Thinking (and writing) about how we think

I read Scibling Jonah Lehrer's How We Decide some time ago, but Moveable Type ate my half-finished review, and it's taken me until now to get back to it. You may have seen quite a few reviews elsewhere by now - Adam Kepecs reviewed it for Nature back in April, and to make a long story short, I largely agree with him: Lehrer is a very good writer, but this is not a great book.

Lehrer starts his book with an airplane anecdote, so I'll do the same - although his opening anecdote is about crashing a plane (albeit a simulated plane), so I'm not sure I'd recommend the book for nervous flyers. Anyway, I was reading How We Decide on a plane from Denver to Washington, DC. The businessman sitting next to me ordered coffee, and I felt a feeling of dread come over me. Sure enough, an hour into the 4-hour flight, he knocked a full cup of steaming coffee into my lap. I wasn't burned, but I did have to endure sopping wet jeans and the stench of burned airline coffee until I could finally change clothes five hours later.

Where did that prescient feeling of dread originate? I'm not psychic. I didn't have any hard evidence that my hapless neighbor would knock his coffee over. I've been on lots of planes, and I don't normally worry about the person next to me being unable to manage their steaming beverage. Yet something about this guy made me uneasy, and that feeling turned out to be accurate. His body language - subtle cues that registered in my consciousness as dread - probably reminded me of other people I've seen knock coffee over (most memorably, a frazzled student who managed to saturate his brand-new textbook in the first five minutes of class.) I couldn't tell you what those cues were, but some part of my brain recognized them and translated them into an emotional alarm. If I'd listened, perhaps I'd have gone to the bathroom or switched seats or preemptively told the guy to be careful with his coffee, dammit. But we're not in the habit of making decisions on strictly emotional bases, are we?

This is the theme of Lehrer's book, How We Decide: that good decisions are surprisingly often motivated by emotion, not logic. Furthermore, a dispassionate Vulcan-style strategy of pure reason is not necessarily superior (though any Star Trek fan could have told you that.) Particularly when it comes to behavioral economics, we've learned all sorts of counterintuitive things about how our own minds work. In some cases, they seem counterintuitive because of our cultural preconceptions, in others because of biological shortcomings we have in understanding our own decision-making process. Lehrer takes us on an engaging tour of anecdotes and experiments illustrating the complexity of decision-making, coming down squarely in the moderate middle: we need both emotional and rational thought processes to make the best decisions possible.

When Lehrer is at his best, his storytelling reminds me of Oliver Sacks. Like Sacks, Lehrer has a gift for telling colorful, memorable human interest stories. He keeps both human subject and science in focus at the same time, striking a fine balance between individual quirks and generalizable principles. His real-world illustrations of decision-making - quarterbacking, poker, test-taking, compulsive shopping, flying a plane - are effective and immediately relevant. He even manages to make cognitive psychology experiments sound engaging; no mean feat, to be sure. But in his quest to simplify the science, Lehrer falsely dichotomizes the "rational" and "emotional" brains. That dichotomy makes good storytelling. But when your central thesis is that there really is no clean distinction between your rational and emotional brain, and you need both to make good decisions, the dichotomy muddies your take-home message. What is the bottom line? Trust your emotions, or not?

Despite Lehrer's undeniable skill with language, How We Decide sounds a lot like the familiar tropes of neuroscience journalism - a field in which brains "light up" at this or that, actions, thoughts, and personality types can be traced to specific brain areas, and everything sounds like it's on the cusp of being figured out. Of course it's not that simple. Lehrer doesn't think it's simple, either, but he's hamstrung by the vernacular he uses. It is incredibly difficult to discuss evolving concepts like agency, the mind, will, and desire. Too often, these concepts are oversimplified in the quest to make them intelligible to a lay audience. Consider my previous post about genetics and weight: can one draw a valid distinction between "genes" and "will," when the brain is the product of genes? Couldn't there be genes that powerfully influence will by shaping the brain during development? How do you disentangle these ideas?

