A Tale of Two Fallacies: the Insanity of Progressive Stacking
As I write this (on a plane en route from Cuba to Toronto), I’m thinking in an undisciplined manner about some recent developments in public intelligentsia. I’m not quite sure how best to describe what I’m calling something of an “academic rift”, as its nature is only dimly formed in my frontal lobe. But it has something to do with the value of evidence and the extent to which we are willing and able to allow evidence (or what passes for evidence) to dictate our view of social and physical reality. Even as I write those words, I am aware that there will be disagreements about what constitutes “evidence”, and a full appreciation that almost nothing exists in an objective vacuum, immune from the infection of personal bias and value.
As with all things upon which I percolate in this blog, I bare my raw cognitive process in hopes that some form of clarity might emerge.
Two disciplines in particular seem to be at the front of this conflict between ideology and (for lack of a better word) evidence: gender studies and evolutionary psychology. Ironically (deliciously so), the two are in some ways on opposite ends of the political spectrum. I’m nowhere near brave enough to offer any criticism or commentary on gender studies (yes, I’m a coward), but the core criticism of evolutionary psychology is that it offers broad generalizations that can be used to rationalize, or even justify, the basest of human actions and desires.
It bothers me, though, that critics of “evo psych” wish to discount the entirety of the field because of the oversteps and overreaches of some of its proponents. There are evolutionary reasons, after all, for many (evo psych evangelists would argue for all) human attitudes, beliefs, and dispositions. There’s an evolutionary reason we rarely look up, for example; threats historically came to us from across the savannah. There’s an evolutionary reason for why we prefer natural vistas to sterile urban panoramas: our genetic memory of which environment most commonly offered us protection and food –those who retained this memory or predilection were more likely to seek out and benefit from such protection and bounty, and were thus more likely to pass on their genes… the theory goes.
But is there an evolutionary reason for more socially and politically problematic behaviours? Rape, for example, is described by evo psych adherents as a strategy for low status males to transmit their genetic material into the next generation without having to succeed at the social game better played by their sexual competitors. Critics of evo psych, however, argue that that explanation is simplistic and ignores the power-seeking and power-expression aspect of rape, as well as lived experiences that might stoke such behaviours, beyond the dry and valueless quality of being described as being “low status”.
(Note: I am not interested in debating in this forum whether or not rape is an evolutionary construct; the possibility is presented herein simply as an example of a touchy subject that gets evo psych adherents into trouble.)
At its core, though, to accept an evo psych explanation of a thing is to also accept that thing’s “naturalness” and, likely, its social inevitability. A postmodern approach to such phenomena, on the other hand, would see us digest rape as a social product mostly divorced from biological need, and therefore deletable given the proper social policy choices.
Similarly, it bothers me that whose who wish to discount the claims of gender studies orthodoxy do so at the peril of ignoring real, systemic and institutionalized power disparities between genders.
It seems to me that, as with all things, such phenomena are sufficiently complicated for all parties to be at least partially correct. A thing like rape can have an evolutionary role and underpinning while at the same time be mutated, extended, and even diminished by the application of social methods.
But I digress. The fundament of controversial disciplines like evo psych (and gender studies) is, frankly, also their tragic intellectual flaw, insofar as the scientific method is concerned: they presume premises that are not falsifiable. Both the presumption of an evolutionary origin to thoughts and behaviours and the presumption of an ill-defined patriarchy that worms its way through all human interaction are not provable by any positivist method, and must therefore be accepted (or rejected) based upon how much one “feels” them to be true. Dangerous stuff, this. (And I am certainly not courageous enough to go further, at least not in this public forum.)
So when the market’s order call for “evidence” needs heeding, some disciplines will find it easier to fulfill than others. And those disciplines’ political brethren, defined in large part by identity dynamics, will gather to defend the herd. The divide between these groups is, I think, that to which I am fast becoming sensitive: those who do and those who don’t believe that some are sacrificing reason on the altar of ideology.
