4 ms·

Taubes interview on Peter Attia's podcast is fascinating. He starts by laying out his career in journalism bringing light to scientific malpractice and otherwis

by kuhewa 5y ago

Taubes interview on Peter Attia's podcast is fascinating. He starts by laying out his career in journalism bringing light to scientific malpractice and otherwise very sharp minds totally deluding themselves when they get stuck on a pet hypothesis. The second half of the podcast was about his later career where he took on sugar. He then illustrates all of the self delusion and poor scientific thinking he had just described. It's kind of amazing. https://peterattiamd.com/garytaubes/ https://peterattiamd.com/garytaubes/

omginternets 5y ago
Can you point to specific instances of self-delusion and poor scientific thinking? The goalpost for these accusations, as far as I'm concerned, is that you would need to show a pattern of ignoring contradicting evidence, which -- it should be noted -- is distinct from arguing that there are fundamental flaws in studies that prima facie invalidate his theory. In the latter case, it's just scholarly disagreement. It's late at night, and I'm not sure how to write this comment without it sounding like a loaded question, so please accept my assurances that it is not. It's been a while since I've read Taubes, and while the sugar/obesity debate is hardly settled, my impression was not one of delusion, nor of poor scientific reasoning. Certainly, it was a cut above the average quality of nutritional science research (though this is admittedly a very low bar).
- kuhewa 5y ago
  My comment was not just with regards to the facts of the subject but the way of thinking epistemologically about hypotheses in science, because he literally spends the first hour talking about how brilliant physicists could get it so wrong in their thought process toward science. He calls it pathological science. Peter even comments on it (listen starting around 2:08 but I think might be closer to 2:12), right after Taubes says there is absolutely no way to reconcile the standard model with his hormone regulation hypothesis and that it isn't even wrong, he says that 99.9% of the time someone says that they are a quack. Not long after, he says the problem with pathological science is they establish what they are seeing based on premature data, go public, and then dig in their heels at that point — he's saying it about an adversary but could easily be said about himself. Its like he's almost self aware, but he's so sure of his view that his framework for understanding bad science doesn't apply to him. I think there's a bit of a motte and bailey that happens with Taubes proponency, where one can say there is this underappreciated role of insulin in obesity, vs the hard line he takes which is the standard energy balance model is "not even wrong", i.e., absolutely no role for a caloric surplus outside of his framework. So sure, demonstrating that a keto diet can be a game changer for a type 2 diabetic is great and Taubes has a hand bringing that to the mainstream. But that doesn't also affirm his hard, and hard to believe stance on the topic. I think what I mean is clear if you listen to the conversation about Doritos starting around 2:16, where Taubes suggests there is no hyperphasia without near-instant hormonal disregulation from adipose tissue in response to a dorito entering the mouth, and it has nothing to do with the team of food scientists that make sure they are additively delicious. Taubes could be 50-90% correct in his hypotheses, but he takes it beyond that point to the absurd and further that there is no reconciling his beliefs with competing hypotheses. I hate to be the 'just listen to the 200 minute podcast' guy, but it really is fascinating in the juxtaposition I mentioned and I'd be keen to know how much you agree or think I am off base if you did give it a listen.
  - omginternets 5y ago
    Ok, I suppose I agree, at least with your motte and bailey analysis. As for the first part, concerning "pathological" science, I'm less sure to the extent that one can reasonably (at least in theory) hold the following position: "the evidence that contradicts my model all suffers from the same methodological flaw, and that explains why they get the results that they get". It's true that this type of argumentation tends to cluster around an individual researcher and his groupies, which might raise some red flags, but isn't fallacious in and of itself. Whether or not that's what Taubes is actually doing is something of a judgement call, and I think I'm concluding a bit differently than you here, which is okay. :)
    - kuhewa 5y ago
      > "the evidence that contradicts my model all suffers from the same methodological flaw, and that explains why they get the results that they get" Note that Taubes says this constantly about a wide range of evidence that contradicts his model, including studies like the one lead by Hill that were funded by NuSI and personally signed off on by Taubes himself.
      - omginternets 5y ago
        Yeah, you can't just go throwing that around willy-nilly either... Hmm. I shall read more. It's been a while since I've looked at this stuff.
        kuhewa 5y ago
        I wanted to get a dose of his recent thinking outside of that podcast and this article and just listened to Joe Rogan pod #1267 (2019) when driving the past couple days, it is a debate with Taubes and Guyunet, a neurologist and former Taubesian that Taubes criticised on a former podcast and comes back. Again hate to be the 'listen to a 3 hour podcast' guy but it was really interesting and a better format for this subject than essays or back-and-forth in blog posts where people tend to talk past each other- transcript is available on Spotify so probably can find it elsewhere as well.
1_player 5y ago
I find it fascinating how this comment with no explanation whatsoever other than "he's self-deluded, look for yourself" is voted to the top.