In How We Decide, Lehrer constantly pits the "emotional" brain against the "rational" brain - with the kicker that decisions by the "emotional" brain may be surprisingly adaptive. Well, if they're adaptive, then aren't they quite possibly rational after all? (Kepecs makes this point in his review, and I strongly agree with him.) The book's constant use of rational vs. emotional is downright unhelpful in handling something as complex as neuroscience.

In fact, I think Lehrer unfortunately props up a straw man by using "rational" and "emotional" to describe our decision making processes - because this language reinforces the cultural preconceptions that he then turns around and debunks. If emotions are the opposite of "rational," one naturally expects them to lead to poor decisions. One is then more surprised by Lehrer's revelation that emotions sometimes do better than Vulcan logic. But think about it: if emotions weren't adaptive, why would we have them? Should it really be a surprise that good decisions often involve emotional input? Consider the fight-or-flight response. Epinephrine surges, the heart pounds, and sweat breaks out: a perfectly appropriate set of responses to a cougar attack, yet completely inappropriate when triggered by public speaking. Terror is adaptive if it propels one away from imminent danger; from an evolutionary perspective, it's perhaps even rational. Context is everything.

Over and over, the book takes a concept, like the prefrontal cortex as the "charioteer" of the brain, then turns around and undermine it, as if to say "aha! neuroscience has been wrong all along!" But I studied neuroscience, and I never got the impression that the prefrontal cortex was an all-powerful dictator - any more than I had the impression that emotions always lead to bad decisions. Things just aren't that simple when you're mucking about in a squishy ball of tissue firing millions of electrical impulses at itself. I understand that neuroscience has to be simplified somewhat for popular consumption; I've struggled to write press releases that were both accurate and catchy. But to simplify the field and then attack it for being simplistic seems a bit unfair.

This series of passages from two pages of the book's Introduction should demonstrate the vague feeling of whiplash I got as I read it:

Ever since the ancient Greeks, these assumptions have revolved around a single theme: humans are rational. When we make decisions, we are supposed to concsiously analyze alternatives and carefully weigh the pros and cons. In other words, we are deliberate and logical creatures. This simple idea underlies the philosophy of Plato and Descartes; it forms the foundation of modern economics; it drove decades of research in cognitive science. Over time, our rationality came to define us. It was simply put, what made us human.

(I feel the need to invoke that hypothetical Star Trek fan again to rebut this presumption, but I digress.)

There's only one problem with this assumption of human rationality: it's wrong. It's not how the brain works. Look for example, at my decisions in the cockpit. They were made in the heat of the moment, a visceral reaction to difficult events. . .

"There's only one problem with this assumption: it's wrong" is a pretty awesome put-down. Sorry, digressing again.

So how did I make a decision? What factors influenced my choices after the engine fire? For the first time in human history, these questions can be answered. We can look inside the brain and see how humans think: the black box has been broken open. It turns out that we weren't designed to be rational creatures. . Instead, the mind is composed of a messy network of different areas, many of which are involved in the production of emotion. . .

The simple truth of the matter is that making good decisions requires us to use both sides of the mind. For too long, we've treated human nature as an either/or situation. We are either rational or irrational. We either rely on statistics or trust our gut instincts. There's Apollonian logic versus Dionysian feeling; the id against the ego; the reptilian brain fighting the frontal lobes.

Not only are these dichotomies false, they're destructive.

Er, uh. Okay. I agree with the last bit, about destructive dichotomies. But in the end, the entire book is based on a dichotomy between the "emotional" and "rational" brains! Why is this dichotomy okay, when the others are so destructive?

The entire book is an expansion on this progression: simplified historical background + hyperbolic science-is-great language ("For the first time in human history, these questions can be answered. . .") -> reversal of previous simplification -> aha! This is just too much like poor science journalism for me to be comfortable with it. The message is hopelessly muddled, caught in a series of reversals. Consider the prefrontal cortex, which gets a really inconsistent rap through the course of the book. On page 17, Lehrer says,

the rational fourth layer of the brain allowed us to ignore the first three layers. We were the only species able to rebel against primitive feelings and make decisions that were dispassionate and deliberate. But this anatomical narrative is false. The expansion of the frontal cortex during human evolution did not turn us into purely rational creatures, able to ignore our impulses.

