“So you’re saying … we should live like lobsters?” or: Why does politics make us stupid?

A few weeks ago, a TV interview of clinical psychologist Jordan Peterson by journalist Cathy Newman became a minor Internet phenomenon, thanks to the journalist’s extraordinary interviewing style. She handled the conversation so badly that the Atlantic commented on that car-crash of an interview under the title Why Can’t People Hear What Jordan Peterson Is Saying?

To provide some background: Jordan Peterson is somewhat famous for defending anti-political correctness positions, for instance arguing that respect for transgender people does not justify proposals for legislation that would compel people to use particular pronouns when referring to them, of the kind considered in Canada. He also defends a broadly conservative agenda in social and cultural matters.

But that’s not the point here. The reason that interview became an Internet sensation is the bewildering behavior of the interviewer. Like a Theme and variations piece, the conversation between Peterson and Newman follows a simple pattern that is repeated multiple times:

  • Jordan Peterson makes a point, tries to provide arguments and occasionally appeals to some evidence.
  • Then Newman interrupts him (often in mid-sentence) with the words “So, you’re saying that…” followed by some fantastically distorted version of what Peterson just said.

The most egregious example occurs towards the end of the segment, when Peterson tries to argue that surely the fact that we have hierarchies in human societies is not surprising, given that there are such hierarchies in very distant species. The exchange is worth quoting in full:

  • Peterson: There’s this idea that hierarchical structures are a sociological construct of the Western patriarchy. And that is so untrue that it’s almost unbelievable. I use the lobster as an example: We diverged from lobsters evolutionarily history about 350 million years ago. And lobsters exist in hierarchies. They have a nervous system attuned to the hierarchy. And that nervous system runs on serotonin just like ours. The nervous system of the lobster and the human being is so similar that anti-depressants work on lobsters. And it’s part of my attempt to demonstrate that the idea of hierarchy has absolutely nothing to do with sociocultural construction, which it doesn’t.
  • Newman: Let me get this straight. You’re saying that we should organize our societies along the lines of the lobsters?

The interview deserves to be studied in full, because the hapless Newman does it about a dozen times – she mistakes equality of opportunity for equality of outcome, for instance, or seems baffled by the notion of multiple factors – and each of these spectacular bloopers is introduced by the infamous leitmotif “So you’re saying…” (Note that this is NOT a post about whether Peterson is right about this. I think he is only partly right here, as there are unique evolutionary features to human hierarchies – but that’s another story.)

As many have commented – the journalist does not seem very bright. That’s certainly the impression one is left with. But that is not much of an explanation.

Obviously, what Newman is trying to do, as the Atlantic piece points out, is not to find out what Peterson is saying, but to smear him by associating him with extreme, absurd or repulsive beliefs.

Most of us have experienced such conversations:

“Sure, there may be cheaters among people who claim disabled benefits…

– So you are proposing to slaughter all handicapped people, like the Nazis?”

Why would anyone try to sound dumb?

The piece in the Atlantic interprets this willful stupidity as a recent and troubling phenomenon, a symptom of the coarsening and polarization of political debate. But that is not entirely plausible – the reductio ad Hitlerum and other forms of abusive ad hominem arguments have a long past. The famous debates between Aischines v. Demosthenes contain gems like “I will not mention that my opponent’s mother was a prostitute…”

Why do people say such things?

The simplest explanation would be that they hope to convince their audience. If I can get my listener to believe that my adversaries’ relatives engaged in crime, somehow he or she will stop paying attention to their arguments.

But one thing we know from the psychology of reasoning is that such arguments do not in fact work (van Eemerem et al., 2012, 2015; Walton 2000). That is, people are not easily swayed by ad hominem rhetoric. The fact that a mass-murderer was a vegetarian and an amateur painter does not convince people that there is anything repulsive in either the diet or the hobby. Maybe Demosthenes’ mother was a prostitute, which does not invalidate her son’s arguments. So people may try, but they rarely succeed.

So why is this persistent?

Explanation 1: meta-gullibility

One possible explanation is that there is an asymmetry between people’s own vulnerability to bad arguments (which is not very high) and their estimate of other people’s vulnerability (very high).

As Hugo Mercier demonstrates in a recent paper, the experimental record shows that it is very difficult to make people entertain  strange or absurd or counter-intuitive beliefs. Humans are just not very easy to persuade of complete nonsense (Mercier 2017). But, as Mercier adds, one thing we often do believe without much evidence is that others will believe just about anything. The only domain where we are really gullible is our estimate of other people’s gullibility. To coin a phrase, humans are not gullible but they seem really meta-gullible.

