Yep Lee it’s good to have a project. I guess it was originally incited back in 2016 or so blogging at the sites of Massimo Pigliucci and Dan Kaufman. They were always so worried about the evils of scientism and making sure that philosophers be respected. That’s where I hatched my scheme for a new community with the sole purpose of founding science by means of expert opinion. It’s taken far too long for me to act on it! I suspect that your “Institute for Ontic Realism” has been brewing in you for a long time too.
Reality itself? There’s only one thing that I believe I can know about about the ontological truth of what exists. That would be my own existence itself, or the particular mind that is me. Descartes referred to this as “thought”, though in my post #2 (https://eborg760.substack.com/p/i-value-therefore-i-am) I refined it further as “I value therefore I am”. Ultimately my existence reduces to something that feels good/bad, (which my last post also argued should be a neurally produced electromagnetic field).
Regardless, from the perfectly certain founding truth of me that exists, science is used to empirically model all else. So what I see can only be theoretical. Humanity may have developed strong theory by this point, though as Mike and I were just saying, it’s all provisional. The perfectly certain me might instead exist by means of some sort of magical situation. And that gets to my coming post about why science must be founded upon the premise that systemic causality never fails — science is rendered obsolete to the degree that causality does fail. So science needs to be reformed with professionally agreed upon principles of metaphysics, epistemology, and axiology. And I do not merely mean to become the founder of this proposed new “metascience”. The specific principles that I propose should also go on to found future science in general.
Hey, we’re bloggers man! If we enjoy what we’re doing then it’s worthwhile for that reason alone. If we don’t then our free time goes to gardening, TV, games, or whatever one enjoys more. And if blogging is only fun because we know we’ll eventually be validated and actually won’t, well yes, that could be precarious. But fun’s still fun!
I’m in the midst of reading and outling your lengthy argument. Please forgive me. As you know I like to take my time. However, as I read, I feel the need to share a thought. This is really not germane in any way to your argument. But it’s on your use of words. “Metaphysics” is a term philosophy inherited from Aristotle. However, it was not Aristotle’s term. He never used the word himself. Aristotle’s term for this part of his philosophy was “first philosophy” or just “wisdom.”
When later editors began to compile Aristotle’s works they placed a collection of Aristotle’s works after his work on physics. The word meta in Greek means after. So, as a working title they simply called those works metaphysics—meta ta physika. The title stuck. However, that inadvertent title had a tendency to influence later philosophers as to Aristotle’s deeper meaning. However, the metaphysics title works fairly well and has endured because meta in Greek also can mean beyond or above as well as after. As I read your newly coined term meta-science to me it sounds like a synonym for metaphysics. I’m easily befuddled I suppose. I’ll be back with some relevant thoughts, my friend.
Interesting Matti! So originally the “meta” term was probably just a minor convenience at the time for classifying the different writings of Aristotle, though the term serendipitously turned out to provide something far more structurally appropriate? This sort of thing clearly does happen in evolution so it doesn’t surprise me for the highly evolved English language. I’ll go deeper into my thoughts on terms when we get to epistemology and Occam’s nominalism. And I do appreciate you taking your time here. There are a couple of ways that my stated positions may not fit in with some of your beliefs so I’ll be interested to see how they play out.
Eric, we've discussed your proposals along these lines many times. I'll only point out a few things here.
The logical positivist standard was verification, which it was eventually pointed out, is itself unverifiable. That standard is too strict,. It would require throwing out too much productive science. Falsifiability is actually a weaker standard introduced by Karl Popper, who was a critic of the logical positivists. Popper argued that while we can never verify a scientific theory, we can, at least in principle, falsify them, if they are indeed false.
Of course, falsifiability has its own issues, as many have pointed out. The failure of Newtonian mechanics to predict Uranus' orbit led to the discovery of Neptune. But the failure to predict Mercury's eventually falsified the theory. Although it wasn't really accepted until there was an alternate theory: general relativity.
My other point is one I've made before, but concerns your causality standard. Causation, in the sense of an asymmetry between cause and effect, is not fundamental. It emerges with the second law of thermodynamics and the direction of entropy increasing. Without that, what we have are symmetric structural relations across time and space, which would put us in structural realism territory. And while my expectation is that these are ultimately deterministic, I hold that metaphysical assumption loosely, always prepared to bracket it if necessary to accept a successful scientific theory.
