5 ms·

There's not much in theoretical physics nowadays that you can verify (or falsify) with a direct experiment. Beyond a certain level, physics becomes philosophic

by kijin 12y ago

There's not much in theoretical physics nowadays that you can verify (or falsify) with a direct experiment. Beyond a certain level, physics becomes philosophical, with all of the usual drawbacks of philosophy: lots of outlandish conjectures with no way to verify any of them anytime soon. The Popperian ideal of science that you describe is only practiced within what Kuhn calls "normal science", i.e. pushing the cutting edge, bit by bit, safely from behind the blade. Sure, this makes up 99% of scientific research, but most of it is boring. Significant progress almost always requires "revolutionary science", i.e. jumping in front of the blade, into the bleeding edge. When you live in the bleeding edge, the normal process of cojectures-and-refutations no longer applies. (How do you verify/falsify the origin of life without going back billions of years in a time machine?) Unfortunately, the unkempt bleeding edge of science tends to attract journalistic sensationalism, just like a pool of blood attracts flies.

vixen99 12y ago
'The Popperian ideal of science that you describe is only practiced within what Kuhn calls "normal science" '. Where do you get this definitive statement from? Popper writes about 'bold conjectures'. Maybe he should have used more vivid adjectives. Popper recognized that the best scientists appreciate the tentative nature of their conjectures and expect that they will be superseded.
- kijin 12y ago
  > Where do you get this definitive statement from? Kuhn's Structure of Scientific Revolutions was published in 1962. For over half a century since then, sociologists, historians, philosophers, and political scientists have been studying how modern science works, both on a small scale (in the lab) and on a large scale (as a society). Most of the results seem to confirm the basic idea set out by Kuhn, although many details remain controversial. Rest assured, the boldness and tentativeness of scientific conjectures are not under any serious doubt. The part of Popper's model that later researchers have challenged is mostly about refutation and/or verification. For political and psychological reasons, as well as practical limitations, it is exceedingly difficult for scientists to conduct experiments that conclusively and unambiguously refute a hypothesis. As a result, the bleeding edge of scientific research often follows fads rather than solid experimental evidence. Personal charisma, funding decisions, and political concerns dominate decision-making processes, and evidence is often so ambiguous at this stage that anyone can cite it to their advantage. It takes a while for this craziness to settle down. In fields like theoretical physics where it is especially difficult to perform meaningful experiments, the time it takes for the dust to settle tends to be longer than in other fields. It can take decades, compared to months or years in other fields. Only when a theory is no longer a fad and falls safely behind the cutting edge (i.e. it becomes "normal science"), do scientists finally get a chance to scrutinize it with solid experiments. Even then, politics can get in the way of proper refutation while the proponents of the theory occupy authoritative positions in academia. As Max Planck once said: > "A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it." (The above quote is often summarized as "Science advances one funeral at a time.") Summary: A lot of conditions need to be just right in order for bold conjectures to be boldly refuted on the basis of solid experimental evidence. Bleeding-edge science often doesn't meet such conditions, due to both technical and human shortcomings. Source: I have a Ph.D. in Science and Technology Studies (STS) [1], an umbrella term for interdisciplinary studies of science and technology in social, political, and historical contexts. [1] https://en.wikipedia.org/wiki/Science,_technology_and_society https://en.wikipedia.org/wiki/Science,_technology_and_societ...
  - guscost 12y ago
    Most of what you call "the bleeding edge" I would probably call "not yet science." It's totally fine that coming up with a plan to falsify some of these theories is beyond our current capabilities. I don't expect that, and I don't think it's necessary before we can discuss interesting ideas. But the second someone gives up on falsifying their theory and starts evangelizing, it ceases to be science. That should not be in dispute.
    - kijin 12y ago
      Fair enough. We might disagree about the exact boundaries of "science" (I tend to think of it more as a spectrum than as a matter of black-and-white boundary), but I totally agree with you that scientists, in general, should try to falsify instead of evangelizing. Not even trying to falsify would be gross negligence.
  - drostie 12y ago
    If people are interested in this sort of thing, it also helps to read some of the works of Imre Lakatos. In my words, not his, he advocates that real scientific progress happens by a sort of natural selection of theories. The basic idea is that new grad students pick whatever theories are most likely to easily let them publish science papers; so that basically certain theories make their grad students more likely to write papers with them inside. Those theories thereby "reproduce" and/or "die". So, for example, we had an article here the other day which lamented why quantum mechanics was still Copenhagen-y rather than being dominated by pilot waves. The basic answer to this, following Lakatos, is that doing stuff with naive quantum mechanics and probabilistic thinking is easier for grad students to work with, as opposed to extending the pilot wave theory to encompass your new direction and then examining it for statistics.
    - kijin 12y ago
      Yes, the similarity between science and evolution is one of the major themes of STS, emphasized by both Kuhn and Lakatos. The reason this makes a lot of people uncomfortable is that it messes with our notion of scientific progress. This notion is rooted in 19th century romanticism, and predicts inevitable progress toward totality. It sees science as a matter of manifest destiny, guided by the truth and approximating it ever more closely. Evolution, however, does not have any goals. It might or might not tend to produce complex and intelligent organisms, but it sure ain't guided by any overarching ideal or telos apart from the laws of thermodynamics. Evolution wanders aimlessly, making do with whatever the universe randomly throws at it. This is not the kind of model that most people want to associate with science. Meanwhile, transhumanists who rally around the idea of the technological singularity represent the last holdout of the old, romantic ideal. They can maintain a semblance of consistency because they also reject the relevance of evolutionary biology to the future of humanity: the age of evolution is over, and AI will take over from here. I'm not quite sure what to make of this idea, but it will be really interesting to watch how their predictions unfold (or not).