So perhaps people use ad hominem and other absurd non-arguments because they mistakenly over-estimate their epistemic effects on listeners.

Only politics can make you that willfully stupid

But that may be only part of the explanation, because the use of abusive rhetoric seems uniquely frequent in the political domain. It is in politics that people call an adversary a drunk baboon, as Lincoln was described by the Democratic party, as a supposedly powerful argument against the abolitionist cause.

That is of course not the only damage politics inflicts on people’s intellects. Living among academics, it is of course always a wonder to witness how people who display great sophistication in understanding multiple intertwined factors, or the way some variable modulate the interaction between tow other factors, etc., suddenly turn into four-year olds when they talk about politics. It is a wonder that the same people, who are so careful with the logic of arguments, suddenly get into a passionate refutation that b could possibly imply a, when all you suggested to them was that perhaps a implies b.

Why does it happen specifically in that domain?

Explanation 2: signaling one’s affiliation

The special factor about politics is that a) it seems to be about arguments, for or against particular policies, but b) it is of course mostly motivated by coalitional psychology. The point is to build and sustain an alliance with strong cooperation and diminish the recruitment potential of other alliances, in what is clearly construed as a zero-sum competition for social support (Pietraszewski, 2013; Tooby & Cosmides, 2010).

Seen from this angle, Cathy Newman’s majestic displays of stupidity make more sense. Newman is signaling to her friends or allies that she is so strongly opposed to Peterson and his conservative views that she will use absurd distortion and insulting comments, rather than engage with and discuss any of his arguments. Sure, that makes her sound like a bit of a simpleton. But the point is that people now know very clearly where she stands.

This would make sense, because an interview is always an ambiguous process. A good journalist should get the interviewee to provide the clearest possible expression of their views. But this may be easily mistaken for support. And, as it happens, many journalists owe their jobs as much to partisan affiliation as to reporting skills or interview technique. So this ambiguity may be particularly damaging. Hence the need for signaling.

Signaling would be a fine explanation, but… the rhetoric used by Newman (and other people in such debates) also conveys incompetence, which is not optimal if you want to recruit people. An uncommitted third-party may watch that extraordinary interview and walk away with the impression that Newman’s “camp”, whatever it is, probably does not have good arguments at all.

So, would the signaling advantage of  really bad arguments over-ride their implications about competence? Does this happen only when the “camps” are so clearly antagonistic that trying to appear competent is entirely redundant?

Does it make sense to signal incompetence?

I offer these reflections as conjectures. The cognition and culture community should tell us, whether these two explanations make sense, and whether there is any evidence for the contribution of meta-gullibility and signaling.


Mercier, H. (2017). How gullible are we? A review of the evidence from psychology and social science. Review of General Psychology, 21(2), 103-122.

Pietraszewski, D. (2013). What is group psychology? Adaptations for mapping shared intentional stances. In M. R. Banaji, S. A. Gelman, M. R. Banaji & S. A. Gelman (Eds.), Navigating the social world: What infants, children, and other species can teach us. (pp. 253-257). New York, NY, US: Oxford University Press.

Tooby, J., & Cosmides, L. (2010). Groups in Mind: The Coalitional Roots of War and Morality. In H. Høgh-Olesen (Ed.), Human Morality & Sociality: Evolutionary & Comparative Perspectives (pp. 191-234). New York: Palgrave MacMillan.

van Eemeren, F. H., Garssen, B., & Meuffels, B. (2012). The disguised abusivead hominem empirically investigated: Strategic manoeuvring with direct personal attacks. Thinking & Reasoning, 18(3), 344-364. doi: 10.1080/13546783.2012.678666

van Eemeren, F. H., & Grootendorst, R. (2015). The history of the argumentum ad hominem since the seventeenth century Reasonableness and Effectiveness in Argumentative Discourse (pp. 611-629): Springer.

Walton, D. N. (2000). Case Study of the Use of a Circumstantial Ad Hominem in Political Argumentation. Philosophy and Rhetoric, 33(2), 101-115.


