Propaganda and Bullshit, Academic Style

Bullshit, Academic Style

In lay terms, bullshit refers to lies and distortions. I do not use it this way, at least not exactly.  Bullshit in the academic sense refers to “impressive-sounding claims without a direct concern for truth.”  They may be false and they are statements no one should believe, but it is not because they are lies or mistakes.  They are statements made that flagrantly disregard standards for truth, or, in the social sciences, validity.  The rumor mill has it that Frankfurt, the Princeton philosopher who wrote the book, On Bullshit, was inspired to do so by what passed for postmodern and “studies” scholarship.  Here are few of his tips (quoted or closely paraphrased) for detecting bullshit:

• It is closer to a bluff than a lie.

• It is impossible to lie unless one thinks one knows the truth.  Producing bullshit requires no such conviction.

• Bullshitters do not care whether the things they state describe reality or not;  bullshitters just pick them out, or makes them up, to suit their purposes.

• The contemporary proliferation of bullshit also has deeper sources, in various forms of skepticism which deny that we can have any reliable access to an objective reality.

Some of my favorite examples of bullshit come in the form of ideas that certain groups of people have been “granted Whiteness” or are not authentically Black because they have the wrong politics.  My Wooden Stakes entries are basically exposés of bullshit masquerading as science (for examples, go here, here or here). 

Illustration by Dirk-Jan Hoek. This appears to be a bullseye, not bullshit. But it is bullshit. Is it obvious how?

Propaganda Scholarship

Propaganda is related to bullshit but not quite the same.  Here is the definition provided by Encyclopedia Britannica:

• Propaganda is the dissemination of information—facts, arguments, rumors, half-truths, or lies—to influence public opinion. Deliberateness and a relatively heavy emphasis on manipulation distinguish propaganda from casual conversation or the free and easy exchange of ideas.

Propaganda is not the same as bullshit, but it is ready, willing and able to deploy bullshit (“rumors, half-truths, lies..”) to influence other people. But sometimes propaganda can exploit truth to promote falsehoods. A list of rich or powerful pedophiles (e.g., Epstein, priests) could surely be true but imply far more than “some pedophiles have been successful in life.” This type of thing is commonly deployed to advance Qanon conspiracy theories without having to explicitly mention anything about secret cabals of Satan-worshipping pedophiles bent on ruling the world. The point is not Qanon; it is that something can be true and still be deployed for propaganda purposes.

The first two lines of this are clearly true. Would you describe this as mostly true? Kinda true? They got some things right? Or propaganda?

My main interest, here, however, is not general political propaganda, but the use of propaganda-like techniques in academic scholarship. I care particularly about this for two reasons: 1. Academia, via natural & social science, history, math, etc. is tasked with truthseeking; if acacdemia, as one of our society’s main truthseeking institutions becomes corrupted by political activist agendas, it has the potential for very bad downstream effects on the society writ large (see here or here for some big picture analyses); 2. As an academic myself, seeing it on a disturbing basis (go here, here, or here for examples) is just appalling and I can’t help but want to do something about it (which is one reason Unsafe Science exists).

I note that although propaganda can be deployed to camouflage politics as scholarship, it can also be deployed to promote a non-political theory or conclusion. For example, the peer reviewed literature touts the effectiveness of anti-depressants to a vastly greater extent than is warranted on the basis of the actual findings.

A paper that should, in my view, be much more widely read by social scientists, Gambrill & Reiman (2011), developed a propaganda index to be used when one is a reviewer evaluating whether to publish an article in a peer reviewed journal. However, one does not need to be a peer reviewer to use it; anyone can (including you!).  Here are some of the signs identified in the Gambrill & Reiman paper that one can use to figure out whether a paper is propaganda:

• The paper fails to acknowledge that some of its claims are controversial in the sense that at least some, perhaps many, do not accept those claims as justified.

• If an issue is controversial, is only one view presented?

• Are citations presented to support the view presented? If so, do they actually support the view?  (see this Wooden Stake entry for a good example of citations not supporting a claim purporting that some STEM fields constitute a “hostile obstacle course” for certain groups).  Beware what Brett Weinstein and Peter Boghossian have called “idea laundering,” aka as the Heffalump Effect, named after a Pooh story in which he tried to track down the elusive Heffalump and took his own tracks as evidence for the Heffalump.  This involves situations where there are citations, and they do actually state things that support the authors’ claims.  But the cited articles themselves, despite stating “X is true” provide no evidence that X is actually true (particularly common in social science claims about the supposed inaccuracy of stereotypes).

• Are applications or real world interventions recommended without evidence documenting their effectiveness? Have their (perhaps unintended) side effects been evaluated?  Has research actually shown that the benefits exceed the costs?

• Use of thinly-disguised causal language when one only has correlational data.  Researchers are often savvy enough to avoid “cause” or even “influence” (though not always).  Instead, they often will use other types of less obvious causal language, such as “impact” “drives,” “increases,” “reduces,” “produces,” “affects,” or “the effect of” etc.  If A “drives” or “reduces” B, then A causes B.  If coefficient X is “the effect” of A on B, the claim is causal.  Such claims are disturbingly common in observational, non-experimental, correlational studies (go here for a recent example), despite any evidence in those studies that the authors engaged in the very very very difficult steps needed to infer cause from correlation.

This was not in Gambrill & Reiman’s paper, but I think it is common, so:

• Refuting claims that have not been made, aka strawman arguments, also functions as propaganda because it slyly creates the appearance, but not the reality, of support for the whatever the strawmanner is arguing for, typically, the opposite. A common modern version of this are attempts to refute evolutionary psychology by accusing it of “genetic determinism,” the idea that evolutonary psychologists claim or assume that behavior and psychology are determined by genes alone. It is true that if this was a general claim it would be manifestly wrong and render evolutionary psychology noncredible. But it isn’t. No evolutionary psychologist I have ever read, and certainly none of the most influential (Pinker, Buss, Tooby, Cosmides), have ever made such an argument; all acknowledge not only that culture, socialization, and experience also influence behavior and psychology, but, sometimes, are the main influences.

Why is this Important?  

Everyone already knows that the politically savvy are very good at promoting both propaganda and bullshit.  But so can academics and scientists (some of whom are quite politically savvy). It can be even harder to detect among academics because they often adopt the mantle of being some sort of objective expert scientist, which provides a veneer of plausibility that makes it much more difficult to see through to the bullshit or propaganda.  It is still difficult, but I hope this essay has made it slightly easier.

I am going to end with a statement of principle.  I use the term “scientific fact” as articulated by the famous Harvard evolutionary biologist, Stephen Jay Gould:

• “In science, ‘fact’ can only mean ‘confirmed to such a degree that it would be perverse to withhold provisional assent.’” 

Assent (agreeing that something is true) is always provisional (as per Gould and I concur) in the sense idea that any scientific claim, no matter how well-established, might be overturned some day in the face of new, overwhelmingly powerful contrary evidence. I am not dismissing all academic scholarship as either bullshit or propaganda, even on politicized topics. I will say, however, that my overall take is that only a small proportion of social science work on politicized topics has been “confirmed to such a degree that it would be perverse to withhold provisional assent.”

So in this paper, we added this corollary to Gould’s definition:

• Anything not so well established that it would not be perverse to withhold provisional assent is not an established scientific fact. When there are con­flicting findings and perspectives in a literature, it is not perverse to believe otherwise, rendering it premature for scientists to present some claim as an established fact.

They may “believe” their own propaganda, but, if they promote something as truth before it has been so well established that it would be perverse to withhold provisional assent, they are promoting bullshit.