This is the fourth in a series of diaries in which I revisit the most common maddeningly witless memes with which I've been confronted on dKos over the past decade.
The first entry addressed the stupidity and wrongness of people who believe that American can "win" the global economic competition. It doesn't really matter why they believe it, but god they're boring.
The second entry addressed the excruciating tedium of trying to explain to the advocates of MMT ... anything. It can't be done, because once you've engaged their MMT defense mechanisms, they stop processing language, and you simply cannot get answers that are responsive to your arguments. It's weird the first few times you encounter it, but then the novelty passes and it's just boring.
The third entry addressed the deeply stupid and unforgivably anti-democractic insistence by certain people that one or another candidate for office is unelectable. (What, one wonders, do such folk think will happen should both major parties nominate an "unelectable" candidate? Will a rift open in the time-space continuum and George Washington step forth to take office on a write-in wave?) When combined with any argument that one is therefore obliged not to support that candidate, the argument transcends mere stupidity and wrongness, and becomes actively Evil.
All three of these BSW memes (and there are more I could have diaried) are rooted in broken epistemology. Too many folk -- including well-meaning, progressive, likable folk -- don't know what knowledge is, or where it comes from. Even those who have appointed themselves the guardians of empiricism on dKos have critical flaws in their understanding of the principles of scientific evidence, inquiry, and "proof" -- especially with respect to statistics and the proof of the negative.
This really gets my goat. I am all about epistemology. Just search the dailykos comments for "epistemology" and "epistemological", and look at how many of the results were written by me. And so, once more over the orange naughty bits, dear friends. Time is running short.
So ... what is epistemology, anyway?
Just as "Biology" refers to both the collective processes of life and the study of the processes of life, "Epistemology" refers to both:
A. The processes by which we come to know things
and
B. The philosophical investigation of such processes.
When a person's epistemology is broken, it means the person "comes to know" things -- facts, or principles, or concepts -- that are false. Often, we describe broken epistemologies in terms of rhetorical fallacies, such as post hoc, propter hoc, or ad hominem.
On dKos people get "ad hominem" wrong so often and consistently that I am going to explain it to y'all, one last fucking time: "ad hominem" does not refer generally to any and all personal insults one might level at a rhetorical opponents.
Insult:
Your argument is fucking stupid, and therefore you are fucking stupid.
vs.
ad hominem:
You are fucking stupid, and therefore your argument is fucking stupid.
.
The insult is pretty well-argued, though it leaves out the possibility that the person making the stupid argument knows full well that the argument is stupid. By contrast, the
ad hominem is nonsense. The quality of any argument always stands apart from the qualities of the person making the argument. Truth is just as true when spoken by a liar, a fool, or a lunatic, as when spoken by a saintly genius.
The odds are good that a kossack accused of arguing ad hominem is instead simply being insulting. The latter is rude. The former is theoretically Boring, Stupid and Wrong.
However, there can be more strength to ad hominem argument than traditional epistemology allows. Most commonly, ad hominem on dKos happens after a person posts a link (or a quote) from a highly disreputable source, and subsequent commenters dismiss the substance of the comment based on the disreputability of the source. There is a "fair enough" quality to such ad hominem, in that any asserted facts in the source should be doubted until verified. Still, it is Boring, Stupid, and Wrong to dismiss arguments based solely on the source, and it can be Boring, Stupid, and Wrong to dismiss out of hand asserted facts from questionable sources. Several weeks ago, a diarist published an excerpt from a post on some website known to be deep in MRA or PUA or some such. The diarist's focus was on a verbatim quotation from another source. Critics of the diary dismissed the significance of the verbatim quotation because the diarist sourced it from its citation on a blog devoted to jackassery. And that is Boring, Stupid and Wrong.
More acutely Boring, Stupid and Wrong is when debates go awry at the point where one's opponent, crippled by some unknown array of epistemological defects, simply can not or will not acknowledge error.