  • comment-avatar
    Hugo Mercier 1 February 2018 (16:28)

    How much does incompetence matter?
    Thanks Pascal for a great post — the lobster bit had me laughing out loud. My intuition would support the coalitional signal interpretation, but I’m a bit unsure about the importance of signaling one’s incompetence. You’re only mentioning this in passing, and I’m not sure it would work in this case. For incompetence to work as a signal, I imagine that the members of the coalition one is trying to woe would have to realize that the members of other coalitions recognize the incompetence. And I’m not sure that’s what happening here (or in other similar cases). My intuition would be that if Newman’s arguments appeal to anybody because of their content, then these people are unlikely to realize how silly they sound to other people. But it would be fun to do experiments on this!

  • comment-avatar
    Thom Scott-Phillips 1 February 2018 (17:43)

    Anne Applebaum at CEU
    A few days ago Anne Applebaum – columnist for the Washington Post and a Pulitzer-prize winning historian – gave a talk here at Central European University on ‘The enduring appeal of the one-party state’. She began by talking about the Soviet party system, and she mentioned in particular how individuals, recognising that pledging allegiance to the party was the best / only way to get in life, would wilfully make statements that were obviously silly; and she seemed to suggest, if I understood right, that they did this because it signalled to others that, as far as party matters were concerned at least, the individual was manifestly willing to ignore / turn off one’s epistemic vigilance. (She didn’t put it in these cognitive terms, of course, but that was the thrust of it.) It’s anecdotal and post-hoc, of course, but it seems in line with hypothesis #2 here.

  • comment-avatar
    Cathal O'Madagain 1 February 2018 (19:35)

    “so you’re saying, Meletus… “
    Socrates did his best to sound stupid, even to insist that he was stupid. And that was no accident: if someone who presents him or herself as sounding dumb, manages to make their interviewee look like an idiot, then all the worse for the interviewee. He also asked leading questions, putting words in his witnesses mouth that draw out idiotic conclusions. Contrary to what Friedersdorf says in the Atlantic, this is not “an unfortunate trend in modern communication…” and Newman likely knows exactly what she’s doing.

    In the case of the lobster, I thought her comment was spot on – what she highlights is that Peterson is constantly on the verge of committing the naturalistic fallacy – that something about how nature is (lobsters have hierarchies too) tells us how we should live (so we should accept that there will be hierarchies). Maybe Peterson isn’t making that mistake, but he gets very close…

  • comment-avatar
    Mauricio Martins 1 February 2018 (23:58)

    Persuade to agree vs. persuade to keep viewing
    I would only add that from a capitalist point of view, Newman’s interview was a success. I am not sure the intention was to persuade the viewer to agree with her *apparent* opinion or to persuade the viewer to keep viewing (and talk about it afterwards).

    This is very much how social media makes its profit, and “normal” media has followed the same strategy. We more likely click/tune in to get an emotional response rather than to be informed/persuaded. And the agents/contents more likely to provide this emotional jolt are rewarded and highlighted. Isn’t this how Trump managed to have the highest media coverage with the least money spent? He tuned his message to the reaction, regardless of the persuasiveness of his arguments.

    Now, this creates a pressure for journalists/politicians to display and nurture these kinds of personas. So Newman’s signalling might not have been to the ideological or political tribe, but rather to the economical one, and the signal was in fact the capacity to entertain and engage.

    In sum, it is possible that Newman was neither dumb, nor she was trying to persuade. Rather the goal might have been to use a set of tools (ideology, disagreeableness and hyperbole) to keep us watching and talking about it, and to display these abilities to her current and future employers. This could have been either an explicit rational goal, or merely a epiphenomenal consequence of the system that rewards these features.

  • comment-avatar
    Joel Mort 2 February 2018 (01:22)

    “Don’t bring that weak a** s**t in here!”
    It might be helpful to use trash talk, as in boastful rhetoric used during sporting events, as an analogue of this interviewer’s behavior.
    1. Meta-gullibility doesn’t seem like a motive here because EVERYONE uses the same language. “I am the greatest”, “You can never beat me”, “Don’t even try it”, “Don’t bring that weak a** s**t in here” etc. Without any evidence to support my claim, this sort of talk doesn’t seem to be really taken seriously by the exhibitors. Every person who uses trash talk can not, by definition, be the greatest. If it is true that people are good at identifying bad arguments then trash talkers know this. Are they meta-gullible? Again, it seems unlikely. They use these phrases knowing that the outcome of the contest is yet to be definitively decided as well as persistently use them when they are trailing in a match or in all matches whether they are undefeated or not.
    2. So the notion that people are signaling something, be it coalitional affiliation or something else, seems more likely. But the coalitional affiliation, in terms of a sporting event is very much in evidence already so it would be difficult to argue that such rhetoric has much effect on members of the group. It might be the case that this type of behavior is meant as a tool for enhancing status, both with one’s group (not in a literal sense – “I am the greatest” – but in the sense of a willingness to espouse the tenets of the group) and in relation to “the Other” (a method of demoralization for example). It seems clear that people with the Obama-bashing license plate “Don’t blame me, I didn’t vote for a monkey” don’t believe or expect others to believe that Obama is a monkey. But this clearly marks the person as uncaring when it comes to compromising the nominal principles of his/her group as well as very effective in provoking “the Other”. This is not the same as meta-gullibility and highly speculative but perhaps helpful in trying to reframe the issue.

