The New Jonah Lehrer: More Honest, More Boring, Still Spreading Bunk

A Book About Love proves the best practices of science journalism don’t offer much insurance against distortions and half-truths.

Is this love?
A Book About Love doesn’t waver from its author’s standard blueprint. He first asks the straw-man question, “is love really make-believe?”

Photo illustration by Natalie Matthews-Ramo. Photo by Thinkstock.

A Book About Love, Jonah Lehrer’s latest work, starts off with a somber author’s note. “I broke the most basic rules of my profession,” he writes. “I am ashamed of what I’ve done. I will regret it for the rest of my life.”

ICYMI: Lehrer fabricated quotes from Bob Dylan as well as other people; he embellished stories and distorted facts; he plagiarized a lot. Of his three best-selling books, two—Imagine and How We Decide—were deemed so flawed as to be taken out of print. But now he’s back, a better man. Lehrer’s new book has been double-checked, with references for every detail. Where possible, he says, he ran the scientific findings he describes by the researchers who made them. “A few simple procedures,” his introduction promises, will “prevent these mistakes from happening again.”

So, what’s a Jonah Lehrer book without the fibs and cheats? It’s just like every other Jonah Lehrer book, but less fun to read.

A Book About Love doesn’t waver from its author’s standard blueprint. First ask a straw-man question (“is love really make-believe?”), then tell a famous story (“a boy goes to a party … his name is Romeo”), relate it to some data from an aging academic (“a spry 80-year-old” who “talks slowly, always pausing thoughtfully”), and finish with a platitude (“our lives become the sum of everyone we have loved”). Repeat, repeat, repeat.

In the bad old days of Evil Jonah Lehrer, this template sometimes had a bit of buoyancy. A cloud of stolen and invented facts would float around a gripping anecdote—like Tom Brady’s final scoring drive in the 2002 Super Bowl—and imbue the scene with unearned drama. Now his method’s made of lead. Good Jonah Lehrer wants to soar again, but without recourse to his former, irresponsible panache, the book remains a dull concatenation of research studies and the bromides they inspire.

“It’s not enough to describe the hormones of Romeo, or the fMRI results of Juliet,” Lehrer writes, or “to reduce the emotion to a set of wires and ingredients.” Yet here he’s done little more than switch out some neuroscience for psychology, one reductive brand of science for another. It seems to me the jargon swap has only made things worse: Where the Lehrer of 10 years ago might have talked of love in terms of chemicals and neural cortex, the new one uses cottonmouth-inducing phrases like habituation, attunement, companionate love, limerence, and sexual communal strength. It’s just about as dreary as it sounds.

Still, let’s give credit where it’s due. This bloodless book may not appeal to everyone, but at least it follows the first and most important rule of journalism: Nothing’s been made up. Lehrer’s latest work may be somewhat boring, but at least it isn’t bullshit. Right?

Here’s the disheartening truth: Despite its thorough vetting, A Book About Love has its share of suspect claims and wonky data. Lehrer may have given up on outright fraud, but he’s still prone to spreading bunk. At this point, what we’re dealing with is less a Jonah Lehrer problem than a science journalism problem. Lehrer’s new book proves that even our field’s best practices don’t offer all that much insurance against distortions and half-truths.

Take Lehrer’s endnotes, 566 of them in total, apparently a bulwark to keep mistakes from slipping in. These notes, piled high like sandbags, function more like security theater than legitimate quality assurance. Most of his notes cite books instead of published papers, even when the text describes findings from a single study. At best, that turns a reader’s search for sources into a treasure hunt through nested references. At worst, it’s a dead end.

I was surprised to read, for instance, that “nearly every language uses a broken heart as the representation of a failed love affair,” and that one of the few exceptions to this rule is Indonesian. Is that really true? His endnote led me to The Little Book of Heartbreak, a 2012 book by Meghan Laslocky. She makes more or less the same assertion, but without the same degree of certainty, and includes a list of eight examples from different languages. (Six employed a broken-heart metaphor; two did not.) Laslocky gives no further references, so I emailed her a question about the “nearly every language” formulation. “Well, that’s taking slight liberties with what I wrote,” she answered. “What I wrote was intended to be a somewhat informed extrapolation, not a declaration of universal fact.”