Finally, along the lines of one of my recent posts, I'm leery of anything being declared fundamental or axiomatic. Historically, it seems like everything declared fundamental by someone eventually turns out to be emergent from something else. For instance, atoms were supposed to be the fundamental building blocks of matter, but turned out to be composed of electrons, protons, and sometimes neutrons. And protons and neutrons themselves turned out to be composed of quarks. It seems unwise to be too comfortable that we've hit bedrock with our current understandings.
Thanks for the clarification on the logical positivists Mike! It was AI that told me they were essentially all about working with falsifiable ideas exclusively. I’m not too disappointed about that specific error though, since this was merely a side point to my inquiry itself. I asked if it knew of anything similar to my proposal? Nope! It said the best it could find was efforts to improve existing science, like the logical positivists and current metascience efforts. But it couldn’t find signs of anyone beyond me calling for principles from which to found science by means of expert opinion providing explicit principles. So at the time for clarity it seems to have fudged its description of the logical positivists a bit. Of course when directly asked it does agree with you that they naïvely wanted to “prove things true”, and with Popper providing a “not false” alternative. Both I think have the same problem.
Regarding causation, you’re getting into my coming metaphysics post. I figured I should at least quickly provide my own proposed three founding positions to help others grasp the concept of what explicit principles from which to found science might be? This was merely an introduction however and next time I’ll address metaphysics in detail. If you do end up deciding that your metaphysical position is both different from mine however, and better, that could be fun! I hope for some competing proposals to test the idea that even distinguished experts might develop reasonable explicit conclusions from which to found science.
Currently you might have the impression that I’m proposing or seeking “truth”? Actually this is all entirely provisional.
I guess we’ll get to metaphysics next time Mike. But one more thing if you don’t mind. What about the theme of this post itself? Are you happy with the situation in traditional science where there has only been implicit conceptions of metaphysics, epistemology, and axiology from which to work? Or do you think that conceptually there’s a case to be made for improvement? Is it reasonable to suspect that a group of professional philosophers might figure out certain explicit rules from which to found science so that it functions more effectively than today?
Eric, given the history of science, I think having any kind of prescriptive philosophy about how or what scientists might discover is dangerous. I think science should stay relentlessly pragmatic and focus on what works. Should Newton have not put forth his theory of universal gravitation because he couldn't account for the causality of its central concept? Should Darwin have withheld natural selection because he couldn't account for inherited traits? Or should Planck have stopped because he didn't like or understand the necessity of adding in quanta to his math?
I'm fine with philosophers interpreting what's happening with these discoveries. But our philosophies should follow from them, not the other way around. They may provide expectations about future discoveries, potentially promising directions for scientists to investigate, although history shows those expectations should only be loosely held.
But science will lose its way if the data and what reliably predicts it ever fail to hold primacy.
It sounds like you’re wanting to disagree Mike, though it isn’t clear that what you’re implying disagreement about is what I’m proposing. Clearly your Newton, Darwin, and Plank examples demonstrate science when it works well. So yes, let’s deny rules which hinder that sort of thing and instead promote rules that advance it. I suspect that my own particular rules would help. Furthermore much of what you’re saying that you do want seems to support the concept of having certain explicit rules about what exists and how to explore it. Sabine Hossenfelder would say that in certain quarters of physics, the point has been reached where data and what predicts it fails to hold primacy. Instead “beauty” has become prime. Without your explicit rule by which “data and the models that predict it remain prime”, science has been permitted to wander.
For my coming post I’ve started looking at ontic structural realism to see if it’s different from the concept of systemic causality? I was going to ask you about the apparently platonic point of “equations are the reality”, though fortunately when asked explicitly, AI straightened out that nonsense. With only some minor caveats it decided in the end that OSR is essentially the same, though without quite using the “causal” word.
I hope you are unlocking the power of AI by means of specific questions rather than just for summarizing long essays. Beyond such a massive encyclopedia at its disposal, I’m quite enjoying how I can get more “objective” answers from it than an actual human. We’re all biased to some degree, and even when the best of us try not to be. Beyond being programmed to make people feel good about themselves in order to get more use, AI is not subjective.
Eric, the difference between my stance and Hossenfelder's is that I'm not opposed to rigorous speculation. Consider that Copernicus' justification for his system over Ptolemy's was that the Ptolemaic systems was "a monster" and his was "more pleasing to the mind." Where would his theory have been under your rules? All it had going for it at first was that the math was easier. Lots of astronomers adopted that math while wanting nothing to do with the implied cosmology. The theory didn't have greater predictive success until Galileo started making telescopic observations (observations motivated by his being a Copernican).
