Could you taboo/define ‘philosophy,’ ‘math,’ and ‘science’ for me in a way that clarifies exactly how they don’t overlap? It’d be very helpful. Is there any principled reason, for example, that theoretical physics cannot be philosophy? Or is some theoretical physics philosophy, and some not? Is there a sharp line, or a continuum between the two kinds of theoretical physics?
if that’s a philosophy department’s purpose then it doesn’t need to be funded beyond that.
If that’s a philosophy department’s purpose, and nothing else can fulfill the same purpose, then philosophy departments are vastly underfunded as it stands. (Though I agree the current funding could be better managed.)
But the real flaw is that we think of philosophy as a college thing. Philosophical training should be fully integrated into quite early-age education in logical, scientific, mathematical, moral, and other forms of reasoning.
I didn’t say they don’t overlap. I said the distinctions have become less blurred (I think because of the need for increased specialization in all intellectual endeavours as we accumulate more knowledge). I define philosophy, math, and science by their professions. That is, their university departments, their journals, their majors, their textbooks, and so on.
Hence, I think the best way to ask if “philosophy” is a worthwhile endeavour is to asked “why should we fund philosophy departments?” A better way to ask that question is “why should we fund philosophy research and professional philosophers (as opposed to teachers of basic philosophy)?”
And though while I think basic philosophy can be helpful in getting a footing in critical thinking, I also think CFAR is considerably better at teaching critical thinking.
I don’t see any principled reason for why we can’t all be generalists without labels. Practical reasons, yes.
I thought you were saying that the distinctions have become less blurred? Now I’m confused.
I define philosophy, math, and science by their professions.
That’s fine for some everyday purposes. But if we want to distinguish the useful behaviors in each profession from the useless ones, and promote the best behaviors both among laypeople and among professionals, we need more fine-grained categories than just ‘everything that people who publish in journals seen as philosophy journals do.’ I think it would be useful to distinguish Professional Philosophy, Professional Science, and Professional Mathematics from the basic human practices of philosophizing, doing science, or reasoning mathematically. Something in the neighborhood of these ideas would be quite useful:
mathematics: carefully and systematically reasoning about quantity, or (more loosely) about the quantitative properties and relationships of things.
philosophy: carefully reasoning about generalizations, via ‘internal’ reflection (phenomenology, thought experiments, conceptual analysis, etc.), in a moderately (more than shamanic storytelling, less than math or logic) systematic way.
science: carefully collecting empirical data, and carefully reasoning about its predictive and transparently ontological significance.
Do you think these would be useful fast-and-ready definitions for everyday promotion of scientific, philosophical, and mathematical literacy? Would you modify any of them?
I thought you were saying that the distinctions have become less blurred?
Yup, my bad. You caught me before my edit.
Do you think these would be useful fast-and-ready definitions for everyday promotion of scientific, philosophical, and mathematical literacy? Would you modify any of them?
I think you’re reifying abstraction and doing so will introduce pitfalls when discussing them. Math, science, and philosophy are the abstracted output of their respective professions. If you take away science’s competitive incentive structure or change its mechanism of output (journal articles) then you’re modifying science. If you install a self-improving recursive feedback cycle with reality in philosophy, then I think you’ve recreated math and science within philosophy (because science is fundamentally concrete reasoning while math is abstract reasoning and philosophy carries both).
If I’m going to promote something to laypeople, it’s that a mechanism of recursive self-improvement is desirable. There’s plenty to unpack there, though. Like you need a measure of improvement that contacts reality.
I think you’re reifying abstraction and doing so will introduce pitfalls when discussing them.
I think your definitions are more abstract than mine. For me, mathematics, philosophy, and science are embodied brain behaviors — modes of reasoning. For you, if I’m understanding you right, they’re professions, institutions, social groups, population-wide behaviors. Sociology is generally considered more abstract or high-level than psychology.
