5 ms·

I call this the "abstraction paradox". There is no way that an abstraction can fully equate what it abstracts. "Apple" could never equal a real apple. But abst

by unabst 11y ago

I call this the "abstraction paradox". There is no way that an abstraction can fully equate what it abstracts. "Apple" could never equal a real apple. But abstraction allows us to think about real apples. In science, all theories and models are abstractions. Hence, they are always approximations, and can be invalidated or nullified with edge cases that can indeed be backed by evidence. But the opposite is also true. If a theory is based on evidence, then it will always be an approximation of the truth to some degree. Hence, there will always be hard evidence that validates the theory also -- ideally the evidence the theory is based on. The ultimate test is application. If a theory or model has purpose, as in, has a use, then the value of that theory can be measured by its consequences. Otherwise, frankly, academics arguing over what has no consequence will have no consequence outside the circles of those who argue over it. I sense "fuck nuance" as a message to them. We still use Newtonian mechanics where objects are big and slow enough. But we also use general relativity, quantum mechanics, quantum field theory, etc, where more accuracy is required. So we already know theories have limits and degrade with constraints, and we switch theories. This is true of EVERY THEORY. And that is the abstraction paradox. If you're arguing over someone just to prove them wrong, congratulations, you've just proved you're completely redundant. So fuck you. (in the spirit of the author)