I don't agree with everything he says, but philosopher of science Michael Strevens points out that scientists come up with hypotheses and theories from all kinds of motivations: logic and observation, but also philosophical, religious, and yes, even aesthetic grounds. It's not the creativity of such theories that need to be constrained, only what we require to consider them settled science. And we always have to be aware that nothing is ever settled forever.
I do ask AI questions, but I don't trust it. Whatever answers I get, I take as the starting point of an investigation. I immediately follow its citations (if provided). You'd be surprised how often those citations don't validate what it's saying. Just conversing with AI is like having a discussion with some random dude in a Reddit thread. You might get good info, but you should independently verify it.
I’m not sure what I’ve said to imply that I’m against rigorous speculation. I suspect Sabine would ask the same. For me however, I’ve got a problem when admittedly supernatural ideas are considered in science. When someone resorts to “God did it”, as well as “that’s just how reality works beyond causal function itself”, that’s a problem for me. And from what I gather, OSR has a problem with such statements too. Mind you that I’m not saying this sort of speculation is wrong. Maybe it’s right. I’m just saying when people posit such things they should be considered under an explicitly “causal plus” variety of science, not the standard form that I believe requires protection. I know that you personally don’t want magic running amok in science any more than I do. But do you also believe science would be harmed, or at least not helped, if theoretical contradictions to systemic causality were explicitly banned from standard science to instead be considered in something more inclusive?
I'm sure Hossenfelder would say she's for rigorous speculation. But her definition of "rigorous" doesn't seem to include mathematical extrapolation of well tested existing theories beyond observables, just judging by the history of what she criticizes, much less anything with only one or a limited number of additional assumptions. Unless it's an idea she likes, like superdeterminism.
Would the theories I asked about above count as "causal plus"? They all had causal gaps, gaps pointed out by critics when they were proposed. Yet despite their imperfections, they worked. They were progress. If they don't count as "causal plus," what does?
To answer your question, I think science is harmed by trying to censor or discourage ideas. The only thing that should do that is failure to fit the available data.
Eric, I must confess (again) that your work demands a lot from me. I really don’t think I can engage you at anything but a rudimentary level. So let’s begin: what is the problem you are trying to address? I just don’t see it. You devise a new complex system of approaching philosophical inquiry—but why? As a side note, you refer to the scientific method as if it is one thing. To my understanding there is no one scientific method—i.e., astronomy is not conducted the same way as microbiology. But aside from that what is the big problem that your system will fix? If we can start there perhaps I can get a handle on what you are trying to do.
Okay sounds good Matti. Stepping back to a more basic question can be a great way to potentially help grasp someone’s perspective. The essential problem as I see it, is that science should function better if it were founded better. This is merely the logical observation that things which are not founded well, whether a house, a tree, or anything at all that requires grounding to function as such, should also function better by means of better founding.
As I see it, in practice there are three pillars that found science. Two of them are inherent, or epistemology and metaphysics. They’re essentially yin and yang — one address the potential understander while the other addresses what’s to be understood. Better rules of what exists and how to potentially understand should thus give scientists a better position from which to build science.
Then regarding the third pillar, because certain scientists study value driven forms of life, their studies should also suffer without an effective grasp of value itself. This is something that I can display quite tangibly. I’d say that modern economists have a good theoretical grasp of value from which to found their science, and have also developed a vast collection of professionally undisputed models. Conversely I don’t consider psychologist to yet formally accept this same value concept for their work, and so it makes sense to me that their modeling hasn’t been nearly as successful as the modeling done in economics.
Whether or not you end up agreeing with my coming metaphysical, epistemological, and axiological principles from which to potentially improve the foundation upon which science rests, that’s the project here. Though merely implicit rules have gotten science this far, logic suggests to me that improvements could be made by means of explicit rules of science developed by a community of professionals tasked exclusively with their development.
OK, a perfectly general explanation, “science should function better if it were founded better.” But it contains no argument that it functions badly or could be improved with a different set of assumptions. And in particular your set of assumptions.
Well yes, I believe that science in general should be suffering given that it functions by means of implicit rather than professionally accepted explicit rules, and especially when attempting to model the function of value driven forms of life like ourselves. That theme was even used to name my blog. This post however is merely an introduction to the concept of founding science by means of a proposed new community which is exclusively tasked with developing such rules. It’s my coming three post that are meant to argue the ways in which science should be suffering given that the specific principles which I propose have not yet become established as such.