When user coffeetalk was bojo'd, she had her defenders, but I saw not one of them attempt to rebut what was, for many people, the principle indictment: That when she was clearly in the wrong on the facts, when her wrongness had been well-documented with quotations and citations, she stuck her fingers in her keyboard's ears and doubled down on her wrongness. She couldn't even bring herself to just quietly fuck off out of a thread, but would just keep posting the same debunked misinformation in comment after comment. That is exquisitely Boring, Stupid and Wrong.
In my BSW: Electability diary, I asserted that George W. Bush had been "widely believed to be unelectable". kossack johnnywurster -- one of the sharp minority who didn't think coffeetalk was an odious troll -- replied, "Bush was eminently electable in 2000. That was widely seen as one of his strengths, in fact." And I replied,
I did not say that Bush was not eminently electable, I said he was, "widely believed to be unelectable," which is a true statement: "widely believed" is not equivalent to "universally believed", or even "generally believed".
To which JW responded with
Like I said, you're just wrong. Bush was thought of as electable.
Enjoy your bitterness!
Aside from the fact that JW is (as briefly evidenced by that tag line, but abundantly evidenced all over dailyKos) a primo fucking asshole,
he is wrong; but
nothing is ever going to get him to admit that. (It is not an accident that JW was supportive of CT. Both are weak bitter tea, steeped in the arrogant anti-humanist and intellectually adolescent epistemological tradition that I have elsewhere dubbed "legalism".) We can neither diagnose nor treat his epistemological disease, because we can never get the patient to engage forthrightly and ingenuously. Does JW assert that every single one of the hundred or so million Americans who had given it any thought, had concluded that GWB was "eminently electable"? His implicit argument is this:
Premise A.
some people -- and in particular, people to whom he happened to be paying attention -- blathered that GWB was eminently electable
Premise B.
other people (or in some cases the same people) -- again, whose opinions caught the eye or ear of JW -- blathered that this electability was a big selling point amongst the movers and the shakers of the GOP,
and therefore.
Conclusion X: There did not exist a very large number of people who thought GWB was unelectable.
Do I really need to continue the analysis, or is it plain enough that JW's epistemological dory is loaded to the gunwales with excrement?
Accommodating such epistemological vapidity is a serious suck on my soul, even when it is not accompanied by smug dismissals of my (or anybody else's) existential sorrows.
And sadly, I see it even with individuals with whom I generally agree, such as SkepticalRaptor and Lenny Flank. These two are the binary stars around which an entire community of dailyKos Defenders of Science revolves. As I consider myself also a stalwart defender of science (and reason, and materialism, and logical positivism, and whatever), it really ticks me off when these two guys go off the rails. SR and LF are fundamentalists -- they are so fully committed to the Platonic Ideal of the scientific endeavor, that they cannot accommodate the scarred, marred, and muchly flawed reality.
According to John Ioannidis, a highly respected mathematician and critic fucking rockstar of science epistemology, the muchly flawed reality is that Most Published Research Findings Are False. (Someone let me know if any of these links are paywalled ... I am writing from a location with free access to just about everything in the world.)
The hoi polloi are neither stupid nor foolish to entertain doubts about what "scientists" are telling them. Scientists are humans too, and contrary to the cherished Scientific Ideal worshipped by LF and SR, the science done by those human scientists reflects their human weaknesses. Yes, in some sort of infinite long run, science may indeed correct all of its errors and frauds and self-delusions, but we humans operate on shorter time scales. Some of the most egregious errors of science have been promulgated for decades, and at horrific cost, even in the presence of vocal gadflies opposing the much-worshipped Scientific Consensus.