Whoa. If any neuroanatomist really thought that a prefrontal cortex made one a "purely rational creature", they would have to stipulate that we all have brain damage, because normal people aren't rational. Just watch the nightly news or get on Facebook for five minutes - or if you live somewhere like I do, look out your window. Clearly, the "purely rational creatures" bit has to be an overstatement. Yet it's presented like a hallowed dogma being overthrown. Hmmm.

What is the take-home message? That the frontal cortex is not a powerful "charioteer;" it does not allow us to resist urges and suppress emotions, that model is false. Right? Here's the end of the chapter, just to be sure:

While Plato and Freud would have guessed that the job of the OFC [orbitofrontal cortex] was to protect us from our emotions, to fortify reason against feeling, its actual function is precisely the opposite. From the perspective of the human brain, Homo sapiens is the most emotional animal of all.

Okay, we get it. But fast forward a few chapters, and the prefrontal cortex is right back to being the rational brain, the all-important charioteer:

"If the emotional brain is pointing you in the direction of a bad decision, you can choose to rely on your rational brain instead. You can use your prefrontal cortex to discount the amygdala."

"The biological reality of the brain, however, is that it's severely bounded, a machine subject to all sorts of shortcomings. This is particularly true of the charioteer, who is tethered to the prefrontal cortex."

"This research can also help explain why we get cranky when we're hungry and tired: the brain is less able to suppress the negative emotions sparked by small annoyances. A bad mood is really just a run-down prefrontal cortex."

"The prefrontal cortex is easy to hoodwink. All it takes is a few additional digits or a slightly bigger candy scoop, and this rational brain region will start making irrational decisions."

So which view is true? Is the prefrontal cortex rational or not? Is it in charge or not? Is it important or not? The answer, of course, is that it's complicated. The prefrontal cortex is not constantly in opposition to emotional input, so much as striving to reconcile and balance between many different inputs, including emotion - a task at which it sometimes succeeds and sometimes fails. But the language chosen by Lehrer to discuss the situations illustrating this complexity embodies the oversimplified dichotomies he shot down in the first place. That's downright confusing.

It's all the more frustrating because Lehrer's engaging portrait of the inconsistent, complex ways in which we make decisions is largely spot-on. Particularly when it comes to economics, we routinely underestimate the role of emotions and overestimate our ability to make evidence-based, objective decisions. But Lehrer argues that much of this new knowledge of decision-making comes from a recent revolution in our understanding of brain anatomy and function (even though his personal approach to the brain seems awfully similar to psychologist William James - he quotes James' 1890 textbook with approval). If so, I was left baffled as to the anatomical correlates of the "Platonic driver and his emotional horses." It sure seems like he's equating the prefrontal cortex with the "rational brain," and the dopamine system with the emotional brain, but didn't he deride that model as insufficient?

In the end, the book is confusing because it tries to fit everything into that Jamesian "emotional" vs. "rational" brain model frame, when it isn't a good fit. Lehrer recognizes that the prefrontal cortex as infallible charioteer is an oversimplification. But the "emotional" vs. "rational" model is also an oversimplification, as demonstrated by the many anecdotes he presents in the book. And while that may go right over the heads of the lay audience, it's like a nasty, itchy splinter to a neurobiologist.

Plus, the take-home self-improvement message of the book is completely unclear to me. Okay: sometimes emotions drive better decisions than logic. But when and why? How do we know when to trust our gut instincts, and if we do, how can we keep from becoming slaves to every paranoid misgiving our brains devise?