Well, let’s see if I can make another stab at it. The following (slightly edited) claim by you needs some unpacking: “science in general … functions by means of implicit rather than … explicit rules, and especially when attempting to model the function of value driven forms of life like ourselves.” Ok, first, I agree that the fairly short history of science as well know it, roughly from the 16th century to now, has been messy. So, yes, the way it has been conducted can be described as following “implicit” rules rather than “explicit” rules. The so-called scientific method (or methods) has stumbled along with folks we now call scientists apparently without a rule book on how one ought to do science properly. If I summarized that fairly and considering the astounding advances in our knowledge of the natural world since the 16th century, then what the heck is your problem with that? That is, shouldn’t your claim also contain an argument that it could be so much more successful if it had some tidy “explicit” rules all scientists had to follow? One wonders how our knowledge of the natural world could have possibly advanced more. We’ve gone from oxen pulling plows to landing on the moon in the last half millennium! Second, I am perplexed by your follow up claim that we need such an explicit rule book because we (human beings) seek to fulfill our values or as you say we are “value driven forms.” I agree we all have purposes, wants, needs, ideals, desires and goals. In short our actions are driven by our values. So what? In other words this sounds like you’re tapping your way to make your explicit rule book for conducting science properly and then taking over everything that you place under the title “axiology” within your rule book as a science. As you can imagine I find such a thought unacceptable for a number of reasons. But here’s an easy one, if the short history of science has been so messy then why the heck do you think we can do ethics and all that we’d consider aesthetics like the fine arts, etc., better as a science? Or, have I completely misunderstood you?
(Strangely the web version of this comment cuts off the last paragraph though it’s printed in full in Substack versions.)
Deep question there Matti, I like it! No it’s more than that. I need it! Your question can be broken into two essential parts. The first asks how I could have a problem with science using merely implicit metaphysical and epistemological guidelines given where science has nevertheless gotten to in only three or four centuries? As I see it however, science would be doing far better today if it were armed with my simple explicit rules of metaphysics and epistemology centuries ago. Essentially the power that nature provides us when we grasp effective models of its function, was destined regardless of our inefficiencies grasping such models.
(By the way, I do consider there to be only one essential scientific method. It’s a process where models of how things work are assessed on the basis of their consistency with empirical evidence.)
Then secondly you’re asking how I’m incorporating axiology into all this? A sure sign that my ideas haven’t been reduced enough here is when the term “values” is used. All values ultimately reduce back to the physics of value itself. So actually I’m not talking about transforming any art into a science. Instead this is an attempt to found the still primitive central science of psychology by means of the same value premise that founds the non central but highly advanced science of economics. For example, trying to develop a science of music should not just practically fail, but categorically fail. Fortunately I plan nothing of the sort.
You also mentioned the term “ethics” (which is commonly associated with “morality”, or how we judge the actions of people in praise worthy to shameful ways). Indeed, I did open up an ugly can of worms in my essay by stating that presumptions of morally right/wrong *behavior* results in a category error — instead there are only *feelings* of rightness to wrongness. Thus while moral feelings should ultimately be modeled in psychology, moral realists seeking to identify feeling independent right/wrong behavior could remain under the domain of philosophy, or even the new “causal plus” form of science that I propose. I won’t get into that until the final post of the series however.
At the risk of being tediously redundant, my first question was simply that your claim should contain an argument that science could be more successful if it had, according to your proposal, some explicit rules all scientists had to follow? That is a serious claim and needs justification. You now claim (again) that “…science would be doing far better today if it were armed with my simple explicit rules of metaphysics and epistemology centuries ago.”
Eric, I hate to be redundant or difficult but that is your initial claim and not any sort of an argument as to how you arrived at that conclusion. It appears that you’re asking that such a conclusion be accepted on faith. To quote the late Christopher Hitchens: What can be asserted without evidence or reasoning can be dismissed without evidence or reasoning. So, at this time I respectfully I dismiss your assertion.
I will refrain from any comments regarding my second inquiry, in general my perplexity on why such a set of explicit rules is also needed because we are as you say “value driven forms.” I simply don’t understand what that means as yet. But we can leave that for now since my first inquiry has not been satisfied and thus I feel no need to inquire further.