(Of course, I don’t reject your definitions on that account; denying the existence of philosophizing or of professional philosophy because one or the other is ‘abstract’ would be as silly as denying the existence of abstractions like debt, difficulty, truth, or natural selection. I just think your abstraction is of somewhat more limited utility than mine, when our goal is to spread good philosophizing, science, and mathematics rather than to treat the good qualities of those disciplines as the special property of a prestigious intellectual elite belonging to a specific network of organizations.)
Feedback cycles are great, but we don’t need to build them into our definition of ‘science’ in order to praise science for happening to possess them; if we put each scientist on a separate island, their work might suffer as a result, but it’s not clear to me that they would lose all ability to do anything scientific, or that we should fail to clearly distinguish the scientifically-minded desert-islander for his unusual behaviors.
Also, it’s not clear in what sense mathematics has a self-improving recursive feedback cycle with reality. Actually, mathematics and philosophy seem to function very analogously in terms of their relationship to reality and to science.
If I’m going to promote something to laypeople, it’s that a mechanism of recursive self-improvement is desirable.
I’m not sure that’s the best approach. Telling people to find a recursively self-improving method is not likely to be as effective as giving them concrete reasoning skills (like how to perform thought experiments, or how to devise empirical hypotheses, or how to multiply quantities) and then letting intelligent society-wide behaviors emerge via the marketplace of ideas (or via top-down societal structuring, if necessary). Don’t fixate first and foremost on telling people about what our abstract models suggest makes science on a societal scale so effective; fixate first and foremost on making them good scientists in their daily lives, in every concrete action.
For you, if I’m understanding you right, they’re professions, institutions, social groups, population-wide behaviors. Sociology is generally considered more abstract or high-level than psychology.
You’re kind of understanding me. Abstractly, bee hives produce honey. Concretely, this bee hive in front of me is producing honey. Abstractly, science is the product of professions, institutions, ect. Concretely, science is the product of people on our planet doing stuff.
I’m literally trying to not talk about abstractions or concepts but science as it actually is. And of course, science as it actually is does things that we can then categorize into abstractions like feedback cycles. But when you say science is a bunch of abstractions (like I think your definitions are), then you’re missing out on what it actually is.
Feedback cycles are great, but we don’t need to build them into our definition of ‘science’ in order to praise science for happening to possess them; if we put each scientist on a separate island, their work might suffer as a result, but it’s not clear to me that they would lose all ability to do anything scientific, or that we should fail to clearly distinguish the scientifically-minded desert-islander for his unusual behaviors.
This is exactly why I want to avoid defining science with abstractions. It literally does not make sense if you think of science as it is. “Scientific” imports essentialism.
Also, it’s not clear in what sense mathematics has a self-improving recursive feedback cycle with reality.
Mathematics is self-improving while at the same time hinging on reality. This is tricky to explain so I might come back to it tomorrow when I’m more well rested (i.e., not drunk).
I’m not sure that’s the best approach. Telling people to find a recursively self-improving method is not likely to be as effective as giving them concrete reasoning skills (like how to perform thought experiments, or how to devise empirical hypotheses, or how to multiply quantities) and then letting intelligent society-wide behaviors emerge via the marketplace of ideas (or via top-down societal structuring, if necessary). Don’t fixate first and foremost on telling people about what our abstract models suggest makes science on a societal scale so effective; fixate first and foremost on making them good scientists in their daily lives, in every concrete action.
No, I think that kernel (and we are speaking in the context of “fast-and-ready”) of thought is really the most important thing to convey. Speaking abstractly, even science doesn’t take that kernel seriously enough. It doesn’t question how it should allocate its limited resources or improve its function. This is costing millions of lives, untold suffering, and perhaps our species continued existence. But it does employ a self-improving feedback cycle on reality which is just enough for it to uncover reality. It needs to install a self-improving feedback cycle on itself. And then we need a self-improving feedback cycle on feedback cycles. I can’t think of any abstraction more important in making progress with something.
Abstractly, bee hives produce honey. Concretely, this bee hive in front of me is producing honey. Abstractly, science is the product of professions, institutions, ect. Concretely, science is the product of people on our planet doing stuff.