wyager 11y ago
Abstractions need not be approximations. Schroedinger's equation is an abstraction, but as far as we know it's exact.
- ludamad 11y ago
  Abstractions of social sciences do need to be approximations :-)
- unabst 11y ago
  The moment the theory is the reality, it's exact. Like cartoons. Like definitions. Like math. But physics and science is about reality. It's about the physical apple, not the word apple. A real apple is physical. There is no amount of theory or symbolism that will produce a real apple in the minds and systems that simulate it. That's the paradox.
  - coldtea 11y ago
    >A real apple is physical. There is no amount of theory or symbolism that will produce a real apple in the minds and systems that simulate it. That's the paradox. Actually, I see two problems with this notion. First, there's no "paradox". That "no amount or theory or symbolism will produce a real apple" is not paradoxical. If anything, it's common sense. Second, this common sense might very well be wrong (and that would be a real paradox this time). In fact, what you call "real apple" might itself be a simulation (along with all the universe we know) -- created by higher order beings. Still, you might say, our theories, us being inhabitants of the (simulated) universe, won't produce a real apple for us. But who's to say that? It depends on the rules of the simulated universe. And we can't say with 100% certainty we know the rules of ours (if it's one such). E.g. the rules could very well allow the thoughts and symbolism of its inhabitants to create new objects in the universe. In that sense, the universe would be like Lisp (where data are code), or like using "reflection".
- smaddox 11y ago
  Schrodinger's equation doesn't even include special relativistic effects. It's not exact (at least not if you mean it exactly describes nature in general). Neither is Dirac's equation; it includes special relativistic effects for a single particle, but not general relativistic effects or multi-particle interactions (at least as far as I understand it).
  - wyager 11y ago
    The general form using the unspecified Hamiltonian operator is compatible with relativity.
- cyphar 11y ago
  Schroedinger's equation isn't a theory. It's a law. And like other laws (F = ma) doesn't account for a bunch of other things. So yes, it still is an approximation.
- Confusion 11y ago
  What is Schroedinger's equation an abstraction of? What details does it abstract from? How can it be that the omission of those details does not result in an inexact representation of the unabstracted, more complete, situation? BTW, there are no exact solutions to the Schroedinger equation of most interesting systems: we only get numerical approximations to the values of interest from it.
  - keithpeter 11y ago
    https://en.wikipedia.org/wiki/Hartree%E2%80%93Fock_method https://en.wikipedia.org/wiki/Hartree%E2%80%93Fock_method Some of those numerical solutions were really important historically and lead to the further development of numerical techniques. Back to main thread: I'm thinking of Lakatos' Proofs and Refutations, itself a play on Popper's Conjectures and Refutations, designed itself to escape from 'induction' based thinking, e.g. Hans Reichenbach Long time since I've thought of those, thanks.
TheOtherHobbes 11y ago
>Otherwise, frankly, academics arguing over what has no consequence will have no consequence outside the circles of those who argue over it. Sociology is mostly a series of status games played by academics. It's part of the humanities, which means the methods of play are peer group signalling and persuasive story-telling, not reality testing. It's not engineering, where models actually have to predict reality with at least a minimal baseline of accuracy and reliability. Engineers don't have a problem with nuance, and are suspicious of toy models that don't include details - because they're just not useful. If you model flight characteristics for an airliner but don't deal with edge cases, the plane can fall out of the sky. If you offer a grand social theory of X but get the details wrong, no one really cares, because sociologists have very little influence on practical policy, and the influence they do have is carefully chosen for political applicability by non-sociologists. Occasionally useful insights fall out of social psychology and sociology. But they have limited applicability because our understanding of social psychology is pre-Newtonian - some things seem to work, but no one is really sure why - and our understanding of politics and economics is barbaric and barely better than medieval, on a good day.
- danharaj 11y ago
  Your entire post is a series of sociological assertions about academics and engineers without scientific evidence. I'm sure the irony was intentional.
  - AnthonyMouse 11y ago
    That entire post can be reduced to two core assertions: 1) Details matter in engineering because products will actually be manufactured according to those details and if the details are wrong then planes would fall out of the sky. (Which we know they rarely do.) 2) Details don't matter in social theory because practitioners don't use academic theories to make policy. The theories can't be tested because they aren't used. Most social science "theories" are therefore a hypothesis without an experiment. The only statement about people or social behavior is that engineers put their theories into practice with more frequency and rigor than sociologists, which is empirically true.
raverbashing 11y ago
I see this problem as such: sciences beyond physics have phenomenon dependent on many more variables, many of which are unknown On Physics, beyond the obvious, the rest can be derided as noise (you won't have an exact measure of anything, sometimes this is a new phenomenon yet to be discovered, or something known but overlooked in this situation, but usually "noise") Now when you depend on more variables, your first guess might only predict the result 60% of the time (as an example). Better than random choice, but still has a low predictability. Then you add more "obvious" variables and can result in, 80% prediction, which is good, but not as good as you can get with physics. This can be social sciences or even something more hard as biology or medicine. In the end it's an issue of overfitting/underfitting
- unabst 11y ago
  This. It's absolutely about overfitting/underfitting. Just as our eyes see a lot more but can only focus on one thing, abstractions are just that. They focus on specific things. But the problem with the physical reality is that it does not end, and everything is connected. Every time we make an abstraction we are severing the subject from it's surroundings which physically are inseparable. They are only theoretically separable. Hence scientific research begins with closing an otherwise open system. Even when we weigh an apple, it's weight is influenced by the gravitational pull of the sun and everything else. Hence what matters depends on application. If you're buying apples by weight, we can safely exclude these influences as inconsequential to the price of your apples. Good enough is what is fitting, and they make good abstractions and good science. It's easy to lose track of utility in social sciences, because history only happens once, and theories often appear to have no predicting power or future use. But if anything, social sciences provide us with introspection. It tells us who we are and how we've been. If you already practice this on a personal level, then you already know how valuable this is at any scale. If not, you should begin to practice it on yourself first, because the value of things is most easily admired through experience.
3minus1 11y ago
> Verisimilitude is a philosophical concept that distinguishes between the relative and apparent (or seemingly so) truth and falsity of assertions and hypotheses. The problem of verisimilitude is the problem of articulating what it takes for one false theory to be closer to the truth than another false theory https://en.wikipedia.org/wiki/Verisimilitude https://en.wikipedia.org/wiki/Verisimilitude
julie1 11y ago
Well, science has not your "abstraction paradox". Science uses a notion called domain of validity for their abstraction. Domain of validity gives conditions for which there is actually no ambiguity. There are domains where science knows it knows nothing. It is called transitions. In science you take more time learning these "domain of validity" than the actual laws. In fact, you have to learn to calculate them by yourself. When Einstein's relativity apply? When you go to fast. When is a photon behaving like a corpuscule ? When you try to count them. When does photon behave like a wave? When you try to measure the interferences. When does deterministic laws fails? When you have to much independent entities, then you use statistical physics. According to the possibility for entities to share the same energy level you then choose your distribution to integrate all the states... Science is not about knowing formula, it is about knowing when you can apply them.
- unabst 11y ago
  The abstraction paradox is with regard to any abstraction. Theories are abstractions. But science is not just theories. It's application as you say. Arguments should be over accuracy and application. To simply prove a theory wrong is redundant because it will always be possible. It's what can be referred to as a "dick move". Your tone was misleading, but everything you said backs the theory. I must thank you.
  - julie1 11y ago
    A map is not the territory it represents, but, if correct, it has a similar structure to the territory, which accounts for its usefulness. — Alfred Korzybski, Science and Sanity (1933, p. 58)

