If Mort Divine ruled the world

You're trying to project rationality onto people whose tempers are flaring and all of whom are already on edge. Nothing about the situation is "normal," and people won't act as expected.

I also said I don't think it makes sense if we're talking instincts.

You're actually the one choosing to believe something based on--I can only assume--some predisposition you have to whatever occurred. And unless you have forensic training and experience analyzing body movements in violent situations, what expertise can you claim?

I didn't even know about this Harris incident until I saw it here and I've already implied several times that it's plausible the Canuck was aiming for the woman's phone and hit her face because both potential targets are inches apart. Harris however, going by the footage, didn't do anything to indicate he was aiming at waist level and the arc of his swing was aimed high and what a shock, he landed high.

Dunno about you but I've been in/seen enough fights to understand something as basic as how body language telegraphs intent. The Canuck telegraphed his kick by pivoting his hip and doing a slight bounce with his feet. Harris telegraphed his head-strike with the angle of his shoulder and elbow and just now I've also noticed that he's seemingly looking at the man's head area as he swings, if he was aiming for the pole at the man's waist he would have been looking downward in that direction because who the fuck strikes a target while not looking at it? Is he Bruce Lee? Or is he some kind of sociopath who looks at targets with only his eyes while he keeps his head level?

I personally just think you're desperate to have a black victim at the hands of the whitenats with no ambiguity or chance that the black guy was in the wrong.
 
And unless you have forensic training and experience analyzing body movements in violent situations, what expertise can you claim?

Dunno about you but I've been in/seen enough fights to understand something as basic as how body language telegraphs intent.

:rofl: Seems super sound.

I personally just think you're desperate to have a black victim at the hands of the whitenats with no ambiguity or chance that the black guy was in the wrong.

Thanks for the psychoanalysis.

My entire point has been that there is ambiguity, but fine.
 
I've read a lot about this "grievance studies" hoax. They published some crazy shit, but overall I don't think they proved much of anything. This piece from Slate sums up my view:

https://slate.com/technology/2018/10/grievance-studies-hoax-not-academic-scandal.html

The following paragraphs are particularly lucid:

The dog-rape study is supposed to have resulted from nearly 1,000 hours of observation at three dog parks in southeast Portland. The dildo paper pretends to draw from multihour interviews with 13 men—eight straight, two bisexual, three gay—about their sexual behaviors. And the breastaurant research claims to have its basis in a two-year-long project carried out in northern Florida, involving men whose educational backgrounds, ages, and marital statuses were duly recorded and reported.

How absurd was it for such work to get an airing? It may sound silly to investigate the rates at which dog owners intervene in public humping incidents, but that doesn’t mean it’s a total waste of time (as psychologist Daniel Lakens pointed out on Twitter). If the findings had been real, they would have some value irrespective of the pablum that surrounds them in the paper’s introduction and discussion sections. Indeed, one can point to lots of silly-sounding published data from many other fields of study, including strictly scientific ones. Are those emblematic of “corruption” too?

It’s true that Pluckrose, Lindsay, and Boghossian tricked some journals into putting out made-up data, but this says nothing whatsoever about the fields they chose to target. One could have run this sting on almost any empirical discipline and returned the same result. We know from long experience that expert peer review offers close to no protection against outright data fraud, whether in the field of gender studies or cancer research, psychology or plant biology, crystallography or condensed matter physics. Even shoddy paste-up jobs with duplicated images and other slacker fakes have made their way to print and helped establish researchers’ careers. So what if these hoaxers did the same for fun? These examples haven’t hoodwinked anyone with sophistry or satire but with a simple fabrication of results.

Pluckrose, Lindsay, and Boghossian employed this made-up-data method for five of their 21 papers, and three of those were accepted for publication—yielding a hoax-success rate of 60 percent. When they wrote up papers without this added layer of deception, just four of 16 were accepted.

That's Ein's entire driving philosophy of life.

th
 
The data is irrelevant. It's the arguments that are the point. Data can be created, legitimately or otherwise, to support any number of hypotheses. p-hacking is where one collects data, finds a significant result from them, then configures a hypothis(eses) that seem to explain the data. These authors didn't p-hack, they created absurd arguments that sounded good to the journals, with some "data" to go with it. Argument-hacking? The data was a bonus. Do you think the reason they were rejected from sociological journals was the data?
 