It sounds like you’re conflating abstract/concrete with general/particular. But a universal generalization might just be the conjunction of a lot of particulars. I prefer to think of ‘abstract’ as ‘not spatially extended or localized.’ Societies are generally considered more abstract than mental states because mental states are intuitively treated as more localized. But ‘lots of mental states’ is not more abstract than ‘just one mental state,’ in the same way that thousands of bees (or ‘all the bees,’ in your example) can be just as concrete as a single bee.
But when you say science is a bunch of abstractions (like I think your definitions are)
We’re back at square one. I still don’t see why reasoning is more abstract than professions, institutions, etc. We agree that it all reduces to human behaviors on some level. But the ‘abstract vs. concrete’ discussion is a complete tangent. What’s relevant is whether it’s useful to have separate concepts of ‘the practice of science’ vs. ‘professional science,’ the former being something even laypeople can participate in by adopting certain methodological standards. I think both concepts are useful. You seem to think that only ‘professional science’ is a useful concept, at least in most cases. Is that a fair summary?
This is exactly why I want to avoid defining science with abstractions. It literally does not make sense if you think of science as it is. “Scientific” imports essentialism.
Counterfactuals don’t make sense if you think of things as they are? I don’t think that’s true in any nontrivial sense....
‘Scientific’ is not any more guilty of essentializing than are any of our other fuzzy, ordinary-language terms. There are salient properties associated with being a scientist; I’m suggesting that many of those clustered properties, in particular many of the ones we most care about when we promote and praise things like ‘science’ and ‘naturalism,’ can occur in isolated individuals. If you don’t like calling what I’m talking about ‘scientific,’ then coin a different word for it; but we need some word. We need to be able to denote our exemplary decision procedures, just to win the war of ideas.
‘Professional science’ is not an exemplary decision procedure, any more than ‘the buildings and faculty at MIT’ is an exemplary decision procedure. It’s just an especially effective instantiation thereof.
I can’t think of any abstraction more important in making progress with something.
Maybe we’re just not approaching the problem at the same levels. When I ask about what the optimal way is to define our concepts, I’m trying to define them in a way that allows us to consistently and usefully explain them (in any number of paraphrased forms) to 8th-graders, to congressmen, to literary theorists, such that we can promote the best techniques we associate with scientists, philosophers, and mathematicians. I’m imagining how we would design a scientific+philosophical+mathematical+etc. literacy pamphlet that would teach people how to win at life. It sounds like you’re instead trying to think of a single sentence that summarizes what winning at life is, at its most abstract. ‘Adopt a self-improving feedback cycle linking you to reality’ is just a fancy way of saying ‘Behave in a way that predictably makes you better and better at doing good stuff.’ Which is great, but not especially contentful as yet. I only care about people understanding how winning works insofar as this understanding helps them actually win.
I prefer to think of ‘abstract’ as ‘not spatially extended or localized.’
I prefer to think of it as anything existing at least partly in mind, and then we can say we have an abstraction of an abstraction or that something something is more abstract (something from category theory being a pure abstraction, while something like the category “dog” being less abstract because it connects with a pattern of atoms in reality). By their nature, abstractions are also universals, but things that actually exist like the bee hive in front of me aren’t particulars at the concrete level. The specific bee hive in my mind that I’m imagining is a particular, or the “bee hive” that I’m seeing and interpreting into a bee hive in front of me is also a particular, but the bee hive is just a “pattern” of atoms.
What’s relevant is whether it’s useful to have separate concepts of ‘the practice of science’ vs. ‘professional science,’ the former being something even laypeople can participate in by adopting certain methodological standards. I think both concepts are useful. You seem to think that only ‘professional science’ is a useful concept, at least in most cases. Is that a fair summary?
Framing those concepts in terms of usefulness isn’t helpful, I think. I’d simply say the laypeople are doing something different unless they’re contributing to our body of knowledge. In which case, science as it is requires that those laypeople interact with science as it is (journals and such).