And the conflagrations exist at both ends of the political spectrum, to be clear.
A handful of famous public intellectuals appear to be at the forefront of this controversy, in broad strokes. The obvious name of Jordan Peterson comes to mind, he whose citation of the statistical observation that gender identity, biological sex, and sexual orientation do not vary independently has put him at odds with those who insist that some or all of those things are social constructs.
Joining Peterson in this category of scholars now vociferously opposing the increasingly mainstream positions of self-styled progressives are celebrated psychologist Jonathan Haidt, who is on a mission to encourage all academies to embrace ideological heterodoxies, and Bret Weinstein and Heather Heying, both evolutionary biologists, who point to the creep into school curricula of politicized expressions of biology, most notably gender and sex, in seeming denial of laboratory and statistical evidence.
And foremost on my mind today is controversial Harvard psychologist Steven Pinker, whose latest book, “Enlightenment Now”, I am currently partway through.
Now, my ignorance is so profound that I did not even know that Pinker is considered “controversial” until I took a peek at my Facebook and Twitter feeds and saw all the smart people I admire (whom I affectionately call “nerds with words”) disparaging Pinker’s otherwise celebrated name.
Pinker’s book sings loving paeans to a time in history of which I have some affection, The Enlightenment, whose values, simply and broadly stated, both Pinker and I feel are definitive of most of what is admirable in Western civilization.
I am by no means a trained historian or philosopher, so take Pinker’s claims at face value, which might be a mistake. I am happily learning from my smarter friends in the Humanities about all the ways that Pinker’s thinking might be flawed. And I will reserve commentary on the specifics of those analyses for a later date. That is not the topic of today’s post. But I mention this to note, as always, that my thinking on this and other topics is always in flux. One should consider this post to be a public description of ideas with which I struggle, and by no means my concluding position.
All of this is rather concerning for an epidemiologist like me, whose mantra is supposed to be celebratory of the primacy of positivist evidence. The most prominent public manifestation of the “values and feelings” diktat, in contrast to the stance of the evidence evangelists, is the growing reach of so-called “identity politics”.
Both Peterson and Haidt –both figures of some public derision for one reason or another– have identified identity politics as natural outgrowths of neo-Marxist thinking that reduces all human relations to dominance hierarchies based on power, wherein one’s group affiliation (i.e., identity) is assumed to be the best estimator of his true societal power. And increasingly, they argue, such power ironically flows to he who is the greater perceived victim, or the greater claim to having experienced historical injustices .
I am not sufficiently intellectually equipped to determine whether any of this is based in reality. But one precipitate of identity politics that does concern me is the extent to which such acceptance of dominance hierarchies based on victimhood can be a barrier to fair and effective scholarship and –most concerning to me– paedagogy. At this stage of my career, increasingly all I care about is how best to service the intellectual growth of all my students.
Not surprisingly, Pinker himself recently commented on the dangers of identity politics to the academy: “it undermines one of the greatest epiphanies of the Enlightenment: that people are equipped with a capacity for sympathetic imagination, which allows them to appreciate the suffering of sentient beings unlike them.” To suggest that one gender cannot empathize with the experiences of the other, or that one race cannot empathize with the struggles of other races, in Pinker’s view, fails to “give force to the very movements for moral improvement that originally inspired identity politics.”
He further states: “The slave trade and the Holocaust are not group-bonding myths; they objectively happened, and their evil is something that all people, regardless of their race, gender, or sexual orientation, must acknowledge and work to prevent in the future.” In other words, by reducing experiences to being “owned” solely by the people specifically implicated in the events in question, we risk failing to project the universality of the moral lessons to the entire world. (Pinker then dismisses accusations of “cultural appropriation” as “assinine”, which, I suppose, hints at the shallowness of his thinking that my friends warned me about.)