Yep Lee it’s good to have a project. I guess it was originally incited back in 2016 or so blogging at the sites of Massimo Pigliucci and Dan Kaufman. They were always so worried about the evils of scientism and making sure that philosophers be respected. That’s where I hatched my scheme for a new community with the sole purpose of founding science by means of expert opinion. It’s taken far too long for me to act on it! I suspect that your “Institute for Ontic Realism” has been brewing in you for a long time too.
Reality itself? There’s only one thing that I believe I can know about about the ontological truth of what exists. That would be my own existence itself, or the particular mind that is me. Descartes referred to this as “thought”, though in my post #2 (https://eborg760.substack.com/p/i-value-therefore-i-am) I refined it further as “I value therefore I am”. Ultimately my existence reduces to something that feels good/bad, (which my last post also argued should be a neurally produced electromagnetic field).
Regardless, from the perfectly certain founding truth of me that exists, science is used to empirically model all else. So what I see can only be theoretical. Humanity may have developed strong theory by this point, though as Mike and I were just saying, it’s all provisional. The perfectly certain me might instead exist by means of some sort of magical situation. And that gets to my coming post about why science must be founded upon the premise that systemic causality never fails — science is rendered obsolete to the degree that causality does fail. So science needs to be reformed with professionally agreed upon principles of metaphysics, epistemology, and axiology. And I do not merely mean to become the founder of this proposed new “metascience”. The specific principles that I propose should also go on to found future science in general.
Hey, we’re bloggers man! If we enjoy what we’re doing then it’s worthwhile for that reason alone. If we don’t then our free time goes to gardening, TV, games, or whatever one enjoys more. And if blogging is only fun because we know we’ll eventually be validated and actually won’t, well yes, that could be precarious. But fun’s still fun!
I’m in the midst of reading and outling your lengthy argument. Please forgive me. As you know I like to take my time. However, as I read, I feel the need to share a thought. This is really not germane in any way to your argument. But it’s on your use of words. “Metaphysics” is a term philosophy inherited from Aristotle. However, it was not Aristotle’s term. He never used the word himself. Aristotle’s term for this part of his philosophy was “first philosophy” or just “wisdom.”
When later editors began to compile Aristotle’s works they placed a collection of Aristotle’s works after his work on physics. The word meta in Greek means after. So, as a working title they simply called those works metaphysics—meta ta physika. The title stuck. However, that inadvertent title had a tendency to influence later philosophers as to Aristotle’s deeper meaning. However, the metaphysics title works fairly well and has endured because meta in Greek also can mean beyond or above as well as after. As I read your newly coined term meta-science to me it sounds like a synonym for metaphysics. I’m easily befuddled I suppose. I’ll be back with some relevant thoughts, my friend.
Interesting Matti! So originally the “meta” term was probably just a minor convenience at the time for classifying the different writings of Aristotle, though the term serendipitously turned out to provide something far more structurally appropriate? This sort of thing clearly does happen in evolution so it doesn’t surprise me for the highly evolved English language. I’ll go deeper into my thoughts on terms when we get to epistemology and Occam’s nominalism. And I do appreciate you taking your time here. There are a couple of ways that my stated positions may not fit in with some of your beliefs so I’ll be interested to see how they play out.
Eric, we've discussed your proposals along these lines many times. I'll only point out a few things here.
The logical positivist standard was verification, which it was eventually pointed out, is itself unverifiable. That standard is too strict,. It would require throwing out too much productive science. Falsifiability is actually a weaker standard introduced by Karl Popper, who was a critic of the logical positivists. Popper argued that while we can never verify a scientific theory, we can, at least in principle, falsify them, if they are indeed false.
Of course, falsifiability has its own issues, as many have pointed out. The failure of Newtonian mechanics to predict Uranus' orbit led to the discovery of Neptune. But the failure to predict Mercury's eventually falsified the theory. Although it wasn't really accepted until there was an alternate theory: general relativity.
My other point is one I've made before, but concerns your causality standard. Causation, in the sense of an asymmetry between cause and effect, is not fundamental. It emerges with the second law of thermodynamics and the direction of entropy increasing. Without that, what we have are symmetric structural relations across time and space, which would put us in structural realism territory. And while my expectation is that these are ultimately deterministic, I hold that metaphysical assumption loosely, always prepared to bracket it if necessary to accept a successful scientific theory.
