Politicization of research funding is largely a result of changes in public management culture over decades. Governments are no longer willing to fund research just because people in the field find it interesting. They expect that the research should somehow benefit the society. With such expectations, funding agencies now require political justifications for every single grant.
Andrew Dalkesays:
The first paper uses the presence of terms like “gender” to identify if an abstract is politicized, and more specifically, “associated with left-wing political causes”. If Stephen Colbert is right and “reality has a left wing bias”, then shouldn’t we expect papers to use reality-based terms more often?
For example, the term “global warming” is politicized. That doesn’t mean any increase in its use is uncalled for.
The paper starts with an explanation of Feynman’s cargo-cult. By using “gender” as an identifier, the paper asserts without proof that gender studies is essentially politicized, and by implication, a cargo-cult.
We know some terms are politicized because we know the Trump administration wanted to prohibit the CDC budget process from using the terms “entitlement,” “diversity,” “transgender,” “fetus,” “evidence-based,” and “science-based” – https://www.ucsusa.org/resources/attacks-on-science/ban-scientific-words-cdc-causes-uproar . This suggests that a *decrease* in the use of certain terms may indicate increasing politicization. At the very least, it seems like relevant background the paper should have mentioned.
I assert without proof that this paper is a cargo-cult use of natural language processing. In addition to the problems I listed above, the author ‘s bio show no training or other expertise in this sort of analysis, and it’s published in by a right-of-center think tank which funds people “to study incidents or trends of bias against conservatives” (quoting https://www.influencewatch.org/non-profit/center-for-study-of-partisanship-and-ideology-cspi/ ).
Figure 2 B suggests “not much”. Note that the intervention itself will cause some stress (it involves invasive surgery) so you would expect the young mice to suffer somewhat.
Maynard+Handleysays:
Is that “immediate outcome” or “longterm lifespan”?
Presumably the baseline model is
– young bodies produce some magic chemical that is of value to old bodies but no longer produced by them.
The obvious two questions are
– does taking it out of the young body reduce the lifetime (or otherwise harm the old age) of the young body
– what is this magic stuff? Can it be made in a factory? AND
– are there adverse consequences of having more [or less] magic stuff when you’re younger (eg does it create a live-fast-die-young sort of life pattern)?
One can obviously write the dystopian movies but the more interesting question to my mind is what exactly is going on here. Versions of this sort of thing have been claimed since at least the early 1900s with Serge Voronoff, following up on Charles-Édouard Brown-Séquard, and it’s still unclear to me how much of it is more than just fancy sympathetic magic (ie analogous to “heart shaped leaves help the heart”; or “rhino horn helps the penis”).
Parabiosis for rejuvenation has been reproduced independently. It is a robust scientific finding.
The effect is not well understood but scientists do not believe that the young mice produce something magical. The weight of the evidence is that the intervention dilutes some undesirable components or, alternatively, normalizes the body to an healthier state. Clinical trials are upcoming.
The therapeutic end goal is not to pump the blood of kids into the elderly. That would be highly impractical. They do these experiments in animals to gain some additional knowledge.
It is too soon to know what a therapy for human beings would look like but I imagine that they might hook you up to a machine and normalize the content of your blood.
Whether it would work, we do not know yet.
Maynard+Handleysays:
Dilution is interesting; the reverse of the model I suggested — too much X rather than not enough Y.
So one idea for the end-point is a sort of generalized dialysis?
I worry that so much of this stuff is fumbling in the dark, that we can maybe make the muscle work better, yes, but the lungs still fall apart. Or the brain.
It feels like eg cancer treatment in 1870, before we even know how mutations work, let alone have the unified paradigm of today within which we can understand and design treatment options.
Of course I don’t have a better investigative paradigm…
So one idea for the end-point is a sort of generalized dialysis?
Yes. I expect that there are clinical trials in this direction right now, although maybe the pandemic is slowing things down.
I worry that so much of this stuff is fumbling in the dark, that we can maybe make the muscle work better, yes, but the lungs still fall apart. Or the brain. It feels like eg cancer treatment in 1870, before we even know how mutations work, let alone have the unified paradigm of today within which we can understand and design treatment options.
We definitively do not know enough but it is unlikely that we will stumble on a conceptual breakthrough. We need experiments.
I’m a bit dubious about the methodology of the first study. The authors claim that usage of certain terms is indicative of a left-leaning bias.
A proposal to, for example “investigate gender income disparity” would contain the trigger-words for the authors to consider it left-leaning, regardless of the study’s result. An alternative interpretation to the authors’ interpretation could be that: as the authors’ chosen trigger words have become more widely discussed, more proposals are submitted to do _real_ science, to determine whether a popular growing theory has any legitimate truth to it.
Moreover, even if the authors’ claim is true, and diversity-studying proposals are being submitted by primarily left-leaning groups, I would think that the solution would be for right-leaning groups to also get funding for research on the same topics, and present a contrary viewpoint. Importantly, if this solution were implemented, the authors’ methodology would simply see it as more evidence of left-leaning bias! In fact, if the authors’ methodology is to be believed, the only way to get rid of left-leaning bias is to censor questions about diversity altogether, and prevent _anyone_ from talking about it.
Separately, I wasn’t able to find any explicit claims that the authors make about the word-embedding-based distance between abstracts. They have it listed in their methodology, but I wasn’t able to find any conclusions that they made based on it.
Joern Engelsays:
I wonder if the first study is an explicit troll. Two quotes from the introduction:
“during a commencement speech at Caltech in 1974.”
“first expressed it almost 60 years ago.”