  • Like
Reactions: HamburgerBoy
Do you think the reason they were rejected from sociological journals was the data?

mmmmm, no. That's the not the author's point though. His point is that hoaxes like this are always aimed at journals on the basis of argumentation. They never aim at scientific journals which might promote plausible arguments based on false data. There's an implicit assumption that objectionable arguments based on legitimate data are somehow worse than non-objectionable arguments based on false data.

Additionally, there's an operating flaw in their methodology that equates mimicry with comprehension, but I'm tired and don't have time to go into it now. Long story short, it's questionable whether they actually "know" they've created faulty arguments if all they're doing is imitating a discourse.
 
mmmmm, no. That's the not the author's point though. His point is that hoaxes like this are always aimed at journals on the basis of argumentation. They never aim at scientific journals which might promote plausible arguments based on false data. There's an implicit assumption that objectionable arguments based on legitimate data are somehow worse than non-objectionable arguments based on false data.

Additionally, there's an operating flaw in their methodology that equates mimicry with comprehension, but I'm tired and don't have time to go into it now. Long story short, it's questionable whether they actually "know" they've created faulty arguments if all they're doing is imitating a discourse.

The point is that the author's point is retarded, and probably indicative of someone that has never published a paper in a remotely-decent scientific journal before. Even in social sciences there's an expectation of some statistical analysis of data, some formation of a testable model, etc. As Dak said, the issue wasn't the forgery of data, it was the fact that these papers were 90% conjecture based on 10% data.

I don't know what you mean by "objectionable" vs "non-objectionable", but people perform hoaxes against journals in the hard sciences as well. The controversy of such hoaxes isn't just forgery of data, it's primarily that the arguments themselves are bullshit and get accepted either because the reviewer dogmatically believes it to be true, or because the reviewer cannot understand the paper.

d76.jpg
 
  • Like
Reactions: Dak
The point is that the author's point is retarded, and probably indicative of someone that has never published a paper in a remotely-decent scientific journal before. Even in social sciences there's an expectation of some statistical analysis of data, some formation of a testable model, etc. As Dak said, the issue wasn't the forgery of data, it was the fact that these papers were 90% conjecture based on 10% data.

I don't know what you mean by "objectionable" vs "non-objectionable", but people perform hoaxes against journals in the hard sciences as well. The controversy of such hoaxes isn't just forgery of data, it's primarily that the arguments themselves are bullshit and get accepted either because the reviewer dogmatically believes it to be true, or because the reviewer cannot understand the paper.

I'm unaware of hoaxes pulled against hard science journals. I'd like to read about them. I don't doubt someone has tried. Part of the author's point, however, is that those hoaxes don't get the kind of publicity that the Areo and Sokal hoaxes do. The reason they get such publicity is that they appeal to a political motivation among even non-academics who simply despise anything remotely gender-related. They claim to be apolitical, but their work is making a splash precisely because it is political. It's exploding on places like reddit and 4chan, while academics are mostly rolling their eyes.

I'm not trying to shrug the episode off. The Areo authors really did manage to publish some outlandish stuff, and it's worth pausing over the fact that four out of sixteen papers with no data collection got published.

My primary critique of the hoax, which I didn't have time to go into last night, has to do with its methodology--which the authors describe as "reflexive ethnography." This basically means immersing yourself into a culture or community until you can effectively communicate in its discourse, i.e. you figure out what the buzzwords are and how to put certain ideas together. In other words, you can pass yourself off as an insider.

The problem with this is that it's basically the same premise as a computer in a Turing test. A computer might convincingly imitate human language/behavior, but that doesn't mean computer scientists are convinced that it understands what it's saying. John Searle's Chinese room thought experiment basically suggests that a computer might communicate appropriately and effectively without having any knowledge of the semantic content of its language.

This raises a problem for the Areo authors claiming to "know they've made things up." If they don't truly understand the content of the things they've said, then they can't actually know what the meaningful effect of their "intentionally broken" arguments are. Even hard sciences like theoretical physics and abstract mathematics rely on innovative recombinations that surprise readers. Good arguments should simultaneously make sense and surprise us. If they didn't surprise us, then there would be no need to make the argument. So all arguments involve some element of creativity.