It is not crazy or stupid or ignorant to suspect that powerful and interested parties are working to corrupt and constrain science, because all too often, they are -- as when the beef industry attempted to squelch any public discussion of the possibility of mad-cow disease in American beef. Richard Marsh, the scientist who led the early charge in the study of the various spongiform encephalopathies, was vilified and marginalized. His grants were not renewed, and he was persona non grata with some of his immediate colleagues. Here is a cautionary tale in scientific epistemology (from The Atlantic Monthly. The bold is mine:
Stephen Morse, the director of the Program in Emerging Diseases at the Columbia University School of Public Health, worked with Richard Marsh on his studies of TSE agent in minks. In a letter to the journal Nature in March of 1990 he said that he did not believe that what was then generally called scrapie agent would ever find its way through the food chain and into human beings. But he has changed his mind. "Many were wrong about it, including me," he told me recently. "We didn't imagine it could pass from cows into humans. But now we think it can, and it has the potential to be terrifying. Perhaps the best analogy is to the AIDS epidemic. Although it's almost certain that TSE doesn't transmit as readily as HIV, it's similar in a number of other ways. It can remain in the body for long periods without obvious symptoms, and it is fatal. But what I'm thinking about is how we regarded AIDS in the early days, before we really understood it. We underestimated the threat. Perhaps we should avoid making that mistake again."
I don't know when Morse and Marsh started their collaboration, but long before Morse's letter to Nature, Marsh had demonstrated that encephalopathy could jump species when one animal ate an afflicted animal of another species. It suggests a rather startling lack of imagination, that anybody studying disease would fail to
imagine, never mind
assume, that the "agent" could "find its way through the food chain", given that the food chain in that case would require exactly two steps: Downer cow -> rendered into feed -> Beef steer -> hamburger on your grill. This is a textbook case in epistemological failure. It might disappoint LF and SR, but scientists can be foolish in all the same ways as everybody else -- worse, they can be a
special kind of foolish, the kind that biases people who both very smart and very knowledgeable against accepting that their well-established beliefs are, um,
wrong.
That's a variety of epistemological fail that is not rare on dailyKos. It's what happens when people like coffeetalk, but 20 to 50 IQ points smarter, are challenged within their domain of expertise. It's disconcerting to see. I'm sympathetic, because I know that I am subject to the phenomenon. It's Boring and Wrong, but it isn't stupid -- it's an inductive Bayesian inference derived from thousands of personal experiences. The question, "Am I Wrong, or Am I Right," is a simple classification problem, and a classification system that has been "trained" for years on observations of an individual's epistemological success rate may be biased either towards False Negatives (if the individual is almost always wrong) or towards False Positives (if the individual is almost always right). People who are both very smart and very knowledgeable are usually right about things (unless they are mad).
Sidebar: Classification Problems (Sensitivity vs Specificity)
The elementary classification problem is to take a group of samples from a population, and classifying each of them as being either X or Y. Though simple in concept, this describes an enormous epistemological range of human inquiry and cognition. Any medical diagnostic test is a classifier: "is this tumor 'Benign' or 'Cancer'?".
Few classifiers are perfect. If we consider the cancer diagnostic, there are four possibilities for every classification the test issues:
A. True Positive (sample is cancer; classifier says "cancer")
B. True Negative (sample is benign; classifier says "benign")
C. False Positive (sample is benign; classifier says "cancer")
D. False Negative (sample is cancer; classifier says "benign")
If a classifier generates few False Positives, it is said to have high Specificity. If it generates few False Negatives, it is said to have high Sensitivity (because it successfully "detects" almost all of the positive cases). Generally, there is a trade-off between Specificity and Sensitivity. If I don't want to miss any of the positive cases, I must set my classifier to generate False Positives: to be "over-sensitive" (or under-specific).
The thing is, quite apart from technological classifiers like medical diagnostics, we all employ our own cognitive classifiers continuously as we bumble through the world. At every instant we are simultaneously making dozens of decisions about the categories of the things around us and the concepts within us. Classification defines most of what passes for argument on dailyKos (e.g. "Electable" or "Unelectable")?
