In the earlier discussion, however, I was referring to things that impair our ability to solve problems. Now I’m talking about the loss of a much more fundamental ability, namely the ability to know our own mind. Furthermore, in this case we have a much more specific explanation for why introspections mess up our reactions. It’s that we simply don’t have any way of explaining our feelings about jam. We know unconsciously what good jam is: it’s Knott’s Berry Farm. But suddenly we’re asked to stipulate, according to a list of terms, why we think that, and the terms are meaningless to us. Texture, for instance. What does that mean? We may never have thought about the texture of any jam before, and we certainly don’t understand what texture means, and texture may be something that we actually, on a deep level, don’t particularly care much about. But now the idea of texture has been planted in our mind, and we think about it and decide that, well, the texture does seem a little strange, and in fact maybe we don’t like this jam after all. As Wilson puts it, what happens is that we come up with a plausible-sounding reason for why we might like or dislike something, and then we adjust our true preference to be in line with that plausible-sounding reason.

Jam experts, though, don’t have the same problem when it comes to explaining their feelings about jam. Expert food tasters are taught a very specific vocabulary, which allows them to describe precisely their reactions to specific foods. Mayonnaise, for example, is supposed to be evaluated along six dimensions of appearance (color, color intensity, chroma, shine, lumpiness, and bubbles), ten dimensions of texture (adhesiveness to lips, firmness, denseness, and so on), and fourteen dimensions of flavor, split among three subgroups—aromatics (eggy, mustardy, and so forth); basic tastes (salty, sour, and sweet); and chemical-feeling factors (burn, pungent, astringent). Each of those factors, in turn, is evaluated on a 15-point scale. So, for example, if we wanted to describe the oral texture of something, one of the attributes we would look at is slipperiness. And on the 15-point slipperiness scale, where 0 is not slippery at all and 15 is very slippery, Gerber’s Beef and Beef Gravy baby food is a 2, Whitney’s vanilla yogurt is a 7.5, and Miracle Whip is a 13. If you taste something that’s not quite as slippery as Miracle Whip but more slippery than Whitney’s vanilla yogurt, then, you might give it a 10. Or take crispiness. Quaker’s low-fat Chewy Chocolate Chunk Granola Bars are a 2, Keebler Club Partners Crackers are a 5, and Kellogg’s Corn Flakes are a 14. Every product in the supermarket can be analyzed along these lines, and after a taster has worked with these scales for years, they become embedded in the taster’s unconscious. “We just did Oreos,” said Heylmun, “and we broke them into ninety attributes of appearance, flavor, and texture.” She paused, and I could tell that she was re-creating in her mind what an Oreo feels like. “It turns out there are eleven attributes that are probably critical.”

Our unconscious reactions come out of a locked room, and we can’t look inside that room. But with experience we become expert at using our behavior and our training to interpret—and decode—what lies behind our snap judgments and first impressions. It’s a lot like what people do when they are in psychoanalysis: they spend years analyzing their unconscious with the help of a trained therapist until they begin to get a sense of how their mind works. Heylmun and Civille have done the same thing—only they haven’t psychoanalyzed their feelings; they’ve psychoanalyzed their feelings for mayonnaise and Oreo cookies.

All experts do this, either formally or informally. Gottman wasn’t happy with his instinctive reactions to couples. So he videotaped thousands of men and women, broke down every second of the tapes, and ran the data through a computer—and now he can sit down next to a couple in a restaurant and confidently thin-slice their marriage. Vic Braden, the tennis coach, was frustrated by the fact that he knew when someone was about to double-fault but didn’t know how he knew. He is now teamed up with some experts in biomechanics who are going to film and digitally analyze professional tennis players in the act of serving so that they can figure out precisely what it is in the players’ delivery that Braden is unconsciously picking up on. And why was Thomas Hoving so sure, in those first two seconds, that the Getty’s kouros was a fake? Because, over the course of his life, he’d experienced countless ancient sculptures and learned to understand and interpret that first impression that crossed his mind. “In my second year working at the Met [Metropolitan Museum of Art in New York], I had the good luck of having this European curator come over and go through virtually everything with me,” he says. “We spent evening after evening taking things out of cases and putting them on the table. We were down in the storerooms. There were thousands of things. I mean, we were there every night until ten o’clock, and it wasn’t just a routine glance. It was really poring and poring and poring over things.” What he was building, in those nights in the storerooms, was a kind of database in his unconscious. He was learning how to match the feeling he had about an object with what was formally understood about its style and background and value. Whenever we have something that we are good at—something we care about—that experience and passion fundamentally change the nature of our first impressions.

This does not mean that when we are outside our areas of passion and experience, our reactions are invariably wrong. It just means that they are shallow. They are hard to explain and easily disrupted. They aren’t grounded in real understanding. Do you think, for example, that you can accurately describe the difference between Coke and Pepsi? It’s actually surprisingly difficult. Food tasters like Civille and Heylmun use what they call a DOD (degree-of-difference) scale to compare products in the same category. It goes from 0 to 10, where 10 is for two things that are totally different and 1 or 2 might describe just the production-range differences between two batches of the same product. Wise’s and Lay’s salt and vinegar potato chips, for instance, have a DOD of 8. (“Ohmigod, they are so different,” says Heylmun. “Wise is dark, and Lay’s is uniform and light.”) Things with a DOD of 5 or 6 are much closer but still possible to tell apart. Coke and Pepsi, though, are only a 4, and in some cases the difference may be even less, particularly if the colas have aged a bit and the level of carbonation has decreased and the vanilla has become a little more pronounced and pruney.

This means that if we are asked to give our thoughts on Coke and Pepsi, most of our answers aren’t going to be very useful. We can say whether we like it. We can make some vague and general comments about the level of carbonation or flavor or sweetness and sourness. But with a DOD of 4, only someone schooled in colas is going to be able to pick up on the subtle nuances that distinguish each soft drink.

I imagine that some of you, particularly those who are diehard cola drinkers, are bristling at this point. I’m being a bit insulting. You think you really do know your way around Pepsi and Coke. Okay, let’s concede that you can reliably tell Coke from Pepsi, even when the DOD hovers around 4. In fact, I urge you to test yourself. Have a friend pour Pepsi into one glass and Coke into another and try to tell them apart. Let’s say you succeed. Congratulations. Now let’s try the test again, in a slightly different form. This time have your tester give you three glasses, two of which are filled with one of the Colas and the third with the other. In the beverage business, this is called a triangle test. This time around, I don’t want you to identify which is Coke and which is Pepsi. All I want you to say is which of the three drinks is not like the other two. Believe it or not, you will find this task incredibly hard. If a thousand people were to try this test, just over one-third would guess right—which is not much better than chance; we might as well just guess.