Counterfactuals don’t make sense if you think of things as they are?
No, I mean thinking of someone as being scientific doesn’t make sense if you think of science as it is because e.g. the sixth grader at the science fair that we all “scientific” isn’t interacting with science as it is. We’re taking some essential properties we pattern match in science as it is, and then we abstract them, and then we apply them by pattern matching.
I’m suggesting that many of those clustered properties, in particular many of the ones we most care about when we promote and praise things like ‘science’ and ‘naturalism,’ can occur in isolated individuals.
We can imagine an immortal human being on another planet replicating everything science has done on Earth thus far. So, yes I think it can occur in isolated individuals, but that’s only because the individual has taken on everything that science is and not some like “carefully collecting empirical data, and carefully reasoning about its predictive and transparently ontological significance.”
If I’m going to apply an abstraction to what I praise in science to individuals, it’s not “being scientific” or “doing science”, it’s “working with feedback.” It’s what programmers do, it’s what engineers do, it’s what mathematicians, it’s what scientists do, it’s what people that effectively lose weight do, and so on. It’s the kernel of thought most conducive to progress in any area.
Maybe we’re just not approaching the problem at the same levels. When I ask about what the optimal way is to define our concepts, I’m trying to define them in a way that allows us to consistently ..
I think we are approaching the problem at the same level. I think I have optimally defined the concepts, and I think “behave in a way that predictably makes you better and better at doing good stuff” is what needs to be communicated and not “science: carefully collecting empirical data, and carefully reasoning about its predictive and transparently ontological significance.” If we’re going to add more content, then we should talk about how to effectively measure self-improvement, how to get solid feedback and so on. With that knowledge, I think a bunch of kids working together could rebuild science from the ground up.
If, in some cataclysm, all of scientific knowledge were to be destroyed, and only one sentence passed on to the next generation of creatures, what statement would contain the most information in the fewest words? I believe it is the atomic hypothesis that all things are made of atoms — little particles that move around in perpetual motion, attracting each other when they are a little distance… -- Feynman
I’d pass on how important “behave in a way that predictably makes you better and better at doing good stuff” is.
I prefer to think of it as anything existing at least partly in mind
That’s problematic, first, because it leaves mind itself in a strange position. And second because, if mathematical platonism (for example) were true, then there would exist abstract objects that are mind-independent.
We’re taking some essential properties we pattern match in science as it is, and then we abstract them, and then we apply them by pattern matching.
You seem to be assuming the pattern-matching of this sort is a vice. If it’s useful to mark the pattern in question, and we recognize that we’re doing so for utilitarian reasons and not because there’s a transcendent Essence of Scienceyness, then the pattern-matching is benign. It’s how humans think, and we can’t become completely inhuman if our goal is to take the rest of mankind with us into the future. Not yet, anyway.
Religions are also feedback loops. The more I believe, the more my belief gets confirmed. Remarkable! The primary problem with this ultra-attenuated notion of what we want is that all the work is being done by the black-box normative terms like ‘improvement’ and ‘better’ and ‘optimal.’ Everything we’re actually trying to concretely teach is hidden behind those words.
We also need more content than ‘working with a feedback loop from reality’; that kind of metaphorical talk might fly on LessWrong, but it’s really a summary of some implicit intuitions we already share, not instruction we could in those words convey to someone who doesn’t already see what we’re getting at. After all, everything exists in a back-and-forth with reality, and everything is for that matter part of reality. Perhaps my formulations of what we want are too concrete; but yours are certainly too abstract and underdetermined.
Could you taboo/define ‘philosophy,’ ‘math,’ and ‘science’ for me in a way that clarifies exactly how they don’t overlap? It’d be very helpful. Is there any principled reason, for example, that theoretical physics cannot be philosophy? Or is some theoretical physics philosophy, and some not? Is there a sharp line, or a continuum between the two kinds of theoretical physics?