Finally, along the lines of one of my recent posts, I'm leery of anything being declared fundamental or axiomatic. Historically, it seems like everything declared fundamental by someone eventually turns out to be emergent from something else. For instance, atoms were supposed to be the fundamental building blocks of matter, but turned out to be composed of electrons, protons, and sometimes neutrons. And protons and neutrons themselves turned out to be composed of quarks. It seems unwise to be too comfortable that we've hit bedrock with our current understandings.
Thanks for the clarification on the logical positivists Mike! It was AI that told me they were essentially all about working with falsifiable ideas exclusively. I’m not too disappointed about that specific error though, since this was merely a side point to my inquiry itself. I asked if it knew of anything similar to my proposal? Nope! It said the best it could find was efforts to improve existing science, like the logical positivists and current metascience efforts. But it couldn’t find signs of anyone beyond me calling for principles from which to found science by means of expert opinion providing explicit principles. So at the time for clarity it seems to have fudged its description of the logical positivists a bit. Of course when directly asked it does agree with you that they naïvely wanted to “prove things true”, and with Popper providing a “not false” alternative. Both I think have the same problem.
Regarding causation, you’re getting into my coming metaphysics post. I figured I should at least quickly provide my own proposed three founding positions to help others grasp the concept of what explicit principles from which to found science might be? This was merely an introduction however and next time I’ll address metaphysics in detail. If you do end up deciding that your metaphysical position is both different from mine however, and better, that could be fun! I hope for some competing proposals to test the idea that even distinguished experts might develop reasonable explicit conclusions from which to found science.
Currently you might have the impression that I’m proposing or seeking “truth”? Actually this is all entirely provisional.
I guess we’ll get to metaphysics next time Mike. But one more thing if you don’t mind. What about the theme of this post itself? Are you happy with the situation in traditional science where there has only been implicit conceptions of metaphysics, epistemology, and axiology from which to work? Or do you think that conceptually there’s a case to be made for improvement? Is it reasonable to suspect that a group of professional philosophers might figure out certain explicit rules from which to found science so that it functions more effectively than today?
Eric, given the history of science, I think having any kind of prescriptive philosophy about how or what scientists might discover is dangerous. I think science should stay relentlessly pragmatic and focus on what works. Should Newton have not put forth his theory of universal gravitation because he couldn't account for the causality of its central concept? Should Darwin have withheld natural selection because he couldn't account for inherited traits? Or should Planck have stopped because he didn't like or understand the necessity of adding in quanta to his math?
I'm fine with philosophers interpreting what's happening with these discoveries. But our philosophies should follow from them, not the other way around. They may provide expectations about future discoveries, potentially promising directions for scientists to investigate, although history shows those expectations should only be loosely held.
But science will lose its way if the data and what reliably predicts it ever fail to hold primacy.
It sounds like you’re wanting to disagree Mike, though it isn’t clear that what you’re implying disagreement about is what I’m proposing. Clearly your Newton, Darwin, and Plank examples demonstrate science when it works well. So yes, let’s deny rules which hinder that sort of thing and instead promote rules that advance it. I suspect that my own particular rules would help. Furthermore much of what you’re saying that you do want seems to support the concept of having certain explicit rules about what exists and how to explore it. Sabine Hossenfelder would say that in certain quarters of physics, the point has been reached where data and what predicts it fails to hold primacy. Instead “beauty” has become prime. Without your explicit rule by which “data and the models that predict it remain prime”, science has been permitted to wander.
For my coming post I’ve started looking at ontic structural realism to see if it’s different from the concept of systemic causality? I was going to ask you about the apparently platonic point of “equations are the reality”, though fortunately when asked explicitly, AI straightened out that nonsense. With only some minor caveats it decided in the end that OSR is essentially the same, though without quite using the “causal” word.
I hope you are unlocking the power of AI by means of specific questions rather than just for summarizing long essays. Beyond such a massive encyclopedia at its disposal, I’m quite enjoying how I can get more “objective” answers from it than an actual human. We’re all biased to some degree, and even when the best of us try not to be. Beyond being programmed to make people feel good about themselves in order to get more use, AI is not subjective.
Eric, the difference between my stance and Hossenfelder's is that I'm not opposed to rigorous speculation. Consider that Copernicus' justification for his system over Ptolemy's was that the Ptolemaic systems was "a monster" and his was "more pleasing to the mind." Where would his theory have been under your rules? All it had going for it at first was that the math was easier. Lots of astronomers adopted that math while wanting nothing to do with the implied cosmology. The theory didn't have greater predictive success until Galileo started making telescopic observations (observations motivated by his being a Copernican).