As I meditate on these issues, one element of my profession bubbles to the surface of my murky thoughts: the interface between cognitive and statistical biases. There are many examples of this phenomenon, but herein I will describe but two: “ecological fallacy” and “exception fallacy”. Other disciplines, like sociology and psychology, might have different names for these fallacies, but I will proceed with the epidemiological nomenclature. (Please note that I consider these fallacies to be formal biases, and will refer to them as such, though other epidemiologists might disagree with this very fine point.)
Exception fallacy occurs when we assume a group to possess the characteristics of an individual from that group. Let’s say you meet an Asian kid who demonstrates remarkable mathematical acuity. You might problematically conclude that all Asians are good at math.
Is that conclusion racist? It might be, depending on your definition of racism and your threshold for detecting (or being “triggered”) by such things. Evo psych adherents might argue that this bias is a natural, though sometimes not ideal, trait that we developed through millennia of exploring our environment: this red striped snake was aggressive, so I will assume all red striped snakes are aggressive and will seek to avoid them in the future; a winning survival strategy!
Postmodernists seeking deconstructive elements to this fallacy will see the very real intersection of learned behaviours —it’s a prominent stereotype that Asians are good at math, impossible to avoid; and the fact that the protagonist in the example focused on the kid’s race and not other identifying attributes, like the size of his nose or colour of his shirt, is clearly and expression of some learned or constructed sociolinguistic effect or other.
Regardless, I think we can all agree that exception fallacy is a real thing and that it is a problem, despite possibly having had some evolutionary advantages in the past. Frankly, I posit that it is objectively inappropriate to define a group by the characteristics of one person from that group. Not only is it atrociously prejudicial, but it is also ineffective as a means of prediction. Inferential reasoning is not an ideal form of reasoning. It is from the regular failing of single example inferences that we eventually developed the statistical science of representative sampling, to reduce exception fallacy.
Ecological fallacy, on the other hand, is the logical converse. It’s the assumption that an individual has the characteristics of the group from which he or she emerges. Now, committing an ecological fallacy is clearly a racist behaviour… or is it?
“That kid is Asian. And we know that Asian kids are good at math. So that kid must be good at math.”
Is the above statement racist because of the deduction that since the individual came from the group, he must immediately have characteristics of that group? Or is it racist because the assumed characteristics of the group are themselves nothing more than a racist stereotype?
Consider a less inciting example. In a group of 10,000 people, the prevalence of obesity is 25% (i.e., 2500 people are obese). If you randomly select 1000 people from that group, how many people in your sample would you expect to be obese? The correct answer, statistically, is 250. That’s right, there is a kind of deductive logic at play here, wherein one expects the frequencies of characteristics to be the same in the sample as in the reference group. If the descriptors of the group characteristics are true, and not merely assumed racist stereotypes, then I contend that ecological fallacy in this sense is not innately racist (or sexist, or any other kind of “ist”).
But… but… but… this is only statistically true if the sample is not small and is a random selection of the reference group. Had I chosen a single person from that reference group of 10,000, the likelihood of that person being obese would still be 25%, but I certainly would not want to bet any money on that probability: the small sample size makes that statistic practically meaningless.
Further, had I asked for 1000 volunteers from the group of 10,000 to bravely come forward to have their obesity status publically declared, it is likely that mostly the non-obese ones would have volunteered. This would give my sample of 1000 a so-called “selection bias” and the sample’s obesity prevalence would probably be much lower than 25%.
This is all to say that ecological fallacy is problematic on its face. But when the sample is very small and/or nonrandom, the bias is monstrous indeed.
What does all this have to do with identity politics? In our present political climate, it is not unusual to find people declaring their identities based upon their group subscriptions. Take a gander at the way many people describe themselves in their Twitter profiles. Here are a few:
Exhibit A: Prominent activist Linda Sarsour self-identifies first as a “Palestinian-American-Muslim”:
Exhibit B: This person wants us to know, first and foremost, that she is a US Air Force veteran:
In a strong sense, the act of declaring one’s identity to be reflected in one’s group membership(s) is to also declare that ecological fallacy does not apply. One is actively asking for the groups characteristics to be assumed to apply to the individual. What does it mean to be a “feminist lesbian” or a “Christian mom”? Clearly, there are assumed characteristics shared by “feminists”, “lesbians”, “Christians”, and “moms” that these individuals would like the world to know also apply to them as individuals. And that is certainly their right, to define their identities as they choose.