Unfortunately, How We Decide doesn't really answer that question. Lehrer exhorts us to think about how we think, but simultaneously warns us not to overthink our decisions. This Catch-22 is the elephant in the room, one he never satisfactorily discusses. The prefrontal cortex, we are told, is not as important as once thought; but later, the prefrontal cortex is the "charioteer" steering the brain. Don't be paralyzed by emotions, we are told; then later, trust your gut instincts. The message is profoundly conflicted, sometimes to the point of incoherence. When Lehrer advises apartment-hunters and couch-shoppers to "think less about those items that you care a lot about. Don't be afraid to let your emotions choose," it may be good advice for some situations, but doesn't he make a big assumption, that you have the needed information for your emotional brain to draw the right inferences? Our emotions aren't tea leaves. They can't predict the right answer based on insufficient information any more than the prefrontal cortex can.

Ah well. Perhaps in the end all that's necessary to improve our decision-making process is some awareness that it is not infallible (in that vein, did you read Malcolm Gladwell's piece in the New Yorker about overconfidence and the financial crisis?) If so, How We Decide is an useful, entertaining read. But it's not going to reframe the way we think about the brain - it just has too much in common with the oversimplifications it criticizes. I only wish it had actively investigated and critiqued the confusing, insufficient language we use to talk about how we think - because that would have been a great service to the field.

Categories

More like this

I think a more useful distinction than between "rational" and "emotional" is between "propositional" and "heuristic". Propositional reasoning is based on explicit logic, while heuristic reasoning is based on rules of thumb.

Yeah - that's the sort of alternative language I'd have preferred. It's not that these concepts don't exist, they're just so rarely used in popular writing about neuroscience.

Thanks for the very interesting review.

A pragmatic approach

I suspect, based on my intuition that we have a representational mind, that "deliberate" decision-making requires us to have a conceptual model of ourself as a thinker and how our mind works. We need such a model, neccessarily simplified, so that we know what strategies or tools based on that model to invoke in our planning.

The rational/emotional distinction (and many variations like it) provides a way for us to divide our decision-making resources into those that emphasize a mindset suitable for effective use of formal methods, with narrowly focused attention on a target or a sequence of steps, and those that emphasize a more receptive mindset more suitable for listening for something new or being ready to respond to subtle internal feedback. I suspect that these kinds of mindset are in some sense rooted in competing systems and at least for most of us are a tradeoff of effective characteristics.

Even though clearly many of us agree that "emotions" and "reason" are inextricable aspects of thinking, there is still in decision making a matter of emphasizing different mindsets which bring different resources with different characteristics to bear. So the question becomes: what is the best way to think of our mental tools, and how can we better develop the ability to use the right ones in the right situations?

I understand that neuroscience has to be simplified somewhat for popular consumption; I've struggled to write press releases that were both accurate and catchy. But to simplify the field and then attack it for being simplistic seems a bit unfair.

Some would say that there are two sorts of experts. Those that get air time, and those that really are experts. Cynics (yes, like me) would add that if you look good then you get the option to be the former. I am not one that ascribes to the idea that all those who oversimplify are doing science a disservice: provided they provide some context that this is their intention, and some references to more detailed/accurate material, it can be a good thing. It does seem in this instance that the author went in the opposite direction though, and a common denominator in these instance seems to be that the author wants to place themselves as being somewhat above their field â I make that as a general observation, because I have not read the book.

If any neuroanatomist really thought that a prefrontal cortex made one a "purely rational creature", they would have to stipulate that we all have brain damage, because normal people aren't rational. Just watch the nightly news or get on Facebook for five minutes - or if you live somewhere like I do, look out your window.

My belief is that most people are instinctual consumers. It not really logic or emotions that drive most people most of the time, but the accumulated memories of those emotions that drive people. So I believe it is less about how people feel at a point in time (although that obviously is a factor), and really a habitudinal thing. This obviously alters when people are placed in new situations, but I think most people do similar things most days.

I would like to see this sort of book written by a team of people.

Someone who tells a good story (regardless of whether they have any scientific training)
A neuroscientist
A mathematician with a background in probability, specialising in decision and risk
Someone with a good broad grasp of computer science, and a strong understanding of the AI branch

Anyhow thanks for the review, and I will probably give this a read at some point.