Skip to content


I think there’s something wrong this op-ed by developmental psychologist Alison Gopnik, “4-year-olds don’t act like Trump,” and which begins,

The analogy is pervasive among his critics: Donald Trump is like a child. . . . But the analogy is profoundly wrong, and it’s unfair to children. The scientific developmental research of the past 30 years shows that Mr. Trump is utterly unlike a 4-year-old.

Gopnik continues with a list of positive attributes, each one which, she asserts, is held by four-year-olds but not by the president:

Four-year-olds care deeply about the truth. . . . But Mr. Trump doesn’t just lie; he seems not even to care whether his statements are true.

Four-year-olds are insatiably curious. One study found that the average preschooler asks hundreds of questions per day. . . . Mr. Trump refuses to read and is bored by anything that doesn’t involve him personally.

Four-year-olds can pay attention. . . . They pay special attention to events that contradict what they already believe. Mr. Trump refuses to pay attention to anything that clashes with his preconceptions.

Four-year-olds understand the difference between fantasy and reality. . . . Mr. Trump seems to have no sense of the boundary between his self-aggrandizing fantasies and reality.

Four-year-olds have a “theory of mind,” an understanding of their own minds and those of others. In my lab we have found that 4-year-olds recognize that their own past beliefs might have been wrong. Mr. Trump contradicts himself without hesitation and doesn’t seem to recognize any conflict between his past and present beliefs.

Four-year-olds, contrary to popular belief, are not egocentric or self-centered. . . . In fact, children as young as 1½ demonstrate both empathy and altruism . . . Mr. Trump displays neither empathy nor altruism, and his egocentrism is staggering.

Four-year-olds have a strong moral sense. Children as young as 2½ say that hurting another child is always wrong, even if an authority figure were to say otherwise. . . . Mr. Trump admires authoritarian leaders who have no compunctions about harming their own people.

Four-year-olds are sensitive to social norms and think that they and other people should obey them. . . . Mr. Trump has time and again shown his contempt for norms of behavior in every community he has belonged to. . . .

My problem with what Gopnik here is that she seems to be treating four-year-olds as a single kind of thing. But four-year-olds vary! There are lots of four-year-olds out there who are self-centered, who admire authoritarian leaders, who don’t pay attention, who show their contempt for norms of behavior, etc. This sort of thing happens all the time.

From the other direction, I don’t agree with everything Gopnik says about Trump. Yes, he contradicts himself without hesitation but this could well be a tactic he has learned to be effective in business and politics. Unfortunately, tactics that have worked in the past don’t always keep working in the future—but all of this is a far cry from saying that Trump operates without a “theory of mind.” And I disagree with the statement that Trump “refuses to pay attention to anything that clashes with his preconceptions.” What about all that tweeting of his: often that’s in reaction to news that clashes with his view of the world, right?

But the Trump thing is just an aside. My main concern is that Gopnik is saying things about “four-year-olds” as if average behavior applied to everyone. Adults vary in their personalities—Donald Trump is no Mike Pence, and vice-versa—and children vary too. Trump could well share some noticeable traits with many four-year-olds even if their average behaviors are not quite like his.

It’s easy to fall into a deterministic trap but we should avoid it.

Don’t get me wrong here. I’m not saying that four-year-olds are no different from adults, nor am I questioning the value of studying average patterns in how four-year-olds think and behave. I think Gopnik’s research is important and I like that she conducts very specific studies and also considers the larger implications of the research findings in her subfield. This is what science is all about. In this case, I just think that she has been led astray by a tendency to identify the average as the whole.


  1. Colman says:

    I disagree – I don’t think she’s falling into the trap, I think she’s just using common English “shorthand” to make her points. Sure, if you take her literally, maybe you could read it that way (I’m not 100% sure about that either), but I doubt she intends “four year olds care about the truth” to mean there is one behaviour that four year olds express towards the truth.

    But also, you have to take this in context. She’s writing a criticism of Trump by using a not-uncommon method of talking up a thing that’s being used in a negative comparison (e.g. “I’d call her as dumb as a box of rocks, but that would be unfair to the rocks!”). But obviously she’s taking a smarter approach… Anyway, the point is that you have to take her article as a response to the context – when people say Trump is like a child, what are they saying? Are they saying he’s like a particular child? No – his behaviours are what is expected of a child, i.e. average behaviour. So it makes sense then to contradict this by talking about average behaviours.