If that’s a philosophy department’s purpose, and nothing else can fulfill the same purpose, then philosophy departments are vastly underfunded as it stands. (Though I agree the current funding could be better managed.)
But the real flaw is that we think of philosophy as a college thing. Philosophical training should be fully integrated into quite early-age education in logical, scientific, mathematical, moral, and other forms of reasoning.
I didn’t say they don’t overlap. I said the distinctions have become less blurred (I think because of the need for increased specialization in all intellectual endeavours as we accumulate more knowledge). I define philosophy, math, and science by their professions. That is, their university departments, their journals, their majors, their textbooks, and so on.
Hence, I think the best way to ask if “philosophy” is a worthwhile endeavour is to asked “why should we fund philosophy departments?” A better way to ask that question is “why should we fund philosophy research and professional philosophers (as opposed to teachers of basic philosophy)?”
And though while I think basic philosophy can be helpful in getting a footing in critical thinking, I also think CFAR is considerably better at teaching critical thinking.
I don’t see any principled reason for why we can’t all be generalists without labels. Practical reasons, yes.
I thought you were saying that the distinctions have become less blurred? Now I’m confused.
That’s fine for some everyday purposes. But if we want to distinguish the useful behaviors in each profession from the useless ones, and promote the best behaviors both among laypeople and among professionals, we need more fine-grained categories than just ‘everything that people who publish in journals seen as philosophy journals do.’ I think it would be useful to distinguish Professional Philosophy, Professional Science, and Professional Mathematics from the basic human practices of philosophizing, doing science, or reasoning mathematically. Something in the neighborhood of these ideas would be quite useful:
mathematics: carefully and systematically reasoning about quantity, or (more loosely) about the quantitative properties and relationships of things.
philosophy: carefully reasoning about generalizations, via ‘internal’ reflection (phenomenology, thought experiments, conceptual analysis, etc.), in a moderately (more than shamanic storytelling, less than math or logic) systematic way.
science: carefully collecting empirical data, and carefully reasoning about its predictive and transparently ontological significance.
Do you think these would be useful fast-and-ready definitions for everyday promotion of scientific, philosophical, and mathematical literacy? Would you modify any of them?
Yup, my bad. You caught me before my edit.
I think you’re reifying abstraction and doing so will introduce pitfalls when discussing them. Math, science, and philosophy are the abstracted output of their respective professions. If you take away science’s competitive incentive structure or change its mechanism of output (journal articles) then you’re modifying science. If you install a self-improving recursive feedback cycle with reality in philosophy, then I think you’ve recreated math and science within philosophy (because science is fundamentally concrete reasoning while math is abstract reasoning and philosophy carries both).
If I’m going to promote something to laypeople, it’s that a mechanism of recursive self-improvement is desirable. There’s plenty to unpack there, though. Like you need a measure of improvement that contacts reality.
I think your definitions are more abstract than mine. For me, mathematics, philosophy, and science are embodied brain behaviors — modes of reasoning. For you, if I’m understanding you right, they’re professions, institutions, social groups, population-wide behaviors. Sociology is generally considered more abstract or high-level than psychology.
(Of course, I don’t reject your definitions on that account; denying the existence of philosophizing or of professional philosophy because one or the other is ‘abstract’ would be as silly as denying the existence of abstractions like debt, difficulty, truth, or natural selection. I just think your abstraction is of somewhat more limited utility than mine, when our goal is to spread good philosophizing, science, and mathematics rather than to treat the good qualities of those disciplines as the special property of a prestigious intellectual elite belonging to a specific network of organizations.)
Feedback cycles are great, but we don’t need to build them into our definition of ‘science’ in order to praise science for happening to possess them; if we put each scientist on a separate island, their work might suffer as a result, but it’s not clear to me that they would lose all ability to do anything scientific, or that we should fail to clearly distinguish the scientifically-minded desert-islander for his unusual behaviors.
Also, it’s not clear in what sense mathematics has a self-improving recursive feedback cycle with reality. Actually, mathematics and philosophy seem to function very analogously in terms of their relationship to reality and to science.