by kijin 12y ago

vixen99 12y ago

'The Popperian ideal of science that you describe is only practiced within what Kuhn calls "normal science" '. Where do you get this definitive statement from? Popper writes about 'bold conjectures'. Maybe he should have used more vivid adjectives. Popper recognized that the best scientists appreciate the tentative nature of their conjectures and expect that they will be superseded.

kijin 12y ago
> Where do you get this definitive statement from? Kuhn's Structure of Scientific Revolutions was published in 1962. For over half a century since then, sociologists, historians, philosophers, and political scientists have been studying how modern science works, both on a small scale (in the lab) and on a large scale (as a society). Most of the results seem to confirm the basic idea set out by Kuhn, although many details remain controversial. Rest assured, the boldness and tentativeness of scientific conjectures are not under any serious doubt. The part of Popper's model that later researchers have challenged is mostly about refutation and/or verification. For political and psychological reasons, as well as practical limitations, it is exceedingly difficult for scientists to conduct experiments that conclusively and unambiguously refute a hypothesis. As a result, the bleeding edge of scientific research often follows fads rather than solid experimental evidence. Personal charisma, funding decisions, and political concerns dominate decision-making processes, and evidence is often so ambiguous at this stage that anyone can cite it to their advantage. It takes a while for this craziness to settle down. In fields like theoretical physics where it is especially difficult to perform meaningful experiments, the time it takes for the dust to settle tends to be longer than in other fields. It can take decades, compared to months or years in other fields. Only when a theory is no longer a fad and falls safely behind the cutting edge (i.e. it becomes "normal science"), do scientists finally get a chance to scrutinize it with solid experiments. Even then, politics can get in the way of proper refutation while the proponents of the theory occupy authoritative positions in academia. As Max Planck once said: > "A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it." (The above quote is often summarized as "Science advances one funeral at a time.") Summary: A lot of conditions need to be just right in order for bold conjectures to be boldly refuted on the basis of solid experimental evidence. Bleeding-edge science often doesn't meet such conditions, due to both technical and human shortcomings. Source: I have a Ph.D. in Science and Technology Studies (STS) [1], an umbrella term for interdisciplinary studies of science and technology in social, political, and historical contexts. [1] https://en.wikipedia.org/wiki/Science,_technology_and_society https://en.wikipedia.org/wiki/Science,_technology_and_societ...
- guscost 12y ago
  Most of what you call "the bleeding edge" I would probably call "not yet science." It's totally fine that coming up with a plan to falsify some of these theories is beyond our current capabilities. I don't expect that, and I don't think it's necessary before we can discuss interesting ideas. But the second someone gives up on falsifying their theory and starts evangelizing, it ceases to be science. That should not be in dispute.
  - kijin 12y ago
    Fair enough. We might disagree about the exact boundaries of "science" (I tend to think of it more as a spectrum than as a matter of black-and-white boundary), but I totally agree with you that scientists, in general, should try to falsify instead of evangelizing. Not even trying to falsify would be gross negligence.
- drostie 12y ago
  If people are interested in this sort of thing, it also helps to read some of the works of Imre Lakatos. In my words, not his, he advocates that real scientific progress happens by a sort of natural selection of theories. The basic idea is that new grad students pick whatever theories are most likely to easily let them publish science papers; so that basically certain theories make their grad students more likely to write papers with them inside. Those theories thereby "reproduce" and/or "die". So, for example, we had an article here the other day which lamented why quantum mechanics was still Copenhagen-y rather than being dominated by pilot waves. The basic answer to this, following Lakatos, is that doing stuff with naive quantum mechanics and probabilistic thinking is easier for grad students to work with, as opposed to extending the pilot wave theory to encompass your new direction and then examining it for statistics.
  - kijin 12y ago
    Yes, the similarity between science and evolution is one of the major themes of STS, emphasized by both Kuhn and Lakatos. The reason this makes a lot of people uncomfortable is that it messes with our notion of scientific progress. This notion is rooted in 19th century romanticism, and predicts inevitable progress toward totality. It sees science as a matter of manifest destiny, guided by the truth and approximating it ever more closely. Evolution, however, does not have any goals. It might or might not tend to produce complex and intelligent organisms, but it sure ain't guided by any overarching ideal or telos apart from the laws of thermodynamics. Evolution wanders aimlessly, making do with whatever the universe randomly throws at it. This is not the kind of model that most people want to associate with science. Meanwhile, transhumanists who rally around the idea of the technological singularity represent the last holdout of the old, romantic ideal. They can maintain a semblance of consistency because they also reject the relevance of evolutionary biology to the future of humanity: the age of evolution is over, and AI will take over from here. I'm not quite sure what to make of this idea, but it will be really interesting to watch how their predictions unfold (or not).