Admittedly, the Areo authors managed to publish some, ahem, excessively creative pieces. But I don't think they've actually proven, or even suggested, that the journals in which they've published are awash in such arguments. Rather, they've demonstrated the degree to which editors/readers are willing to bend to allow creative arguments. In some cases, it may be too far; but I'm willing to bet the majority of arguments accepted in these journals don't fall into this outlier category (I don't read any of them, so I don't know).
 
Last edited:
Additionally, there's an operating flaw in their methodology that equates mimicry with comprehension, but I'm tired and don't have time to go into it now. Long story short, it's questionable whether they actually "know" they've created faulty arguments if all they're doing is imitating a discourse.

Are they creating faulty arguments? We can only assume no. It's hard to create a faulty argument in "disciplines" where the discourse isn't even wrong. We can even make the assumption that they were engaging in mimicry, but this just makes it more amusing.

I'm not trying to shrug the episode off. The Areo authors really did manage to publish some outlandish stuff, and it's worth pausing over the fact that four out of sixteen papers with no data collection got published.

The data is, in my estimation, the only funny part of the whole thing. If they merely recombined a lot of the jargon together and had it accepted that would simply be Sokal all over again. Just showing that these fields aren't any better twenty some years on. Instead, they add in data to fields that, depending on the academic, reject data in general as irrelevant anyway. These are pretend disciplines who were likely paradoxically giddy that through these articles they were "sciencing".

My primary critique of the hoax, which I didn't have time to go into last night, has to do with its methodology--which the authors describe as "reflexive ethnography." This basically means immersing yourself into a culture or community until you can effectively communicate in its discourse, i.e. you figure out what the buzzwords are and how to put certain ideas together. In other words, you can pass yourself off as an insider.

The problem with this is that it's basically the same premise as a computer in a Turing test. A computer might convincingly imitate human language/behavior, but that doesn't mean computer scientists are convinced that it understands what it's saying. John Searle's Chinese room thought experiment basically suggests that a computer might communicate appropriately and effectively without having any knowledge of the semantic content of its language.

This raises a problem for the Areo authors claiming to "know they've made things up." If they don't truly understand the content of the things they've said, then they can't actually know what the meaningful effect of their "intentionally broken" arguments are. Even hard sciences like theoretical physics and abstract mathematics rely on innovative recombinations that surprise readers. Good arguments should simultaneously make sense and surprise us. If they didn't surprise us, then there would be no need to make the argument. So all arguments involve some element of creativity.

Admittedly, the Areo authors managed to publish some, ahem, excessively creative pieces. But I don't think they've actually proven, or even suggested, that the journals in which they've published are awash in such arguments. Rather, they've demonstrated the degree to which editors/readers are willing to bend to allow creative arguments. In some cases, it may be too far; but I'm willing to bet the majority of arguments accepted in these journals don't fall into this outlier category (I don't read any of them, so I don't know).

There is absolutely no comparison between hard sciences and grievance studies. Novel arguments in hard sciences lead to concrete products or to nothing. Novel arguments in grievance studies are just adding more and more words to the same basic scaffolding provided by Nietzsche and Marx, and substituting out the specific "oppressors" and the "victims" as needed, with no concrete connection required nor indeed available. It's purely conspiracy theorizing, but of the worst sort. This is the reason Mein Kampf was workable into this framework. At least Jews, and particularly Jewish bankers, were a concrete target for Hitler. "The patriarchy" is completely abstract: it's not even attached to men specifically. Just anything a "feminist" doesn't like is "patriarchy". Cite Judith Butler and now we're academicing!
 
Well, a lot to talk about. I don't have the luxury of time, so I'm not sure how often I'll be able to respond. This is all for now:

Are they creating faulty arguments? We can only assume no. It's hard to create a faulty argument in "disciplines" where the discourse isn't even wrong. We can even make the assumption that they were engaging in mimicry, but this just makes it more amusing.

Actually, it does, but not for the reason you think. It's because if their methodology is mimicry then it reinforces a key insight common to both Turing machines and gender performativity: i.e. that all communication is mimicry. The authors of this study don't actually know anything--or if they do, that knowledge doesn't make a difference.