    • Andrew says:


      Sure, but let’s talk averages. It could well be reasonable to say that Trump lies a lot, and so do 4-year-olds all the time; Trump is easily bored, and so are 4-year-olds all the time; Trump contradicts himself without hesitation, and so do 4-year-olds all the time. So, just on the level of averages, I don’t buy Gopnik’s statements. In each category, she’s taking one aspect of Trump and then mapping it into a particular aspect of 4-year-olds to make her point. I guess one could say she’s “cherry picking.”

      For example, 4-year-olds really do get bored easily! Actually, so do adults. But 4-year-olds, yes, definitely, it happens. So how relevant is it that “One study found that the average preschooler asks hundreds of questions per day”? OK, Trump probably does not ask hundreds of questions per day. But that sort of question-asking must be rare of adults in general! So now we’re just saying that adults are different from kids. Which misses the point as it does not respond to the underlying statement which is not, “Trump is just like an average 4-year-old” but rather, “Trump is more like a 4-year-old than most adults are.” Or, “If you put all adults on a scale and ask which ones are most like 4-year-olds, Trump would be in that category.”

      I’m interested in Gopnik’s research, I’m interested in the studies by others that she cites, and I respect her continuing efforts to connect her academic research to the outside world. But in this case I don’t buy her argument, and I think her use of averages isn’t helping.

      • sentinelchicken says:

        I’m interested to know what you think her argument is? Do you think she is arguing that Trump is like EVERY 4 year-old?

        • Andrew says:


          I’m not really sure what Gopnik’s argument is. As I wrote above, she goes through a list of positive attributes, each one which, she asserts, is held by four-year-olds but not by the president. But when I looked at them carefully, the comparisons didn’t seem quite right: in some cases, I think Trump does have the positive attribute, and, in some, I don’t think four-year-olds typically do. It’s kind of an apples-to-oranges comparison in many cases.

          At this point you might say I’m being picky, but then, what exactly is Gopnik trying to say? What do I make of a statement such as, “Mr. Trump refuses to pay attention to anything that clashes with his preconceptions,” given that Mr. Trump is famous for paying lots of attention to all sorts of people who criticize and disagree with him? Etc.

          Gopnik knows a lot about four-year-olds and I think she could write an excellent article comparing politicians and children. I just don’t think the above-linked op-ed is that thing.

    • Ibn says:

      Colman, the real point is that Gopnik, a person who’s research is useful, is engaging in tribal bad-mouthing of a locally unpopular person by taking a cavalier attitude towards both her own expertise and whether what she says is actually a true statement instead of just snappy – one of the very things she accuses Trump with. What could we say – sad.

      Andrew is just being nice as there is not much of a point in generating personal conflict with someone who’s work you value just because an ill-considered partisan column.

    • psyoskeptic says:

      So what you’re saying is that we should take Galopnik seriously but not literally… where have I heard that before?

  2. zbicyclist says:

    Four year olds will be more mature one year from now.

    But the overall comparison is silly — it’s the type of thing you’d put in a Facebook meme for the amusement of people who agree with you. The NYTimes is giving scarce op-ed space to this? On a day when Trump turns theologian and is giving a speech on Islam?

  3. sentinelchicken says:

    I get your point but you’re being a bit pedantic here. Gopnik’s title and thesis contain an unspoken parenthetical: “4 year-olds don’t act like Donald Trump (on average)”. The variance argument is one I here most often coming from undergrads in intro classes when learning about research that makes them uncomfortable. Sure, people vary across time and place. Yes, you might not have ever fallen prey to the hindsight bias and I’m sure your sense of optimism is always grounded in a very careful and prudent weighting of all the relevant facts and evidence. By all means, these findings and claims do not apply to you because people vary!

    I understand the argument and where it comes from, and I don’t disagree. In fact I’ve taken to spending some time at the beginning of every class to explain that findings and claims are most often based on group averages but that people vary in studies as they do in the real world. Nevertheless, it’s a lazy argument that I would mostly expect from lay people. You have definitely made it clear in your writings about psychological topics that you are not a psychologist. I guess I shouldn’t be too surprised to see you be led astray by a tendency toward naive realism, reading the title and article literally, and spinning off a reactionary hashtag in the spirit of #bluelivesmatter.