I’m not sure that’s the best approach. Telling people to find a recursively self-improving method is not likely to be as effective as giving them concrete reasoning skills (like how to perform thought experiments, or how to devise empirical hypotheses, or how to multiply quantities) and then letting intelligent society-wide behaviors emerge via the marketplace of ideas (or via top-down societal structuring, if necessary). Don’t fixate first and foremost on telling people about what our abstract models suggest makes science on a societal scale so effective; fixate first and foremost on making them good scientists in their daily lives, in every concrete action.
You’re kind of understanding me. Abstractly, bee hives produce honey. Concretely, this bee hive in front of me is producing honey. Abstractly, science is the product of professions, institutions, ect. Concretely, science is the product of people on our planet doing stuff.
I’m literally trying to not talk about abstractions or concepts but science as it actually is. And of course, science as it actually is does things that we can then categorize into abstractions like feedback cycles. But when you say science is a bunch of abstractions (like I think your definitions are), then you’re missing out on what it actually is.
This is exactly why I want to avoid defining science with abstractions. It literally does not make sense if you think of science as it is. “Scientific” imports essentialism.
Mathematics is self-improving while at the same time hinging on reality. This is tricky to explain so I might come back to it tomorrow when I’m more well rested (i.e., not drunk).
No, I think that kernel (and we are speaking in the context of “fast-and-ready”) of thought is really the most important thing to convey. Speaking abstractly, even science doesn’t take that kernel seriously enough. It doesn’t question how it should allocate its limited resources or improve its function. This is costing millions of lives, untold suffering, and perhaps our species continued existence. But it does employ a self-improving feedback cycle on reality which is just enough for it to uncover reality. It needs to install a self-improving feedback cycle on itself. And then we need a self-improving feedback cycle on feedback cycles. I can’t think of any abstraction more important in making progress with something.
It sounds like you’re conflating abstract/concrete with general/particular. But a universal generalization might just be the conjunction of a lot of particulars. I prefer to think of ‘abstract’ as ‘not spatially extended or localized.’ Societies are generally considered more abstract than mental states because mental states are intuitively treated as more localized. But ‘lots of mental states’ is not more abstract than ‘just one mental state,’ in the same way that thousands of bees (or ‘all the bees,’ in your example) can be just as concrete as a single bee.
We’re back at square one. I still don’t see why reasoning is more abstract than professions, institutions, etc. We agree that it all reduces to human behaviors on some level. But the ‘abstract vs. concrete’ discussion is a complete tangent. What’s relevant is whether it’s useful to have separate concepts of ‘the practice of science’ vs. ‘professional science,’ the former being something even laypeople can participate in by adopting certain methodological standards. I think both concepts are useful. You seem to think that only ‘professional science’ is a useful concept, at least in most cases. Is that a fair summary?
Counterfactuals don’t make sense if you think of things as they are? I don’t think that’s true in any nontrivial sense....
‘Scientific’ is not any more guilty of essentializing than are any of our other fuzzy, ordinary-language terms. There are salient properties associated with being a scientist; I’m suggesting that many of those clustered properties, in particular many of the ones we most care about when we promote and praise things like ‘science’ and ‘naturalism,’ can occur in isolated individuals. If you don’t like calling what I’m talking about ‘scientific,’ then coin a different word for it; but we need some word. We need to be able to denote our exemplary decision procedures, just to win the war of ideas.
‘Professional science’ is not an exemplary decision procedure, any more than ‘the buildings and faculty at MIT’ is an exemplary decision procedure. It’s just an especially effective instantiation thereof.