Either deliberately or subconsciously, we attack classification from two directions -- deduction and induction. Deduction attempts to analytically prove that certain properties or characteristics must place a sample in one class or the other. Induction attempts to assign a sample to a class according to observed similarities to other samples that are somehow known to belong to that class. (And asks the epistemologist: Whence that a priori knowledge? "Unsupervised classification" attempts to create classifications from a population of samples, simply by looking for differences and similarities among them, with no magic Oracle to authoritatively classify any of the exemplars. This is a process that, when humans do it, is central to intellectual progress.)
Folks are perpetually, and often subconsciously, building induction-based classifiers, in which process they are subject to some serious epistemological errors, mostly statistical in nature. A large chunk of what we call "superstition" is erroneous inference by induction: We confuse induction with deduction, creating the post hoc, propter hoc fallacy.
I can't say what LF, SR, et al, make of the case of the Mink that ate the Mad Cow. I do know that some of the science fundamentalists who frequent those diaries expressly believe that it is more reasonable to assume that a new technology is harmless unless
proven otherwise; they reject any epistemology that draws on generalization from experience, reasoning that most things aren't harmful, or at least, they're no more likely to be harmful than is their absence. I disagree. I suspect that most novel "things" probably
are harmful. Like other animals, we are acutely adapted to our ecological niche. Sure, our technological prowess allows us a lot more leeway than most other critters, but the bottom line is that there's no good reason to assume that the average random synthetic bioactive molecule isn't going to fuck us up. Simply put, the science fundamentalists don't want to sacrifice miscellaneous technological luxuries in the name of unknowable perils, so they
just decide that the perils aren't worth worrying about. I can accommodate that this is what they
believe. What I cannot accommodate their
certainty that they are correct. Neither they nor I can
prove our case. This doesn't agitate me, because my epistemological worldview accepts uncertainty as a fundamental quality of just about every idea anyone ever had. They, however, are dependent on certainty -- which they derive from their misplaced faith in the "evidence-based" epistemology of Ideal Science. Which would be great, if
real science came anywhere near the ideal. But it doesn't.
Can you say, "broken epistemology"?
But don't get me wrong! I agree with those guys about almost everything that pragmatically matters. I understand why they often want to tear their hair out. (The studied ennui of LF's signature signoff -- "shrug" -- is contradicted by the vehemence with which he occasionally blows his cork.)
Somewhere around here is the woman who denounced pasteurization and sang the praises of raw dairy. We danced back and forth through the usual anti-pasteurization slogans. Her concluding comment explained how she handled her (possibly illicitly-obtained?) raw milk: She boiled it. Yes. Because somehow, flash-heating milk to quickly kill the bacteria does more "damage" to the milk than cooking it. I told her she preferred the resulting beverage, not because it was "fresher" than pasteurized, but because she had caramelized the sugars. She was halfway to creme brule, for God's sake.
In a recent diary taking "traditional chinese medicine" to task, Wee Mama weighed in with her personal testimony about how her physician -- who is trained in many things, including chinese herbalism -- was able to jumpstart her ovaries and/or her womb. So. Okay. So:
A. We will never know whether her gladly-found fertility (and please understand, I intend no mockery or belittlement of the significance of WM's physiological breakthrough, it delights me that she attained this fulfillment) was a direct result of her physician's treatment, or was some sort of weird coincidence. post hoc, propter hoc, man. Biology is still mysterious, and the biology of fertility is especially so. The inexplicable is commonplace. Epistemologically, WM has some reason to suspect that her doctor's intervention contributed to the happy outcome -- but that is a personal data point that the rest of us are poorly equipped to evaluate, never mind draw inferences from.
B. Even if her doctor's application of certain chinese herbs did bring the desired outcome, that is not a validation of the body of "knowledge" called "chinese traditional medicine.