  4. Jonathan says:

    I think her piece was kind of off target: it isn’t that people believe Trump is 4 but they have trouble accepting he acts with intention, that he may have a ‘theory of mind’ in which creating chaos occurs intentionally, in which picking meaningless fights is considered a valuable signaling method, etc. She attempts to say, “no, you’re wrong; he doesn’t act like a basic 4 year old” when the actual question or issue is twofold this: does Trump act with a theory of mind and why can’t people think about what that might be. The latter is particularly important because, well, I for one have under-estimated Trump at every turn, even though I’ve read his book and understand his methods, and repeated error speaks to me like a giant flashing warning that my mental conception of what he is doing are affected by my expectations about these contexts (and thus the degree to which my priors are warped by a fairly narrow history – or at least my awareness of that history). To focus on the latter for a moment, take JFK: his behavior was way more out of bounds than Trump’s in many ways, engaging in a sexual affair with a mob boss’ girlfriend, having prostitutes in the White House (including the great story of him having a Secret Service agent hold the girl’s head under water when he was cumming to make his orgasm stronger). But that was ‘private’ (though the press knew about it). If my contextual understanding of the Presidency weren’t so focused on the appearances … I forget Bill Clinton’s sexual harassment of an underling (which it was by definition) and I forget FDR’s mistresses and Eleanor’s not very closeted lesbian relationship – which I don’t put into a category these days of ‘bad’ but it was then – and Presidents who gambled and whored all the way back to Jefferson and his slave children. I don’t even know how to characterize guys like Wilson who embodied many good things but was a retrograde racist even for his era. But there’s an expectation of public dignity, sort of like the Roman Emperor appearing in public in his purple robe – the modern equivalent seems to be a dozen identical Lincoln Navigators – while the orgy goes in the palace. Trump violates that norm but not other Presidential norms, many of which were sanitized in history and which were bluntly ignored at the time by an accommodating press.

    One could argue Trump is our first modern President. Remember the big deal about Obama wanting a smart phone? The iPhone was a year old. Trump not only tweets but engages in social media the way ordinary people do instead of using it as a means of spreading a specific, traditionally Presidential message. It’s difficult to grasp the implications of that because we haven’t seen it before, because so much of the country despised the fact of his election, etc. Before turning to averages, I’d like to say I think we’re under-estimating him again: we don’t grasp, for example, that in his theory of mind picking silly fights is a way of gaining advantage because a) it communicates he will fight and he will fight dirty to win and he might even keep fighting long after you’ve given up and desperately want it to end, and b) it communicates that you don’t know how he will react so you have more trouble planning your moves, so it functions as a strategic disrupter and c) it fits into his general theory of mind that he thrives on the chaos which makes others flail. I think he fell in love with the Kipling poem If – ‘if you can keep your head when all around you are losing theirs’ – and he realize that people don’t keep their heads so he causes them to do that because he keeps his. I give the man more respect these days: if this theory of mind is correct, he’s managed to deflect a vast amount of attention toward bits of nothing much – the daily Trump as soap opera report – while accomplishing huge parts of his agenda. He just put together a huge – rather huuuuge – deal with the Saudis, the Education Dept is turning massively toward incentives like charters, he’s quietly pushing forward toward some repeal of the ACA, he’s fairly deep into the process of de-regulating all sorts of things, etc., all while we’re making fun of him because we think he’s a 4 year old.

    As to averages, my wife is a teacher of this age group and the use of averages in this context is as a yardstick. (She’s actually writing reports now.) But the developmental yardsticks aren’t crap like ‘does he tell lies’ or ‘is she sincerely interested in truth’. Nope: a creative liar is marked out as a good story teller, as a good joker, as smart. Some of the best 4 old stories involve kids saying things to get what they want or to get out doing something. As to ‘interest in truth’, a major yardstick for that age group is the extent to which they observe what’s around them and how they extract information from their observations and then how they interweave that into their lives, into creative play, into their personalities. In 4 year old terms, a ‘sincere’ search for truth is, as my wife says, more that they’re little scientists learning how to observe, not that they’re actually looking for ‘truth’, whatever the bleep that is. A good teacher – see, wife above – can then look at how a kid observes and how that kid internalizes and expresses that. A yardstick of averages is useful for doing this because as a yardstick it automatically says you’re this spot versus that spot. In other words, your objection to the word ‘averages’ appears to miss the usage – a prior conception on your part? – which is not as an average but as a marker, really a series of markers, against which progress is compared. That means the concept in teaching young kids extends to those who are way away from some yardstick average, as is true for the kid she just finished writing about who has a number of language processing issues.

  5. Andrew,

    If Gopnik had prefaced her statements with “Most” or “Nearly all”, would you be OK with the rest? As in, “Most four-year-olds care deeply about the truth” or “Nearly all four-year-olds are insatiably curious”. And then backed these statements up with references to evidence about the distribution and not just about the mean?

    I rush to add that I have no idea if the statements would remain true if modified this way. If they do, then I guess the point would be that Gopnik was being unnecessarily ambiguous; if they don’t, then Gopnik was being misleading.

  6. Cody L Custis says:

    The bigger problem is that she’s used her credibility as a scientist for political propaganda.

Leave a Reply