In other places, Lehrer either doesn’t bother with a reference or else supplies one when it’s hardly needed. Here’s my favorite, from Chapter 2:

I think here of Where the Wild Things Are, Maurice Sendak’s masterwork.39

Checking the endnote, I found:

39            Maurice Sendak, Where the Wild Things Are (New York: HarperCollins, 1984).

But even the best citations can still conceal biased or misleading research. In A Book About Love, Lehrer makes a lot of big claims about the value of “attachment.” Responsive and sensitive parenting leads to secure attachment, which in turn brings “a raft of benefits” to their children, he writes. He cites data showing that people who grow up under those circumstances end up healthier, richer, and better educated than those who don’t. “According to the data,” he writes, “nothing counts more than our attachments. Nothing even comes close.”

Jonah Lehrer
Jonah Lehrer.

Timothy White

The facts check out, at least insofar as Lehrer’s claims derive directly from published academic work. In particular, he draws many of his figures from a well-known pair of studies, one that began in 1975, based on 267 high-risk families in Minnesota, and another that kicked off in the late 1930s and followed the lives of 268 male Harvard undergraduates. Citing data from the Minnesota study, Lehrer notes the extraordinary fact that 32-year-old adults who, as infants, had not been securely attached to their parents were almost three times more likely to say they had a chronic illness.

The fact that this statistic came from scholarly work—in this case a conference poster rather than a published paper—doesn’t necessarily make it true. Even when he’s behaving like an ethical, honest science reporter, Lehrer is only as reliable as his source material. And in this case, his source material has limitations.

I called up Jay Belsky, an expert in child development at UC–Davis and a proponent of attachment theory, to ask about the Minnesota study. He told me it was an important data set, but noted that the study followed a very specific group of kids in a very specific time and place (as all studies do). Furthermore, he said, “it’s a study where the findings [on attachment] are stronger than they are in much of the similar literature.”

Indeed, meta-analyses of the research on attachment theory—which attempt to combine results from many studies instead of focusing on just one—find that Lehrer’s “raft of benefits” may in fact be slight. Recent evaluations of the literature typically find a “small to moderate effect” for early child-caregiver attachment on long-term outcomes such as relationship quality. In other words, attachment does make a difference, but the studies that Lehrer cites (fairly and accurately, no doubt) could overstate the case.

The same uncertainty arises in Lehrer’s treatment of the Harvard study, and its former head, George Vaillant. Men who lacked a warm relationship with their mothers were 2½ times more likely to experience dementia down the road, Lehrer says, citing Vaillant’s work. Men from loving homes earned 50 percent more money. Men who were better at attachment lived longer. “The capacity for love turns out to be a great predictor of mortality,” Vaillant tells him in an interview.

Again, this is honest journalism—Lehrer draws his claims from actual research. But when I looked at other accounts of Vaillant’s work, I found a host of quirky findings. It turns out that among the Harvard men, having a warm father correlates with enjoying one’s vacations more, and being more satisfied with life at the age of 75. Having a warm mother had no effect on these variables. Meanwhile, a mother’s warmth in childhood seemed to make the Harvard men more effective at work late in life, whereas their fathers’ warmth did no such thing. With so many variables in play—their mothers’ warmth, their fathers’ warmth, their chances of dementia, their enjoyment of vacations, their happiness at any given age, and many other measures taken all throughout their lives—I couldn’t help but wonder if these associations might be spurious. Could they be nothing more than arbitrary patterns in the data that happened to comport with Vaillant’s pre-existing, psychoanalytic theory of the world? It would take more than Lehrer’s dutiful transcription—telling us what Vaillant says—to find out.

Fact-checked, straight-laced science journalism may also fail to catch distortions of the evidence, if those distortions come from researchers themselves. Lehrer spends several pages on the work of Angela Duckworth, whose best-selling book on “grit” came out in May. In that book and elsewhere, Duckworth has overstated the predictive value of grit; now Lehrer makes similar, exaggerated claims. Duckworth “demonstrated the importance of grit in loving relationships,” he writes, by collecting data from 6,362 middle-age adults who were either married at the time, or had been married at one point in their lives. Her study found that “gritty men were 17 percent more likely to stay married.” From this he concludes that love requires inner strength and that grit will shape our closest relationships.