  • comment-avatar
    Pascal Boyer 5 February 2018 (18:13)

    More research needed…
    Thanks to all for these comments on incompetence as signaling. I appreciate that, as Joel Mort and Mauricio Martins point out, journalism (especially on TV) requires loud signaling, which of course provides the conditions for the kind of strident silliness evident in the Newman-Peterson case. That and the coalitional imperative would explain, better than simple incompetence, why the journalist does not bother to follow or discuss Peterson’s point. Cathal O’Madagain thinks that Peterson is often “on the verge of” committing a naturalistic fallacy, but… he does not commit it in that exchange, as far as I can see. And the condition for a debate surely is that you should respond to what people have said, not to what you think they might have said if they had not said what they said. Because politics is intensely coalitional, a great deal of political “debate” resorts to the infamous “So you’re saying…”
    Beyond the anecdote, it would be important to understand under what conditions ad hominem and other low-quality arguments appear, and what effects they have, as Hugo Mercier suggests. As far as I know, politics in small-scale groups, e.g., disputes about the use of commons, can be coalitional but does not seem to descend to Cathy Newman levels of distortion. What may be special to modern politics, or to some domains of modern politics, is the combination of coalitional psychology with the intense moralization of affiliation. We need more cognitive anthropologists to explain all that.

  • comment-avatar
    Mika 8 February 2018 (01:12)

    Strengthening ingroup vs. recruiting new members
    Related to Hugo’s comment, I could see the signalling mechanism being more useful if the motivation isn’t so much to recruit others to your side, but to bolster the group that already is on your side – the people who already share your beliefs might not care about what that lobster actually means in the argument, but making the other group look dumb (even if your argument makes you look dumb to them…) could be a way to rouse/pump up your own group.

  • comment-avatar
    chris goble 8 February 2018 (21:37)

    Signalling Strength of a Group Agent?
    I fully agree with the coalition theory presented.

    I also wonder if Group Agent theory (List & Pettit, 2011) might tighten the frame a bit? Big Brothers (group agents) emerge when there is enough biasing between an individual and a aggregation function to make imagined embodiment of a quasi-real group agent an efficient way to predict and understand what group judgments are likely to be.

    Thus, it isn’t just affiliation signalling that is happening, but signalling about the amount of influence the group has on the individual. Extreme influence implies a great deal about the group. From a traditional perspective, it implies the group is robust enough to support costly beliefs (see Atran’s in God’s we Trust). But a Group Agent perspective focusses attention on how an individual is controlled. There can obviously be peer pressure control, but there may also be group agent / moral big brother “control”.

    An interpretation of this difference is the invhiotility of the belief system itself. Is it merely peer pressure or are the morals now bigger than any individual. I’d suggest Newman’s actions may reflect a move to the latter. And thus, the signal may perhaps not just be group affiliation, but a signal that there is a (quasi-real) morality bigger than the group itself.

    To test this you might look for how resistive individuals are to change their belief based on peer changes. If peers change, but the individual doesn’t, then the probability of peer influence/signalling decreases and the probability of group agent influence increases. To filter whether the individual is signalling for the formation of a potential group rather than being biased by a group agent you could test them on transactional exchanges (are you willing to give up belief x it you get lots of supporters or benefits). This would nail down if the belief is sacred. Then you would have to wave a magic wand and hope you could tease out different levels of change consternation for two equally important sacred values, one which is individually controlled, and one which is supposedly controlled by the Big Brother.

  • comment-avatar
    chris goble 8 February 2018 (21:51)

    Active Support or Inevitable Acquiescence
    Maybe the question boils down to:
    -is the individual (Cathy Newman) showing active support, or
    -is the individual showing acquiescence?