The problem percolates when they/we begin to assume that group characteristics apply to members who do not feel that their identities should be reflective of group membership. One question naturally precipitates: when do one’s individual achievements, qualities, experiences, and thoughts accrue to sufficiently outweigh the assumptions of identity made based upon nothing more than group membership?
The obvious answer is: when the individual chooses. But that is not the reality with which we are increasingly faced. A “white male” increasingly has his group characteristics assumed of him (specifically that he has high social privilege), regardless of how much he seeks to have his personal individual qualities recognized. Of course, historically, that has been the complaint of people of color (POC), women, and other minorities for centuries: that they are judged not by the content of their character, but by the colour of their skin.
Not susprisingly, this question abuts one of the aspects of Enlightment thinking upon which Pinker wishes us to cogitate: that the age of supposed objectivism freed us from the tyranny of group identity and allowed us to revel in the terrifying freedom of individual expression.
It is my understanding that historically this freedom was best reflected in the diminishing of religious dogma, insofar as self-described scientists were newly able to define reality based upon their observations of the universe, rather than feel compelled to adhere to Christian orthodoxy and its received cosmological wisdom. And of course, that freedom was never absolute, nor is it, in my view, a natural state of the human animal. With every generation we seek to associate ourselves with communities both large and small, after all: religions, clubs, sports franchises, political movements, or ideological schools of thought. Those who reject such memberships are often the excluded and marginalized ones.
And this is probably why Pinker’s book, flawed as it is, has some value for me: it reminds me that the freedom to eschew groupthink was hard-won, historically rare, and precarious even unto this day.
That, my friends, brings me to the practice that first got me thinking about these matters: something called “the progressive stack.” I’m trying to choose my words carefully in this essay, and to avoid unnecessary hyperbole. But I must be frank and say that I find progressive stacking anti-intellectual, socially regressive, and not only not helpful, but ultimately harmful, especially when applied in the classroom.
Progressive stacking began during the Occupy movement. In a diverse group, when several people wished to speak, they would be called upon in inverse order of their assumed power in society. This means that low-status individuals, defined thus by their group memberships, would have speaking priority over high-status individuals, also defined by group membership.
So, a transgender, LGB, female person of colour would outrank a cis-heterosexual white male. It’s generally accepted that “there are personal, cultural, learning, and social reasons people don’t speak up in class.” When seeking to level the playing field for all students, “we are working against powerful structures of inequality where gender, race, and other aspects of identity often prevent voices from being heard.” Or at least those are the claims of defenders of “stacking”.
In defence of the practice, one writer offers: “it doesn’t mean excluding men or white students from conversations, or forcing underrepresented students to talk. Instead, it means calling on students who want to talk in the reverse order that one might predictably do so, based on social biases.” That was written in defence of Stephanie McKellop, “a graduate teaching assistant [who claimed to be] under attack by fringe-right groups for using progressive stacking in her classes and then tweeting about it.” Yet this is what McKellop herself tweeted about her practice (emphasis mine):
“And if I have to… white men.” Sigh. Sounds like a grudging acceptance of white men, rather than simply a de-ranking of their privilege. This brings me back to my earlier comment, about the point at which individual identity should supercede group identity. There is an assumption made by purveyors of progressive stacking that people being “stacked” are defined more by their group affiliations than by their individual experiences, and this strikes me as (1) presumptuous, (2) racist, (3) authoritarian, and (4) illogical, given the threats posed by exception and ecological fallacies.