But as Iowa State University psychologist Marcus Crede pointed out to me by email, that statistic is a bit misleading. First, it glosses over the fact that when Duckworth looked at all the people in her study, grit had no effect at all. (It only seemed to matter when the women were excluded from analysis.) Second, by focusing on relative instead of absolute benefits, it makes the effect seem bigger than it really is. Looking at the paper, it seems that among all the men in Duckworth’s data set, about 80 percent were still with their spouses. Her statistics show this rate was slightly higher for the grittier men, at 82.4 percent. That is to say, the gritty men were just a few percentage points more likely to stay married—a much smaller effect than Lehrer gives, quoting from the paper.

Analogous mistakes appear elsewhere in A Book About Love. In one example, Lehrer touts a study showing that children with highly involved biological fathers are “43 percent more likely” to get very good grades in school. His endnote shows he pulled that figure from a booklet published by the U.S. Children’s Bureau in 2006. But a closer look at the research cited in that booklet—a 2001 study from the Department of Education—suggests it would be more accurate and more clear to say the likelihood that these children will get good grades goes up by 9 percent, not 43 percent.

These small distortions aren’t fatal, but they add up to a bigger, more important point: Science journalism can go wrong in many different ways, very few of which require explicit acts of fraud. The same is true of science. While con men do emerge from time to time—think of Diederik Stapel, with at least 58 retracted papers—they’re more a sideshow than a threat. More distressing are the researchers who play by the rules and still end up with second-rate and meaningless results. It’s the rank-and-file scientists, not the shameless frauds, who brought on the replication crisis. They’re incentivized to publish more instead of better papers, which leads to sloppy research practices. Likewise, it’s the rank-and-file science journalists, not the sui generis liars, who produce the largest volume of reporting, and thus the greatest number of half-truths. Our own incentives are to blame: The quest for cleaner, more surprising, better-selling stories leads to sloppy work—with studies cherry-picked to make a point, or statistics taken out of context. Fundamental rules of writing (never make things up, never plagiarize) won’t address that market-driven problem.

But I can’t let Jonah Lehrer off the hook completely. In his introduction and conclusion, he describes A Book About Love as the story of his own redemption, and of his search for meaning in the midst of abject misery. (At one point Lehrer more or less compares himself to the Holocaust survivor Viktor Frankl, who also grew from “a suffering man into a loving man.”) “People change,” he writes. “That simple fact is one of the great themes of the longitudinal studies in this book.” Has Lehrer really changed? Has Evil Jonah turned into Good Jonah? He’s daring me to ask the question, so I guess I ought to answer.

I don’t think there are fabrications in the book; I think it’s honest in that sense. But I did find at least a hint of Lehrer’s old pathology—his self-destructive impulse to cut corners.

Take, for example, his account of Harry Harlow’s classic work from 1958, in which baby monkeys were reared with dolls fashioned from either wire mesh or terrycloth. Lehrer’s latest version of this story is pretty much the same, paragraph by paragraph, as the one he used in How We Decide, seven years ago. (In turn, that older version was pretty clearly cribbed, with only passing reference, from Deborah Blum’s 2002 book, Love at Goon Park: Harry Harlow and the Science of Affection.)

Here’s a portion of the recycled text:

A Book About Love (2016)

To minimize the spread of disease, Harlow kept the animals in individual cages, away from parents and siblings. … The resulting litter of primates looked bigger and healthier than their peers in the wild.

But the appearance of these young monkeys hid a devastating loneliness. Because their short lives had been defined by total isolation, they proved incapable of even the most basic social interactions. In the company of other primates, they appeared nervous and withdrawn, staring at the floor. … For the Wisconsin scientists, these troubled primates demonstrated the developing mind needed more than proper nutrition. But what did it need?

How We Decide (2009)

In order to minimize the spread of disease, Harlow never let the babies interact with one another. The result was a litter of primates that were bigger and stronger than their peers from the wild.

But the physical health of these young monkeys hid a devastating sickness: they had been wrecked by loneliness. Their short lives had been defined by total isolation, and they proved incapable of even the most basic social interactions. … When they encountered other monkeys, they would shriek in fear, run to the corners of their cages, and stare at the floor. …

For Harlow, these troubled baby monkeys demonstrated that the developing mind needed more than proper nutrition. But what did it need?

This is so imprudent as to be insane. In 2013, Lehrer’s publisher yanked How We Decide from bookstore shelves and offered refunds to his readers. Now, in his comeback effort, Lehrer has copy-pasted from his own besmirched and error-ridden work.

The passage above hasn’t changed that much. Neither has he.