I don't agree with everything he says, but philosopher of science Michael Strevens points out that scientists come up with hypotheses and theories from all kinds of motivations: logic and observation, but also philosophical, religious, and yes, even aesthetic grounds. It's not the creativity of such theories that need to be constrained, only what we require to consider them settled science. And we always have to be aware that nothing is ever settled forever.
I do ask AI questions, but I don't trust it. Whatever answers I get, I take as the starting point of an investigation. I immediately follow its citations (if provided). You'd be surprised how often those citations don't validate what it's saying. Just conversing with AI is like having a discussion with some random dude in a Reddit thread. You might get good info, but you should independently verify it.
I’m not sure what I’ve said to imply that I’m against rigorous speculation. I suspect Sabine would ask the same. For me however, I’ve got a problem when admittedly supernatural ideas are considered in science. When someone resorts to “God did it”, as well as “that’s just how reality works beyond causal function itself”, that’s a problem for me. And from what I gather, OSR has a problem with such statements too. Mind you that I’m not saying this sort of speculation is wrong. Maybe it’s right. I’m just saying when people posit such things they should be considered under an explicitly “causal plus” variety of science, not the standard form that I believe requires protection. I know that you personally don’t want magic running amok in science any more than I do. But do you also believe science would be harmed, or at least not helped, if theoretical contradictions to systemic causality were explicitly banned from standard science to instead be considered in something more inclusive?
On your disdain for AI, duly noted! 🙂
I'm sure Hossenfelder would say she's for rigorous speculation. But her definition of "rigorous" doesn't seem to include mathematical extrapolation of well tested existing theories beyond observables, just judging by the history of what she criticizes, much less anything with only one or a limited number of additional assumptions. Unless it's an idea she likes, like superdeterminism.
Would the theories I asked about above count as "causal plus"? They all had causal gaps, gaps pointed out by critics when they were proposed. Yet despite their imperfections, they worked. They were progress. If they don't count as "causal plus," what does?
To answer your question, I think science is harmed by trying to censor or discourage ideas. The only thing that should do that is failure to fit the available data.
Eric, I must confess (again) that your work demands a lot from me. I really don’t think I can engage you at anything but a rudimentary level. So let’s begin: what is the problem you are trying to address? I just don’t see it. You devise a new complex system of approaching philosophical inquiry—but why? As a side note, you refer to the scientific method as if it is one thing. To my understanding there is no one scientific method—i.e., astronomy is not conducted the same way as microbiology. But aside from that what is the big problem that your system will fix? If we can start there perhaps I can get a handle on what you are trying to do.
Okay sounds good Matti. Stepping back to a more basic question can be a great way to potentially help grasp someone’s perspective. The essential problem as I see it, is that science should function better if it were founded better. This is merely the logical observation that things which are not founded well, whether a house, a tree, or anything at all that requires grounding to function as such, should also function better by means of better founding.
As I see it, in practice there are three pillars that found science. Two of them are inherent, or epistemology and metaphysics. They’re essentially yin and yang — one address the potential understander while the other addresses what’s to be understood. Better rules of what exists and how to potentially understand should thus give scientists a better position from which to build science.
Then regarding the third pillar, because certain scientists study value driven forms of life, their studies should also suffer without an effective grasp of value itself. This is something that I can display quite tangibly. I’d say that modern economists have a good theoretical grasp of value from which to found their science, and have also developed a vast collection of professionally undisputed models. Conversely I don’t consider psychologist to yet formally accept this same value concept for their work, and so it makes sense to me that their modeling hasn’t been nearly as successful as the modeling done in economics.
Whether or not you end up agreeing with my coming metaphysical, epistemological, and axiological principles from which to potentially improve the foundation upon which science rests, that’s the project here. Though merely implicit rules have gotten science this far, logic suggests to me that improvements could be made by means of explicit rules of science developed by a community of professionals tasked exclusively with their development.
OK, a perfectly general explanation, “science should function better if it were founded better.” But it contains no argument that it functions badly or could be improved with a different set of assumptions. And in particular your set of assumptions.
I think it necessary to make my rebuttal in small steps. I’m deliberately moving is baby steps here.
Well yes, I believe that science in general should be suffering given that it functions by means of implicit rather than professionally accepted explicit rules, and especially when attempting to model the function of value driven forms of life like ourselves. That theme was even used to name my blog. This post however is merely an introduction to the concept of founding science by means of a proposed new community which is exclusively tasked with developing such rules. It’s my coming three post that are meant to argue the ways in which science should be suffering given that the specific principles which I propose have not yet become established as such.