by kuhewa 5y ago

omginternets 5y ago

Can you point to specific instances of self-delusion and poor scientific thinking? The goalpost for these accusations, as far as I'm concerned, is that you would need to show a pattern of ignoring contradicting evidence, which -- it should be noted -- is distinct from arguing that there are fundamental flaws in studies that prima facie invalidate his theory. In the latter case, it's just scholarly disagreement. It's late at night, and I'm not sure how to write this comment without it sounding like a loaded question, so please accept my assurances that it is not. It's been a while since I've read Taubes, and while the sugar/obesity debate is hardly settled, my impression was not one of delusion, nor of poor scientific reasoning. Certainly, it was a cut above the average quality of nutritional science research (though this is admittedly a very low bar).

kuhewa 5y ago
My comment was not just with regards to the facts of the subject but the way of thinking epistemologically about hypotheses in science, because he literally spends the first hour talking about how brilliant physicists could get it so wrong in their thought process toward science. He calls it pathological science. Peter even comments on it (listen starting around 2:08 but I think might be closer to 2:12), right after Taubes says there is absolutely no way to reconcile the standard model with his hormone regulation hypothesis and that it isn't even wrong, he says that 99.9% of the time someone says that they are a quack. Not long after, he says the problem with pathological science is they establish what they are seeing based on premature data, go public, and then dig in their heels at that point — he's saying it about an adversary but could easily be said about himself. Its like he's almost self aware, but he's so sure of his view that his framework for understanding bad science doesn't apply to him. I think there's a bit of a motte and bailey that happens with Taubes proponency, where one can say there is this underappreciated role of insulin in obesity, vs the hard line he takes which is the standard energy balance model is "not even wrong", i.e., absolutely no role for a caloric surplus outside of his framework. So sure, demonstrating that a keto diet can be a game changer for a type 2 diabetic is great and Taubes has a hand bringing that to the mainstream. But that doesn't also affirm his hard, and hard to believe stance on the topic. I think what I mean is clear if you listen to the conversation about Doritos starting around 2:16, where Taubes suggests there is no hyperphasia without near-instant hormonal disregulation from adipose tissue in response to a dorito entering the mouth, and it has nothing to do with the team of food scientists that make sure they are additively delicious. Taubes could be 50-90% correct in his hypotheses, but he takes it beyond that point to the absurd and further that there is no reconciling his beliefs with competing hypotheses. I hate to be the 'just listen to the 200 minute podcast' guy, but it really is fascinating in the juxtaposition I mentioned and I'd be keen to know how much you agree or think I am off base if you did give it a listen.
- omginternets 5y ago
  Ok, I suppose I agree, at least with your motte and bailey analysis. As for the first part, concerning "pathological" science, I'm less sure to the extent that one can reasonably (at least in theory) hold the following position: "the evidence that contradicts my model all suffers from the same methodological flaw, and that explains why they get the results that they get". It's true that this type of argumentation tends to cluster around an individual researcher and his groupies, which might raise some red flags, but isn't fallacious in and of itself. Whether or not that's what Taubes is actually doing is something of a judgement call, and I think I'm concluding a bit differently than you here, which is okay. :)
  - kuhewa 5y ago
    > "the evidence that contradicts my model all suffers from the same methodological flaw, and that explains why they get the results that they get" Note that Taubes says this constantly about a wide range of evidence that contradicts his model, including studies like the one lead by Hill that were funded by NuSI and personally signed off on by Taubes himself.
    - omginternets 5y ago
      Yeah, you can't just go throwing that around willy-nilly either... Hmm. I shall read more. It's been a while since I've looked at this stuff.
      - kuhewa 5y ago
        I wanted to get a dose of his recent thinking outside of that podcast and this article and just listened to Joe Rogan pod #1267 (2019) when driving the past couple days, it is a debate with Taubes and Guyunet, a neurologist and former Taubesian that Taubes criticised on a former podcast and comes back. Again hate to be the 'listen to a 3 hour podcast' guy but it was really interesting and a better format for this subject than essays or back-and-forth in blog posts where people tend to talk past each other- transcript is available on Spotify so probably can find it elsewhere as well.

1_player 5y ago

I find it fascinating how this comment with no explanation whatsoever other than "he's self-deluded, look for yourself" is voted to the top.