I mean, talk about a straw man. Who on earth doesn't believe that you can do all kinds of things to a human body if you feed it teas and herbal extracts? The epistemological beef that empiricists have with "chinese traditional medicine" is that:
A. Little (if any) of it can be shown, in blind trials, to be significantly effective. (See: Acupuncture. Acupressure. Etc.)
B. There is no rational epistemology underlying the "theory". 100% of Chinese medical knowledge is either the result of induction from trial and error, or is bunk -- and a good deal of it is both. The theoretical models by which practioners explain the efficacies of their art are pure balderdash. There are no fucking "lines of energy" in the body upon which their techniques may operate. Somewhere in the long-forgotten past, some guy(s) just made this shit up -- guys were were not deeply studied in the problems of epistemology.
I mean ... Jesus, sometimes we here are expected to politely interact with folks who don't get the fact that the whole homeopathic '10x dilution == 10x more potent' concept was just some random fucking thing some charlatan made up out of nowhere. Literally. He was just sitting around one day, thinking wrong thoughts, and thought that one, and decided he liked it. For fuck's sake, the word "potent" doesn't even mean anything in that formula! How the fuck is the correct "dosage" inferred for any of these little samples of hyper-purified water, when nobody can clinically demonstrate their effectiveness at any dosage at all? Drink a drop or drink a gallon, the effect will be the same, so what the fuck do you imagine "potency" describes here?
And even if potency did mean something, what kind of bizarre magic do these people think is in play? What mysterious law of physics could possibly work its way right down through the chemistry to the molecular biology and on through the cellular biology to produce such a remarkably universal principle? We know of none -- not even anything even suggestive of such a phenomenon. So how is it that every single compound, regardless of its relationship to the body's molecular biology, exhibits the same, identical, universal inverse linear property of potency?
This isn't just broken epistemology, this is drunken, coke-paranoia, fun-house mirror, full-blown red-lined superstition. It's one thing to believe something for no reason other than that somebody told you it was true. It's another thing entirely to believe that same something despite the fact that it contradicts any conceivable rational idea you have ever had in your head about how the universe works. I mean, come on.
And also, chiropractic won't cure cancer.
And Wicca is nonsense. Really. Light your candles, paint your pentagrams, mutter your spells -- the universe is unaffected and indifferent. There are no spirits, there is no Goddess, and you are all just making this up as you go along, which is the really embarrassing part. Or ought to be. (It's hilarious that I'm told that I cannot possibly have a valuable opinion about the imperatives of anyone who doesn't look like me, but that some 21st-century geek-grrl engaged in particularly elaborate and sincere cosplay is experiencing something epistemologically comparable to the hallucinogenic cognition of a socially isolated 9th-century spell caster -- who, for example, lived in real, daily fear that she might inadvertently get on the wrong side of the spirits she invoked, and find herself in some fresh hell unimagined by her modern counterpart.)
You're lying to yourselvees, and you're lying to me, and I'm not going to be polite about it. The whole point of the FSM shtick is to reveal to people the absurdity of their supernaturalisms, and I am no more obliged to treat your bunk seriously than you would be to treat seriously the affectations of any proselytizing pastafarian.
God. I'm sick of it. I once had to sit at a table and not laugh in the face of a college-educated professional who, before we adjourned our meeting, wanted to tell us all about how the earth was entering an astrological cycle of some inane sort, and we should incorporate that into our plans. I once had to sit at a table and stare at my hands when one of the women asserted that her particular skill was clairvoyancy. (Prior to her announcement, the strangest thing I'd heard at that table was the 6'-plus MtoF transsexual referring to her time in the submarine service -- I didn't think it was appropriate to ask, "Aren't you a bit tall for the submarine service"?)
Man.
And it seems like nobody around here understands what the hell a statistical conclusion means, or what kind of practical knowledge can be gleaned from the typical social science research report. And christ, there are whole books written, trying to explain the epistemological subtleties of inference from population studies, so what can I accomplish in a subsection of what has become a barely-hinged rant?