Well, let’s see if I can make another stab at it. The following (slightly edited) claim by you needs some unpacking: “science in general … functions by means of implicit rather than … explicit rules, and especially when attempting to model the function of value driven forms of life like ourselves.” Ok, first, I agree that the fairly short history of science as well know it, roughly from the 16th century to now, has been messy. So, yes, the way it has been conducted can be described as following “implicit” rules rather than “explicit” rules. The so-called scientific method (or methods) has stumbled along with folks we now call scientists apparently without a rule book on how one ought to do science properly. If I summarized that fairly and considering the astounding advances in our knowledge of the natural world since the 16th century, then what the heck is your problem with that? That is, shouldn’t your claim also contain an argument that it could be so much more successful if it had some tidy “explicit” rules all scientists had to follow? One wonders how our knowledge of the natural world could have possibly advanced more. We’ve gone from oxen pulling plows to landing on the moon in the last half millennium! Second, I am perplexed by your follow up claim that we need such an explicit rule book because we (human beings) seek to fulfill our values or as you say we are “value driven forms.” I agree we all have purposes, wants, needs, ideals, desires and goals. In short our actions are driven by our values. So what? In other words this sounds like you’re tapping your way to make your explicit rule book for conducting science properly and then taking over everything that you place under the title “axiology” within your rule book as a science. As you can imagine I find such a thought unacceptable for a number of reasons. But here’s an easy one, if the short history of science has been so messy then why the heck do you think we can do ethics and all that we’d consider aesthetics like the fine arts, etc., better as a science? Or, have I completely misunderstood you?
(Strangely the web version of this comment cuts off the last paragraph though it’s printed in full in Substack versions.)
Deep question there Matti, I like it! No it’s more than that. I need it! Your question can be broken into two essential parts. The first asks how I could have a problem with science using merely implicit metaphysical and epistemological guidelines given where science has nevertheless gotten to in only three or four centuries? As I see it however, science would be doing far better today if it were armed with my simple explicit rules of metaphysics and epistemology centuries ago. Essentially the power that nature provides us when we grasp effective models of its function, was destined regardless of our inefficiencies grasping such models.
(By the way, I do consider there to be only one essential scientific method. It’s a process where models of how things work are assessed on the basis of their consistency with empirical evidence.)
Then secondly you’re asking how I’m incorporating axiology into all this? A sure sign that my ideas haven’t been reduced enough here is when the term “values” is used. All values ultimately reduce back to the physics of value itself. So actually I’m not talking about transforming any art into a science. Instead this is an attempt to found the still primitive central science of psychology by means of the same value premise that founds the non central but highly advanced science of economics. For example, trying to develop a science of music should not just practically fail, but categorically fail. Fortunately I plan nothing of the sort.
You also mentioned the term “ethics” (which is commonly associated with “morality”, or how we judge the actions of people in praise worthy to shameful ways). Indeed, I did open up an ugly can of worms in my essay by stating that presumptions of morally right/wrong *behavior* results in a category error — instead there are only *feelings* of rightness to wrongness. Thus while moral feelings should ultimately be modeled in psychology, moral realists seeking to identify feeling independent right/wrong behavior could remain under the domain of philosophy, or even the new “causal plus” form of science that I propose. I won’t get into that until the final post of the series however.
At the risk of being tediously redundant, my first question was simply that your claim should contain an argument that science could be more successful if it had, according to your proposal, some explicit rules all scientists had to follow? That is a serious claim and needs justification. You now claim (again) that “…science would be doing far better today if it were armed with my simple explicit rules of metaphysics and epistemology centuries ago.”
Eric, I hate to be redundant or difficult but that is your initial claim and not any sort of an argument as to how you arrived at that conclusion. It appears that you’re asking that such a conclusion be accepted on faith. To quote the late Christopher Hitchens: What can be asserted without evidence or reasoning can be dismissed without evidence or reasoning. So, at this time I respectfully I dismiss your assertion.
I will refrain from any comments regarding my second inquiry, in general my perplexity on why such a set of explicit rules is also needed because we are as you say “value driven forms.” I simply don’t understand what that means as yet. But we can leave that for now since my first inquiry has not been satisfied and thus I feel no need to inquire further.
This seems like a good place for a private response, so I’ll try that.