by unabst 11y ago

wyager 11y ago

Abstractions need not be approximations. Schroedinger's equation is an abstraction, but as far as we know it's exact.

ludamad 11y ago
Abstractions of social sciences do need to be approximations :-)
unabst 11y ago
The moment the theory is the reality, it's exact. Like cartoons. Like definitions. Like math. But physics and science is about reality. It's about the physical apple, not the word apple. A real apple is physical. There is no amount of theory or symbolism that will produce a real apple in the minds and systems that simulate it. That's the paradox.
- coldtea 11y ago
  >A real apple is physical. There is no amount of theory or symbolism that will produce a real apple in the minds and systems that simulate it. That's the paradox. Actually, I see two problems with this notion. First, there's no "paradox". That "no amount or theory or symbolism will produce a real apple" is not paradoxical. If anything, it's common sense. Second, this common sense might very well be wrong (and that would be a real paradox this time). In fact, what you call "real apple" might itself be a simulation (along with all the universe we know) -- created by higher order beings. Still, you might say, our theories, us being inhabitants of the (simulated) universe, won't produce a real apple for us. But who's to say that? It depends on the rules of the simulated universe. And we can't say with 100% certainty we know the rules of ours (if it's one such). E.g. the rules could very well allow the thoughts and symbolism of its inhabitants to create new objects in the universe. In that sense, the universe would be like Lisp (where data are code), or like using "reflection".
smaddox 11y ago
Schrodinger's equation doesn't even include special relativistic effects. It's not exact (at least not if you mean it exactly describes nature in general). Neither is Dirac's equation; it includes special relativistic effects for a single particle, but not general relativistic effects or multi-particle interactions (at least as far as I understand it).
- wyager 11y ago
  The general form using the unspecified Hamiltonian operator is compatible with relativity.
cyphar 11y ago
Schroedinger's equation isn't a theory. It's a law. And like other laws (F = ma) doesn't account for a bunch of other things. So yes, it still is an approximation.
Confusion 11y ago
What is Schroedinger's equation an abstraction of? What details does it abstract from? How can it be that the omission of those details does not result in an inexact representation of the unabstracted, more complete, situation? BTW, there are no exact solutions to the Schroedinger equation of most interesting systems: we only get numerical approximations to the values of interest from it.
- keithpeter 11y ago
  https://en.wikipedia.org/wiki/Hartree%E2%80%93Fock_method https://en.wikipedia.org/wiki/Hartree%E2%80%93Fock_method Some of those numerical solutions were really important historically and lead to the further development of numerical techniques. Back to main thread: I'm thinking of Lakatos' Proofs and Refutations, itself a play on Popper's Conjectures and Refutations, designed itself to escape from 'induction' based thinking, e.g. Hans Reichenbach Long time since I've thought of those, thanks.

TheOtherHobbes 11y ago

>Otherwise, frankly, academics arguing over what has no consequence will have no consequence outside the circles of those who argue over it. Sociology is mostly a series of status games played by academics. It's part of the humanities, which means the methods of play are peer group signalling and persuasive story-telling, not reality testing. It's not engineering, where models actually have to predict reality with at least a minimal baseline of accuracy and reliability. Engineers don't have a problem with nuance, and are suspicious of toy models that don't include details - because they're just not useful. If you model flight characteristics for an airliner but don't deal with edge cases, the plane can fall out of the sky. If you offer a grand social theory of X but get the details wrong, no one really cares, because sociologists have very little influence on practical policy, and the influence they do have is carefully chosen for political applicability by non-sociologists. Occasionally useful insights fall out of social psychology and sociology. But they have limited applicability because our understanding of social psychology is pre-Newtonian - some things seem to work, but no one is really sure why - and our understanding of politics and economics is barbaric and barely better than medieval, on a good day.

