I’d originally considered entitling this post simply “More Bullshit”, but that of course would be misleading—The Oikofuge attempts to be a bullshit-free zone.

I’ve posted on the topic of bullshit before, when I wrote about Pennycook et al.‘s classic paper “On the Reception and Detection of Pseudo-Profound Bullshit“. Now I’d like to share with you a new exploration of bullshit—Turpin et al‘s “Bullshit Ability as an Honest Signal of Intelligence“, recently published by the journal Evolutionary Psychology.

In two studies, involving a total of more than one thousand undergraduates (participating in exchange for course credit, as seems to be pretty routine in psychological research), Martin Turpin and his colleagues in Ontario, Canada, investigated the links between bullshitting, intelligence, and other people’s assessment of the bullshitter’s intelligence.

Here’s what they did, while in the process introducing some marvellous technical terms:

First, they assessed the Bullshit Willingness of the participants, by inviting them to score their familiarity with a list of ten concepts using a scale from “never heard of it” to “know it well, understand concept”. The trick being that, while six of the concepts (like “Cognitive Dissonance”) were real things, four (like “Neural Acceptance”) were entirely invented names. Participants were then scored on their willingness to pretend familiarity with fictitious concepts.

The participants then moved to the Bullshit Production phase. A subset of them were presented with the previous list of ten concepts, and asked to produce a “convincing and satisfying explanation” of each—making up an explanation for any terms they were unfamiliar with. After which, Bullshit Raters (the rest of the participants) scored the bullshit explanations of the fictitious concepts according to how accurate and satisfying they found the explanations, from which a Bullshit Ability score for each “explainer” was derived. In the second study the Bullshit Raters were also asked to say how intelligent they thought each explainer was.

Finally, all the participants completed a couple of measures of general intelligence (the Wordsum task and Raven’s progressive matrices) and were then given a task which assessed their Bullshit Receptivity and Bullshit Sensitivity. They were asked to assess the profundity (or otherwise) of thirty statements originally used in the study by Pennycook et al. Ten of these statements are mundane observations (“Some things have very distinct smells”). Ten are motivational quotations that contain some sort of observation on the human condition (“A wet person does not fear the rain”). And ten are meaning-free pseudo-profound bullshit, generated algorithmically by arranging profound-seeming words and phrases into random grammatical sentences (“We are in the midst of a self-aware blossoming of being that will align with the nexus itself”). You can find the whole list in Turpin et al.’s Supplemental Material. Participants who assigned high profundity to the pseudo-profound statements received a high Bullshit Receptivity score; Bullshit Sensitivity was assessed by comparing each participant’s rating of pseudo-profound bullshit with their rating of the motivational quotations.

With me so far? Fair enough. To recap, the experimenters ended up with a bunch of scores relating to a bunch of people: their Bullshit Willingness (readiness to bullshit); their Bullshit Ability (aptitude for bullshitting); their Bullshit Receptivity (readiness to accept bullshit); their Bullshit Sensitivity (ability to distinguish bullshit from non-bullshit); a couple of measures relevant to their intelligence; and an assessment of their perceived intelligence, based on their ability to bullshit.

The researchers than whacked all those data into a big correlation matrix, to see what correlated with what. Here’s what they found:

• Participants’ Bullshit Ability correlated with both measures of their general intelligence
• Participants’ Bullshit Willingness correlated negatively with both measures of their general intelligence
• There was no correlation between Bullshit Ability and Bullshit Willingness
• Participants’ Bullshit Ability correlated with their perceived intelligence
• Participants’ Bullshit Willingness correlated with their Bullshit Receptivity and correlated negatively with their Bullshit Sensitivity

Accepting, for the sake of argument, that high scores on the two little intelligence tests administered in this study actually pick out smarter people, what the authors seem to have discovered is that smarter people are better at bullshitting, but more reluctant to bullshit, and that being a good bullshitter results in people perceiving you as being intelligent. If all this is true, one interpretation is that, in evolutionary terms, good bullshitting is an “honest signal” of intelligence—people correctly judge the good bullshitter to have above-average intelligence. The authors’ interpretation is as follows:

[W]e propose that the ability to produce satisfying bullshit may have emerged as an energetically efficient strategy for achieving an individual’s goals (such as acquiring status or impressing mates). That is, a person can engage in the arduous process of acquiring expert skills in domains that they could then leverage to accomplish certain goals, or can use bullshit as a strategy that potentially produces the same benefits at a much smaller cost. Of course, these strategies need not be mutually exclusive, as the ability to produce satisfying bullshit may help even highly skilled individuals achieve their goals over equally skilled peers.

Pity the poor person who has high Bullshit Willingness but poor Bullshit Ability, however—their audience will form an unfavourable view of their intelligence. And their Bullshit Willingness goes hand-in-hand with increased Bullshit Receptivity and decreased Bullshit Sensitivity. To quote the authors again:

Thus, contrary to the common expression, it may indeed be possible to “bullshit a bullshitter.”