The data is, in my estimation, the only funny part of the whole thing. If they merely recombined a lot of the jargon together and had it accepted that would simply be Sokal all over again. Just showing that these fields aren't any better twenty some years on. Instead, they add in data to fields that, depending on the academic, reject data in general as irrelevant anyway. These are pretend disciplines who were likely paradoxically giddy that through these articles they were "sciencing".

In that case, the exercise sounds not merely flawed but baroque. It contains distractions that actually clutter the central point, which is (presumably) that their arguments are absurd.

If the data is irrelevant--or better, hypothetically speaking, if the data existed--and the arguments are still sound based on the evidence at hand and the discourses in which they occur, then by what do we judge their absurdity? It seems to me that all we have to go by is a kind of... value judgment (*gasp*).

There is absolutely no comparison between hard sciences and grievance studies. Novel arguments in hard sciences lead to concrete products or to nothing.

I don't know what you mean by concrete objects, but it sounds presumptuous. Evolutionary theory hasn't produced any "concrete products" that I'm aware of; it simply provided a means of describing species development over time. That subsequent discoveries in genetics support evolution doesn't mean that evolutionary theory produced these discoveries; and furthermore, genetics is yet another theory. It's not as though science is working toward some all-explaining mechanism. It's theories all the way down (or up... not sure). Provided evolutionary theory is, in fact, true, it need not yield some yet-secret mechanism that we haven't discovered. It's not an ontology for species development, just like structuralism wasn't an ontology for what texts are; both are simply ways of describing the world based on evidence.

It's also worth pointing out that of the "grievance studies" disciplines these authors target, only one--gender and women's studies--have actual programs in many universities. The others--sexuality studies, fat studies, even feminist studies--don't actually have departments on all (or in some cases any) campuses. For the most part, they only vaguely describe theoretical discourses from which humanities scholars draw. In this respect, the target of the Areo piece is also admittedly vague, and really discernible only to those who already lump a lot of these disciplines together.

Novel arguments in grievance studies are just adding more and more words to the same basic scaffolding provided by Nietzsche and Marx, and substituting out the specific "oppressors" and the "victims" as needed, with no concrete connection required nor indeed available. It's purely conspiracy theorizing, but of the worst sort. This is the reason Mein Kampf was workable into this framework. At least Jews, and particularly Jewish bankers, were a concrete target for Hitler. "The patriarchy" is completely abstract: it's not even attached to men specifically. Just anything a "feminist" doesn't like is "patriarchy". Cite Judith Butler and now we're academicing!

This sounds less rational than antagonistic.

The problem here is that this doesn't boil down to logic and facts, much to some people's chagrin, but to what kind of research and analysis you find valuable. I think there's value in pursuing cultural contextualizing and historicizing, and that means that even things like "big data" warrant attention as potential objects of critique.

That doesn't mean that data is worthless, directionless, or even harmful necessarily. It just means that data is a tool. It's composed and designed with ends in mind. It doesn't just present itself. Data appears to those who look for it. It qualifies a perspective.
 
Last edited:
Actually, it does, but not for the reason you think. It's because if their methodology is mimicry then it reinforces a key insight common to both Turing machines and gender performativity: i.e. that all communication is mimicry. The authors of this study don't actually know anything--or if they do, that knowledge doesn't make a difference.

If all communication is mimicry then no one knows anything. It doesn't dig one out of the hole, and leaves one with nowhere to go. Again, just slapping words on scaffolding like so much paper mache. The core is hollow and it's dead.

In that case, the exercise sounds not merely flawed but baroque. It contains distractions that actually clutter the central point, which is (presumably) that their arguments are absurd.

If the data is irrelevant--or better, hypothetically speaking, if the data existed--and the arguments are still sound based on the evidence at hand and the discourses in which they occur, then by what do we judge their absurdity? It seems to me that all we have to go by is a kind of... value judgment (*gasp*).

But we are talking about discourses which, at some junctures, explicitly eschew logic even when they don't obfuscate and obscure meaning. Then we have first principles that are again, cut from conspiracy theory cloth. Where is there soundness? It's just buzzwords and some data about observing people doing stuff. "Men observed sitting with legs not tightly pressed together......PATRIARCHY CONFIRMED!". Academicing.