Politicization of research funding is largely a result of changes in public management culture over decades. Governments are no longer willing to fund research just because people in the field find it interesting. They expect that the research should somehow benefit the society. With such expectations, funding agencies now require political justifications for every single grant.
The first paper uses the presence of terms like “gender” to identify if an abstract is politicized, and more specifically, “associated with left-wing political causes”. If Stephen Colbert is right and “reality has a left wing bias”, then shouldn’t we expect papers to use reality-based terms more often?
For example, the term “global warming” is politicized. That doesn’t mean any increase in its use is uncalled for.
The paper starts with an explanation of Feynman’s cargo-cult. By using “gender” as an identifier, the paper asserts without proof that gender studies is essentially politicized, and by implication, a cargo-cult.
We know some terms are politicized because we know the Trump administration wanted to prohibit the CDC budget process from using the terms “entitlement,” “diversity,” “transgender,” “fetus,” “evidence-based,” and “science-based” – https://www.ucsusa.org/resources/attacks-on-science/ban-scientific-words-cdc-causes-uproar . This suggests that a *decrease* in the use of certain terms may indicate increasing politicization. At the very least, it seems like relevant background the paper should have mentioned.
I assert without proof that this paper is a cargo-cult use of natural language processing. In addition to the problems I listed above, the author ‘s bio show no training or other expertise in this sort of analysis, and it’s published in by a right-of-center think tank which funds people “to study incidents or trends of bias against conservatives” (quoting https://www.influencewatch.org/non-profit/center-for-study-of-partisanship-and-ideology-cspi/ ).
What happened to the young mice?
Figure 2 B suggests “not much”. Note that the intervention itself will cause some stress (it involves invasive surgery) so you would expect the young mice to suffer somewhat.
Is that “immediate outcome” or “longterm lifespan”?
Presumably the baseline model is
– young bodies produce some magic chemical that is of value to old bodies but no longer produced by them.
The obvious two questions are
– does taking it out of the young body reduce the lifetime (or otherwise harm the old age) of the young body
– what is this magic stuff? Can it be made in a factory? AND
– are there adverse consequences of having more [or less] magic stuff when you’re younger (eg does it create a live-fast-die-young sort of life pattern)?
One can obviously write the dystopian movies but the more interesting question to my mind is what exactly is going on here. Versions of this sort of thing have been claimed since at least the early 1900s with Serge Voronoff, following up on Charles-Édouard Brown-Séquard, and it’s still unclear to me how much of it is more than just fancy sympathetic magic (ie analogous to “heart shaped leaves help the heart”; or “rhino horn helps the penis”).
Parabiosis for rejuvenation has been reproduced independently. It is a robust scientific finding.
The effect is not well understood but scientists do not believe that the young mice produce something magical. The weight of the evidence is that the intervention dilutes some undesirable components or, alternatively, normalizes the body to an healthier state. Clinical trials are upcoming.
The therapeutic end goal is not to pump the blood of kids into the elderly. That would be highly impractical. They do these experiments in animals to gain some additional knowledge.
It is too soon to know what a therapy for human beings would look like but I imagine that they might hook you up to a machine and normalize the content of your blood.
Whether it would work, we do not know yet.
Dilution is interesting; the reverse of the model I suggested — too much X rather than not enough Y.
So one idea for the end-point is a sort of generalized dialysis?
I worry that so much of this stuff is fumbling in the dark, that we can maybe make the muscle work better, yes, but the lungs still fall apart. Or the brain.
It feels like eg cancer treatment in 1870, before we even know how mutations work, let alone have the unified paradigm of today within which we can understand and design treatment options.
Of course I don’t have a better investigative paradigm…
So one idea for the end-point is a sort of generalized dialysis?
Yes. I expect that there are clinical trials in this direction right now, although maybe the pandemic is slowing things down.
I worry that so much of this stuff is fumbling in the dark, that we can maybe make the muscle work better, yes, but the lungs still fall apart. Or the brain. It feels like eg cancer treatment in 1870, before we even know how mutations work, let alone have the unified paradigm of today within which we can understand and design treatment options.
We definitively do not know enough but it is unlikely that we will stumble on a conceptual breakthrough. We need experiments.
See https://www.aging-us.com/article/103418/text
I’m a bit dubious about the methodology of the first study. The authors claim that usage of certain terms is indicative of a left-leaning bias.
A proposal to, for example “investigate gender income disparity” would contain the trigger-words for the authors to consider it left-leaning, regardless of the study’s result. An alternative interpretation to the authors’ interpretation could be that: as the authors’ chosen trigger words have become more widely discussed, more proposals are submitted to do _real_ science, to determine whether a popular growing theory has any legitimate truth to it.
Moreover, even if the authors’ claim is true, and diversity-studying proposals are being submitted by primarily left-leaning groups, I would think that the solution would be for right-leaning groups to also get funding for research on the same topics, and present a contrary viewpoint. Importantly, if this solution were implemented, the authors’ methodology would simply see it as more evidence of left-leaning bias! In fact, if the authors’ methodology is to be believed, the only way to get rid of left-leaning bias is to censor questions about diversity altogether, and prevent _anyone_ from talking about it.
Separately, I wasn’t able to find any explicit claims that the authors make about the word-embedding-based distance between abstracts. They have it listed in their methodology, but I wasn’t able to find any conclusions that they made based on it.
I wonder if the first study is an explicit troll. Two quotes from the introduction:
“during a commencement speech at Caltech in 1974.”
“first expressed it almost 60 years ago.”
It should have been ‘almost 50 years ago’.
The question is whether it was a typo, a math mistake or a deliberate tell.