Look. Imagine there's a study. And imagine the study has 1000 male participants and 1000 female participants. And imagine that the participants are each given almost-identical sets of 10 applications to rank for an entry-level clerical job. Imagine that the names on the applications are all women's names. And that two of the names end in "y". And that one thing that changes from one set of applications to another is, one of the "y" names is changed to end with an "i". And that the other thing that changes is that the names are randomly assigned to the applications.
Now imagine that amongst the 1000 female participants, Betti and Patti and Betty and Patty and every other name all get exactly the same average (i.e., arithmetic mean) rank: 5.5. (unlikely, but hey ... Imagine). And imagine that amongst the 1000 male participants, Betti and Patti each average a rank of 5.0, and the other names all average 5.625 (Imagine dammit!).
I'm not going to bother to calculate the "p-value", or the confidence interval, or anything else, I'm just going observe that if such a study were published, the accompanying headline would be, "Men discriminate against women whose names end in 'i'". And then, sooner or later, some dumbass on This. Very. Blog. would make some sort of dumbass argument whose premise was that every single man in American entertains secret doubts about the competence of women whose names end in 'i'".
But that isn't what the data indicate. What the data indicate is that: there is a significant likelihood that in a large sample of men, more of the men will tend to think less of a woman whose name ends in 'i' than will think better of a woman whose name ends in 'i'. That is it.
And that is only an indication -- after all, we only tested 2 names.
I can tell you -- since I made the data up out of nowhere -- that the "Ground Truth" is that 80% of the female participants discriminated marginally against the "i" names, but 20% of the female participants -- the ones whose names ended with "i" -- were so favorably biased towards Betti and Patti that they completely balanced out the bigots. The men, however, were largely uninfluenced by names -- except for the 10% of them who had been dumped in high school by a narcissist with a name that ended in "i".
See how that works?
Bleargh.
Disclaimer. I think there is a very high likelihood that this statement is true: "Vaccines are unrelated to autism".
Disclaimer disclaimer: Contrary to the assertions of some who carry the anti-anti-vax standard on dailyKos, that statement has not been proven. I'm not sure how it ever could be, other than to come up with some other set of explanations that accurately predict every single case of autism. What has been shown, beyond any reasonable doubt, is that if there is a link between vaccination and autism, it is subtle, and it is impossibly improbable that it explains more than a small fraction of autism cases.
Amplified Disclaimer: I support universal vaccination! People who don't vaccinate their children anger me, for several reasons. I oppose religious and other philosophical objections. Get with the program, or find another society to plague with your anti-social paranoia.
Amplified Disclaimer notwithstanding, I'm not going to make epistemologically vacant assertions about what has or has not been "proven" with respect to vaccines and autism.
Autism aside, there are a lot of things that haven't been proven, but are asserted around here as having been proven. People often treat the following statements as equivalent:
A. The study identified no statistically significant relationship.
and
B. The study identified no statistical relationship.
If a study finds a statistical relationship, but the ANOVA gives only a 70% chance that the relationship wasn't due to random chance, the authors will report "A", and unless we dig into the publication ourselves, we'll never know whether they observed a relationship that didn't meet the all-holy 5% confidence threshold. Does that matter? Well it does, if you want to make grand statements like, "A study proved that watching American Idol does not make people stupider."
Oh, fuck it, I can't write any more about this. This topic requires a book, but you know what? That book has already been written -- probably a thousand times. Go read a couple of them. Start with Historian's Fallacies, by David Hackett Fisher. And then maybe Eric Blair's commentary on language and political rhetoric. Richard Feynman has a nice little treatment of "Cargo Cult" superstition in one of his books. Hell, pick up a couple of not-too-difficult science journal articles, and look at how their actual numbers, and compare those to their conclusions, and ask, "What is a true statement I can make about this study and its observations. Oh yeah, read Ioannidis's paper, cited above.
Peace fucking out.