Defenders of stacking will claim that this is not about identity, but about power. Societal privilege is real, after all, and we ignore it at our peril. I am reminded of an interview comedian Norm MacDonald once had with a conservative radio host. Norm had commented something to the effect that “black people are poorer than white people in America.” The host was agitatedly taken aback, and had replied something close to, “I know lots of poor white people. And look at all those rich black basketball players!”
Now, the host was clearly statistically ignorant. On average, white families in America have nearly ten times the net worth of black families. The data on this gap is quite clear. The radio host had fallen hard for exception fallacy, assuming that her examples of poor white people and rich black people were statistically representative of the entirety of their groups.
The element of privilege, in my estimation, goes something like this: “Are there poor white people, just as there are poor black people? Of course there are. But those poor white people aren’t poor because they are white; whereas many of those poor black people are poor in large part because they are black.”
So, to reiterate, privilege is real. And racism, both overt and systemic, in classrooms is often real, as well. Defenders of stacking will argue that by bringing the practice into classrooms –compelling the white male students to speak last while encouraging the others to speak first– is an attempt to correct for the advantages in privilege experienced by the white males outside of the classroom.
To simplify this overlong screed, my objection is in two parts: (1) the practice is offensive, and (2) the practice is illogical.
The illogical: My argument concerning stacking’s illogic is based, again, on exception and ecological fallacies. A classroom full of students is a small nonrandom fragment of society. This means that it is clearly statistically not representative of any larger population. This is increasingly true the higher up (later year) the class is, and when the class is not mandatory. A self-selected small set of individuals cannot possibly be expected to reflect the characteristics of the groups they purport to represent, and certainly not to the exclusion of grander individual variants, like personal experience and family socioeconomic status.
As one writer notes, proponents of stacking have decided that a pre-selected set of students’ characteristics are all that will define their power and privilege within the classroom. Those characteristics, again, are race, gender, orientation, and disability status, partly because those are the characteristics most visible or publicly celebrated. What remains largely invisible are students’ familial wealth, medical issues, psychological struggles, and other individual experiences that have affected their paths through life.
Statistically, what is more likely to hinder your participation in a classroom? Your race or your introversion, language deficiencies, personality, or general psychology? Of course all of these things are related, but it’s an unsupportable error to assume only a couple of those descriptors truly matter.
Is it possible that the white man has more social privilege than the brown woman? Sure. But in my classroom of a small number of non-randomly selected individuals, it is equally likely that the white man is from a poor coal-mining family, is the first in his clan to go to university, and suffers from all sorts of deprecating maladies, while the brown woman is the daughter of wealthy physicians from a long line of privileged and powerful professionals.
Their race and gender tell me nothing about their actual social privilege, because of the ecological and exception fallacies. In fact, by choosing to sort strangers by visible demographic indicators, it is likely that one would be creating new inequalities and new injustices.
Furthermore, progressive stacking is postmodern deconstruction rendered for the practical world, and disastrously so. As Alan Sokal once said, “I confess that I’m an unabashed Old Leftist who never quite understood how deconstruction was supposed to help the working class.”
As essentially an expression of Marxist leveling of power, stacking fails in its primary task, because it ignores the fundament of its philosophy, that social inequalities are class-based, with the working class perpetually underserved. By focusing on race and gender identity, class and wealth are altogether ignored; yet socioeconomic status remains the single biggest predictor of power and privilege. With the aforementioned epidemiologic fallacies, this failing renders progressive stacking an utterly pointless, and possibly harmful, exercise.
The offensive: I’m an immigrant person of colour who comes from poverty. Those are my victim bonafides. All I ever wanted or asked for was a chance to enter the game, for a place on the starting blocks: entry into a university classroom. Once there, I would compete for my credentials on a level playing field with everyone else who had made it in.
The suggestion that I would require intervention from a professor to prioritize my participation in a class is offensive and infantilizing. It is also ironically a subtler expression of the condescending colonial attitude that stacking purports to solve. Stacking is essentially the perfect manifestation of what George W Bush and Michael Gerson called “the soft bigotry of low expectations.” And I will certainly refuse to implement it in any class that I teach.