I don't know what you mean by concrete objects, but it sounds presumptuous. Evolutionary theory hasn't produced any "concrete products" that I'm aware of; it simply provided a means of describing species development over time. That subsequent discoveries in genetics support evolution doesn't mean that evolutionary theory produced these discoveries; and furthermore, genetics is yet another theory. It's not as though science is working toward some all-explaining mechanism. It's theories all the way down (or up... not sure). Provided evolutionary theory is, in fact, true, it need not yield some yet-secret mechanism that we haven't discovered. It's not an ontology for species development, just like structuralism wasn't an ontology for what texts are; both are simply ways of describing the world based on evidence.

Not being in the hard sciences I could be wrong about this, but doesn't evolutionary theory provide useful and relatively consistent predictions which assist in things like finding and identifying fossils and new species, as well as working together with other discoveries (like genetics) to explain biological changes across time? It hasn't given us the internet of course, but it's got some concrete utility.

Grievance studies give us concrete behaviors like screaming meltdowns in the streets and generally justifications for BPD and APD cognitions and behaviors, but it's not quite a unique product, and where's the positive outcomes? It attracts and multiplies the unhappy and broken. And doesn't give us any trinkets in the process either.

The problem here is that this doesn't boil down to logic and facts, much to some people's chagrin, but to what kind of research and analysis you find valuable. I think there's value in pursuing cultural contextualizing and historicizing, and that means that even things like "big data" warrant attention as potential objects of critique.

That doesn't mean that data is worthless, directionless, or even harmful necessarily. It just means that data is a tool. It's composed and designed with ends in mind. It doesn't just present itself. Data appears to those who look for it. It qualifies a perspective.

Data indeed supports a perspective, and everything isn't logic. But we're talking about disciplines which generally eschew data and logic. What's left? Conspiracy theories, psychiatric medications, and riots.
 
If all communication is mimicry then no one knows anything. It doesn't dig one out of the hole, and leaves one with nowhere to go. Again, just slapping words on scaffolding like so much paper mache. The core is hollow and it's dead.

It might be, or it might not be. That's beside the point.

The construction and expansion of knowledge has little to do with what individual minds know.

But we are talking about discourses which, at some junctures, explicitly eschew logic even when they don't obfuscate and obscure meaning. Then we have first principles that are again, cut from conspiracy theory cloth. Where is there soundness? It's just buzzwords and some data about observing people doing stuff. "Men observed sitting with legs not tightly pressed together......PATRIARCHY CONFIRMED!". Academicing.

Logic isn't pre-discursive, meaning it's a valid target of discourse. To employ logic, we have to agree on a ground from which to proceed. That ground is always subject to critique.

Not being in the hard sciences I could be wrong about this, but doesn't evolutionary theory provide useful and relatively consistent predictions which assist in things like finding and identifying fossils and new species, as well as working together with other discoveries (like genetics) to explain biological changes across time? It hasn't given us the internet of course, but it's got some concrete utility.

It sounds like what you're saying is that evolutionary theory produces findings that reaffirm evolutionary theory (i.e. biological changes across time).

Grievance studies give us concrete behaviors like screaming meltdowns in the streets and generally justifications for BPD and APD cognitions and behaviors, but it's not quite a unique product, and where's the positive outcomes? It attracts and multiplies the unhappy and broken. And doesn't give us any trinkets in the process either.

You're being pernicious, and I'm trying to avoid this degrading into a shouting match.

You see meltdowns in the streets and think that "grievance studies" have somehow contributed to the unhappiness in the world. I'd say that they've given those who were already unhappy a vocabulary through which to cognize their unhappiness. And from there, it's a matter of trying to make change through social and political discourse. Screaming in the streets isn't the best example, but that's a common example. It's just the example you always hear about.

Put another way, this hoax confirms what you already believed about these discourses, and so you ignore the glaring omissions. It attempts to dismiss whole fields (and then a whole philosophical paradigm) with a handful of singularities that don't even meet the evidentiary standards of a representative anecdote. It's the reduction of entire fields to a few embarrassing oversights.

Data indeed supports a perspective, and everything isn't logic. But we're talking about disciplines which generally eschew data and logic. What's left? Conspiracy theories, psychiatric medications, and riots.

