t think this is a very enlightening way to frame the debate, and in most contexts I’d be happy to grant that you could be a necessitarian Humean; but it’s an extremely common form of the debate, so it should be kept in mind.
Out of curiosity, do you have on hand any good articles on how Humeans (metaphysically, epistemically, etc.) distinguish ‘mere patterns’ from ‘capital-p Patterns’? (Aside from anthropocentric concepts like explanatory reducibility.)
The classic neo-Humean account is David Lewis’s. It’s described in the SEP here. I also wrote a post about it here. The account does appeal to the simplicity of descriptions, which is arguably a somewhat anthropocentric concept, but I see that as a feature, not a bug. The laws are ways of efficiently organizing our understanding of the universe, not fundamental mind-independent entities (see the last section of my post). I don’t think Lewis would agree with my characterization in this regard, I should say.
Humeanism: The laws of nature are compressed descriptions of salient patterns in the distribution of physical events.
Non-Humeanism: The laws of nature are not mere descriptions. They determine the distribution of physical events.
Leaving aside any putative True Theory of Everything which we don’t know yet, the laws we actually know and use today are definitely Humean. We should know, we made them that way.
I assumed the question was referring to the fundamental laws of the universe, which would be a theory of everything.
A true theory of everything is by definition never wrong. In which case there’s no observable difference between Humeanism and non-Humeanism, and it makes no sense to talk about the theory “determining” events or merely “describing” them.
Define: theory of everything: maximally compressed, true and complete description of the physical evolution of the universe over time.
This was a point of some confusion to me. But “The laws of nature” to me means the fundamental laws of the universe, not the models we come up with. I dismissed my confusion as “oh, this must be another obnoxious thing that mainstream philosophy thinks.”
And “determine” too.
This wording suggests that a major distinction between Humeans and non-Humeans is that the former consider laws of nature ‘descriptions’ (i.e., linguistic/conceptual entities) while the latter consider laws of nature mind-independent. But we could frame the distinction either linguistically or metaphysically. Metaphysically, Humeans think that Nature’s patterns are ultimately just coincidental recurrences, while non-Humeans think that observed patterns have some cause or explanation in a deeper unifying structure.
I suspect that a lot of the support for Humeanism in the above poll can be explained by people pattern-matching ‘laws of nature are… descriptions’ to ‘positivism / metaphysical humility,’ whereas in reality Humeanism is just as substantive and speculative a metaphysical thesis as is non-Humeanism.
As an aside, it should be noted that Hume himself was not a ‘Humean’ in the above sense. So the modern terminology is historically misleading here.
I think contemporary Humeans would disagree with your characterization. They distinguish between accidentally true generalizations (like the fact that no human is over 10 feet tall) and law-like generalizations. Not all patterns count as laws, hence the use of “salient” in my definition. So a law is not merely a coincidental recurrence. Many Humeans believe that any particular instance of a salient pattern of recurrences is explained by the existence and salience of the pattern, so laws are explanatory.
Also, a number of Humeans do cite metaphysical humility as an advantage of their account of laws. Unlike non-Humeans, who need an additional fundamental metaphysical kind besides spatio-temporally distributed properties (i.e. laws of nature), Humeans purport to make do with just the properties themselves, reducing the laws to the properties. If this works, then it does seem like a more metaphysically parsimonious account, at least according to some accounts of parsimony. Of course, the Humean account is still a metaphysical account, so it wouldn’t be congenial to the strict positivist, I suppose.
Out of curiosity, do you have on hand any good articles on how Humeans (metaphysically, epistemically, etc.) distinguish ‘mere patterns’ from ‘capital-p Patterns’? (Aside from anthropocentric concepts like explanatory reducibility.)
‘Salience’ is usually an anthropocentric / psychological concept. What is meant by it here? I ask because we have to be careful not to allow non-Humean ‘salience-making’ properties. If salience is something metaphysical that makes certain patterns objectively ‘special’ and ‘lawful’ and ‘regular’, then I’d normally think of salience as non-Humean. On the other hand, if ‘salience’ is a non-unifying brute fact (‘brute correlation,’ ‘unexplainable recurrence’, ‘irreducible coincidence’, etc.), or is somehow relative to our human projects and interests and standards, then I’d normally think of it as Humean. Does this jibe with your experience?
Yes. And ‘metaphysical humility’ here means ‘metaphysical commitment to a smaller domain of entities or kinds’. My point was just to hammer home that it doesn’t mean ‘avoiding doing difficult speculative metaphysics’ or ‘skeptically withholding judgment’ or ‘treating lawfulness as a human construct’ or ‘positivistically denying that non-Humean laws are intelligible’; the latter four options preclude Humeanism as much as they preclude anti-Humeanism, and my worry was that some positivism-friendly LessWrongers would misconstrue Humeanism as somehow unmetaphysical, rather than just as metaphysically sparse.
(Though, again, this is complicated; non-Humeans claim that Humeanism is metaphysically profligate, say, because brute correlations that continue to recur, and recur, and recur are more startling than unitarily ‘law-governed’ recurrences. Cf. Eliezer’s concerns with treating our universe as metaphysically first-order.)
A side-note, since it hasn’t come up: A lot of people in the Humeanism debate unpack the dispute modally. On this conception, Humeans consider the laws and regularities of nature contingent, while non-Humeans believe that there is some sort of metaphysical necessity behind our world’s regularities. I don’t think this is a very enlightening way to frame the debate, and in most contexts I’d be happy to grant that you could be a necessitarian Humean; but it’s an extremely common form of the debate, so it should be kept in mind.
The classic neo-Humean account is David Lewis’s. It’s described in the SEP here. I also wrote a post about it here. The account does appeal to the simplicity of descriptions, which is arguably a somewhat anthropocentric concept, but I see that as a feature, not a bug. The laws are ways of efficiently organizing our understanding of the universe, not fundamental mind-independent entities (see the last section of my post). I don’t think Lewis would agree with my characterization in this regard, I should say.