Maybe we’re just not approaching the problem at the same levels. When I ask about what the optimal way is to define our concepts, I’m trying to define them in a way that allows us to consistently and usefully explain them (in any number of paraphrased forms) to 8th-graders, to congressmen, to literary theorists, such that we can promote the best techniques we associate with scientists, philosophers, and mathematicians. I’m imagining how we would design a scientific+philosophical+mathematical+etc. literacy pamphlet that would teach people how to win at life. It sounds like you’re instead trying to think of a single sentence that summarizes what winning at life is, at its most abstract. ‘Adopt a self-improving feedback cycle linking you to reality’ is just a fancy way of saying ‘Behave in a way that predictably makes you better and better at doing good stuff.’ Which is great, but not especially contentful as yet. I only care about people understanding how winning works insofar as this understanding helps them actually win.
I prefer to think of it as anything existing at least partly in mind, and then we can say we have an abstraction of an abstraction or that something something is more abstract (something from category theory being a pure abstraction, while something like the category “dog” being less abstract because it connects with a pattern of atoms in reality). By their nature, abstractions are also universals, but things that actually exist like the bee hive in front of me aren’t particulars at the concrete level. The specific bee hive in my mind that I’m imagining is a particular, or the “bee hive” that I’m seeing and interpreting into a bee hive in front of me is also a particular, but the bee hive is just a “pattern” of atoms.
I think that you’re stuck in noun-land while I’m in verb-land, but I don’t think noun-land is concrete (it’s an abstraction).
Framing those concepts in terms of usefulness isn’t helpful, I think. I’d simply say the laypeople are doing something different unless they’re contributing to our body of knowledge. In which case, science as it is requires that those laypeople interact with science as it is (journals and such).
No, I mean thinking of someone as being scientific doesn’t make sense if you think of science as it is because e.g. the sixth grader at the science fair that we all “scientific” isn’t interacting with science as it is. We’re taking some essential properties we pattern match in science as it is, and then we abstract them, and then we apply them by pattern matching.
We can imagine an immortal human being on another planet replicating everything science has done on Earth thus far. So, yes I think it can occur in isolated individuals, but that’s only because the individual has taken on everything that science is and not some like “carefully collecting empirical data, and carefully reasoning about its predictive and transparently ontological significance.”
If I’m going to apply an abstraction to what I praise in science to individuals, it’s not “being scientific” or “doing science”, it’s “working with feedback.” It’s what programmers do, it’s what engineers do, it’s what mathematicians, it’s what scientists do, it’s what people that effectively lose weight do, and so on. It’s the kernel of thought most conducive to progress in any area.
I think we are approaching the problem at the same level. I think I have optimally defined the concepts, and I think “behave in a way that predictably makes you better and better at doing good stuff” is what needs to be communicated and not “science: carefully collecting empirical data, and carefully reasoning about its predictive and transparently ontological significance.” If we’re going to add more content, then we should talk about how to effectively measure self-improvement, how to get solid feedback and so on. With that knowledge, I think a bunch of kids working together could rebuild science from the ground up.
I’d pass on how important “behave in a way that predictably makes you better and better at doing good stuff” is.
That’s problematic, first, because it leaves mind itself in a strange position. And second because, if mathematical platonism (for example) were true, then there would exist abstract objects that are mind-independent.
You seem to be assuming the pattern-matching of this sort is a vice. If it’s useful to mark the pattern in question, and we recognize that we’re doing so for utilitarian reasons and not because there’s a transcendent Essence of Scienceyness, then the pattern-matching is benign. It’s how humans think, and we can’t become completely inhuman if our goal is to take the rest of mankind with us into the future. Not yet, anyway.
Religions are also feedback loops. The more I believe, the more my belief gets confirmed. Remarkable! The primary problem with this ultra-attenuated notion of what we want is that all the work is being done by the black-box normative terms like ‘improvement’ and ‘better’ and ‘optimal.’ Everything we’re actually trying to concretely teach is hidden behind those words.
We also need more content than ‘working with a feedback loop from reality’; that kind of metaphorical talk might fly on LessWrong, but it’s really a summary of some implicit intuitions we already share, not instruction we could in those words convey to someone who doesn’t already see what we’re getting at. After all, everything exists in a back-and-forth with reality, and everything is for that matter part of reality. Perhaps my formulations of what we want are too concrete; but yours are certainly too abstract and underdetermined.