danharaj 11y ago
Your entire post is a series of sociological assertions about academics and engineers without scientific evidence. I'm sure the irony was intentional.
- AnthonyMouse 11y ago
  That entire post can be reduced to two core assertions: 1) Details matter in engineering because products will actually be manufactured according to those details and if the details are wrong then planes would fall out of the sky. (Which we know they rarely do.) 2) Details don't matter in social theory because practitioners don't use academic theories to make policy. The theories can't be tested because they aren't used. Most social science "theories" are therefore a hypothesis without an experiment. The only statement about people or social behavior is that engineers put their theories into practice with more frequency and rigor than sociologists, which is empirically true.

raverbashing 11y ago

I see this problem as such: sciences beyond physics have phenomenon dependent on many more variables, many of which are unknown On Physics, beyond the obvious, the rest can be derided as noise (you won't have an exact measure of anything, sometimes this is a new phenomenon yet to be discovered, or something known but overlooked in this situation, but usually "noise") Now when you depend on more variables, your first guess might only predict the result 60% of the time (as an example). Better than random choice, but still has a low predictability. Then you add more "obvious" variables and can result in, 80% prediction, which is good, but not as good as you can get with physics. This can be social sciences or even something more hard as biology or medicine. In the end it's an issue of overfitting/underfitting

unabst 11y ago
This. It's absolutely about overfitting/underfitting. Just as our eyes see a lot more but can only focus on one thing, abstractions are just that. They focus on specific things. But the problem with the physical reality is that it does not end, and everything is connected. Every time we make an abstraction we are severing the subject from it's surroundings which physically are inseparable. They are only theoretically separable. Hence scientific research begins with closing an otherwise open system. Even when we weigh an apple, it's weight is influenced by the gravitational pull of the sun and everything else. Hence what matters depends on application. If you're buying apples by weight, we can safely exclude these influences as inconsequential to the price of your apples. Good enough is what is fitting, and they make good abstractions and good science. It's easy to lose track of utility in social sciences, because history only happens once, and theories often appear to have no predicting power or future use. But if anything, social sciences provide us with introspection. It tells us who we are and how we've been. If you already practice this on a personal level, then you already know how valuable this is at any scale. If not, you should begin to practice it on yourself first, because the value of things is most easily admired through experience.

3minus1 11y ago

> Verisimilitude is a philosophical concept that distinguishes between the relative and apparent (or seemingly so) truth and falsity of assertions and hypotheses. The problem of verisimilitude is the problem of articulating what it takes for one false theory to be closer to the truth than another false theory https://en.wikipedia.org/wiki/Verisimilitude https://en.wikipedia.org/wiki/Verisimilitude

julie1 11y ago

Well, science has not your "abstraction paradox". Science uses a notion called domain of validity for their abstraction. Domain of validity gives conditions for which there is actually no ambiguity. There are domains where science knows it knows nothing. It is called transitions. In science you take more time learning these "domain of validity" than the actual laws. In fact, you have to learn to calculate them by yourself. When Einstein's relativity apply? When you go to fast. When is a photon behaving like a corpuscule ? When you try to count them. When does photon behave like a wave? When you try to measure the interferences. When does deterministic laws fails? When you have to much independent entities, then you use statistical physics. According to the possibility for entities to share the same energy level you then choose your distribution to integrate all the states... Science is not about knowing formula, it is about knowing when you can apply them.

unabst 11y ago
The abstraction paradox is with regard to any abstraction. Theories are abstractions. But science is not just theories. It's application as you say. Arguments should be over accuracy and application. To simply prove a theory wrong is redundant because it will always be possible. It's what can be referred to as a "dick move". Your tone was misleading, but everything you said backs the theory. I must thank you.
- julie1 11y ago
  A map is not the territory it represents, but, if correct, it has a similar structure to the territory, which accounts for its usefulness. — Alfred Korzybski, Science and Sanity (1933, p. 58)