You just said everything isn't logic and then implied that nothing other than logic constitutes valid argumentation. But there are other valid forms of argument; and those are what's left.
 
It might be, or it might not be. That's beside the point.

The construction and expansion of knowledge has little to do with what individual minds know.

Logic isn't pre-discursive, meaning it's a valid target of discourse. To employ logic, we have to agree on a ground from which to proceed. That ground is always subject to critique.

We do have to agree on the ground from which to proceed, but if logic, at this point, is a "target" of discourse, there is no discourse. There are just ravings. We can even have shared first principles and then any discourse just looks like so much spilled water.


It sounds like what you're saying is that evolutionary theory produces findings that reaffirm evolutionary theory (i.e. biological changes across time).

So we don't find changes? Or are you saying there are other theories which explain changes?

You're being pernicious, and I'm trying to avoid this degrading into a shouting match.

You see meltdowns in the streets and think that "grievance studies" have somehow contributed to the unhappiness in the world. I'd say that they've given those who were already unhappy a vocabulary through which to cognize their unhappiness. And from there, it's a matter of trying to make change through social and political discourse. Screaming in the streets isn't the best example, but that's a common example. It's just the example you always hear about.

Put another way, this hoax confirms what you already believed about these discourses, and so you ignore the glaring omissions. It attempts to dismiss whole fields (and then a whole philosophical paradigm) with a handful of singularities that don't even meet the evidentiary standards of a representative anecdote. It's the reduction of entire fields to a few embarrassing oversights.

I'll agree it gives a vocabulary, but not a helpful one either on an individual or societal level. I do dismiss these/this "philosophical paradigms". They might have a point or two (or more, who knows?), but we can't know because they refuse to be baptized by the fire of critical analysis (which, of course, is why you find "critical" and "analysis" within the names of related domains). You contend the Sokal Squared papers are a few embarrassing oversights. I agree. It's only a few studies, they are embarrassing, and the "sight" is bad. That's what doing nothing but kicking up the proverbial dust does. It blocks vision. And it's systemic.

You just said everything isn't logic and then implied that nothing other than logic constitutes valid argumentation. But there are other valid forms of argument; and those are what's left.

Everything isn't logic. Not everything is argumentation. Not all argumentation is logical. That doesn't mean that argumentation outside of a logical format is valid. Of course, you can question the utility? of "validity" but then you're outside of (verbal) discourse and left with war. War based on vapid conspiracy theories.
 
We do have to agree on the ground from which to proceed, but if logic, at this point, is a "target" of discourse, there is no discourse. There are just ravings. We can even have shared first principles and then any discourse just looks like so much spilled water.

Simply incorrect. Discourse can involve speculation, extrapolation, and analogical reasoning. All of these are valid methods of discourse.

So we don't find changes? Or are you saying there are other theories which explain changes?

I'm not sure what you mean. I'm not saying there are other theories (I mean, there could be); but evolutionary theory hasn't produced new fossils or biological changes. Those fossils and/or biological changes were/are already there. All evolutionary theory provides is the explanation. So again, I'm not sure what you mean.

I'll agree it gives a vocabulary, but not a helpful one either on an individual or societal level. I do dismiss these/this "philosophical paradigms". They might have a point or two (or more, who knows?), but we can't know because they refuse to be baptized by the fire of critical analysis (which, of course, is why you find "critical" and "analysis" within the names of related domains). You contend the Sokal Squared papers are a few embarrassing oversights. I agree. It's only a few studies, they are embarrassing, and the "sight" is bad. That's what doing nothing but kicking up the proverbial dust does. It blocks vision. And it's systemic.

I still don't follow. Now you're saying the Areo hoax is itself a grievance study? If grievance studies are invalid, then how is the hoax itself valid?

From another perspective, if it revealed only a few embarrassing examples, then the implication is that most articles published in critical journals aren't embarrassing. They're not "kicking up the proverbial dust." They're dedicated studies of contemporary culture. I think it's that very premise that you have a problem with.

Everything isn't logic. Not everything is argumentation. Not all argumentation is logical. That doesn't mean that argumentation outside of a logical format is valid.

Yes, it does. As I already said above.
 
Last edited: