Like little stars.
Several years ago the neuroscientist Jaak Panksepp began researching that most elusive of aesthetic experiences: the “chill” that you feel in your spine listening to moving passages of music. Panksepp’s studies make a convincing case that the shiver of pleasure we experience while listening to our favorite music is the release of endogenous opioids, the same molecules implicated in social bonding, parental love, the “runner’s high” — and, of course, in narcotic drugs like heroin and morphine. Panksepp has found that animals appear to have chill responses to music as well. In one widely cited study, he played dozens of records to chickens attached to equipment designed to record their shivers of pleasure. (The chickens turned out to have the strongest positive response to the late-era Pink Floyd record “The Final Cut.”)
Now imagine taking Panksepp’s experiment one step further: instead of a chicken’s brain listening to Pink Floyd, imagine peering into a brilliant composer’s brain as he or she dreams up a new composition. Thus far, most brain-imaging research has focused on normal brains and on brains that suffer from some kind of disability. But we also have the opportunity to scan brains that are unusual in the sense of being unusually gifted. What vista into the world of inspiration will this open up to us?
I don’t know firsthand what moments of true musical inspiration feel like. For me, inspiration revolves around words and sentences, and not melody and harmony. I’m not imagining myself to be a literary Stravinsky, but stringing text into narratives and arguments has been the most fluid of my mental faculties for as long as I can remember. Could brain science have something useful to say about this talent? I wanted to know what was happening in my head when a new insight arrived, usually half formed and barely grasped: a vague connection between two ideas, a new way of introducing a troublesome chapter, a phrasing for a sentence. For reasons probably both genetic and cultural, I am not much of a mystic, but these flashes of insight were the closest thing I had to the experience of mysticism. These sparks were the transcendence that Keats sought when he commanded us to “open wide the mind’s cage’d doors.” An idea shoots in front of my mind’s eye seemingly out of nowhere. Where did it come from?
How extraordinary that we can even begin to answer this question! We can only speculate where new ideas come from in the sense of their evolutionary roots, and we don’t really understand how the firing of neurons creates the rich subtleties of ideation. But we can determine, with split-second precision, the parts of the brain that are active in the creation of a new idea. We can map mental processes as ephemeral as having a hunch. On a fundamental level, we can tell where the hunch comes from. All it takes is a brave, nonclaustrophobic subject and a $2 million magnet.
I thought I was precisely that brave, nonclaustrophobic subject until they strapped my head down to the mechanical gurney, and I began sliding into a two-foot-wide tube, with only a mirror the size of a playing card supplying me with a glimpse of the outside world.
There’s no better way to say it: I was having my head examined. Mechanically speaking, the exam was being conducted by a five-ton GE Twin-Speed fMRI scanner. My guide through the world of advanced brain scanning was Joy Hirsch, director of Columbia University’s Brain Imaging Group, who had graciously offered to help me in my pursuit: to see the brain, from the inside, as it comes up with a new idea.
A week or so before my appointment with the scanner, I suggested an experimental structure to Joy: we would begin with my reading a series of nonsense sentences, followed by my reading someone else’s prose, and then I would read a passage of my own work — a passage from this book, in fact. In reading my own passage, I hoped to spur one of those imaginative leaps: something about the words would make me think of a new line to add, or a new way of phrasing the idea, or some other unpredictable insight. If all went well, the machine would take a snapshot of that idea forming in my head. fMRI scans can capture subtle shifts of activity within a three-dimensional model of the brain by measuring levels of oxygenation in the blood of nerve cells. It is not a perfect view by any means — you have to have roughly 500,000 neurons active in an area for the scan to register them — but it is as close to pure vision of the mind’s inner life as current technology allows us.
When I arrive for my session, Joy and I sit down in her office. She explains that each stage of the experiment will involve three sections of forty seconds each: rest, activity, rest. The scanner will start up, and I’ll do my best to think of nothing for forty seconds. Then the stimuli will begin — the flashing checkerboard or the text — and I’ll process that for another forty seconds. And then I’ll think of nothing again. Each 120-second stage will be repeated twice.
As Joy lays out the sequence, I start to worry that I won’t have time to actually think while in the machine; I don’t want to spend the whole forty seconds reading, particularly once we get to my own words. I want to have the words trigger some new idea or association in my head. So Joy agrees to make a last-minute addition: a final stage during which I’m shown a single sentence from my book and given the entire forty seconds to ruminate.
Then Joy walks me through the risks. “We’re looking at your brain here. So there’s a very small chance that we might see something in these scans, some abnormality.”
I nod. “You mean a brain tumor.”
“Sometimes when we do work with experimental subjects — people who come in to help with our research, and who don’t have any symptoms — they say, ‘If you see something in there, don’t tell me.’”
“Hey, if you see something in there that you don’t like,” I smile ruefully, “by all means let me know.”
Then she moves on to the dangers associated with the scanner itself. “It is a fundamentally safe procedure, noninvasive.” I think of a news story from a few years back in which hospital staff had left a metal trash can in the room with an fMRI. When they began scanning a patient, the magnetic field triggered by the scanner being switched on turned the trash can into a lethal projectile that killed the guy instantaneously.
I choose not to bring this up.
A minute or two later, we walk over to the fMRI room. The machine itself looks like an oversized clothes drier — about ten feet high with a huge GE logo embossed above the hollow tube at its center. I lie down on the mechanical gurney, and the technician gently tapes my forehead to the cradle at the end, hands me a pair of earplugs.
And then I’m in.
Being inside an fMRI machine is definitely more unpleasant than it looks to be from the outside. The space itself is astonishingly small, and the sense of being encased in a huge piece of machinery unsettles more than you think it will. For my experiment Joy and her team have placed a small mirror above my eyes that enables me to see a sliver of the world outside the tube. This sliver lets me read the text that they’ve projected onto a screen, but it also prompts a surge of nausea as I first enter the scanner.
The fMRI machine is capable of capturing two types of images: conventional MRIs that are higher resolution but don’t show specific activity in the brain, and then lower-resolution “functional” images that show the brain actually thinking. (Functional MRI images work because active areas of the brain require an increase in oxygenated blood, which creates a small but detectable disturbance in a magnetic field.) We begin with a round of conventional images of my brain, during which time the machine rattles ominously around my head. Then we move on to our little experiment, starting with the checkerboard pattern.
You can easily tell when the fMRI is in its “functional” mode because it emits an uncomfortably loud, high-pitched, pulsing tone. (Hence the earplugs.) When you’re actually inside the scanner, it sounds like a truck backing up into your head. For the first forty seconds of “rest,” I find myself incapable of thinking about anything other than the excruciating noise. When the flashing checkerboard appears on the screen, it occurs to me that this is like attending some kind of demonic performance-art happening — a tiny, cramped space with strobing black-and-white images projected onto a screen, all accompanied by monotonous, piercing rhythmic tones.
But by the second iteration of the checkerboard stage, I start getting accustomed to the noise and the physical enclosure. I can see Joy smiling at me through the mirror, and the sound becomes more background noise than anything else. In fact, I feel comfortable enough that I start having difficulty shutting off my brain during the “rest” periods. First, I find myself thinking about ways that I could describe the setting, shaping the story of my fMRI experience while my head is still stuck inside the device. When I catch myself doing this, I smile in my dark tunnel. It occurs to me that this is one of those small examples of the brain’s miraculous resilience and flexibility: you stuff your brain into a physical situation that should by all rights overwhelm it, and you tell it explicitly not to think of anything, and yet still it churns away in spite of everything. You couldn’t imagine a more hostile environment for free associating, but here my brain was riffing away, as though I were daydreaming in the shade of an oak tree.
Then I’m reading. It ends up being easier to focus on my own words, but there certainly isn’t time to ruminate. As we finish that stage, I think to myself that I’m glad we added the rumination “bonus round.”
I’m glad, but I’m also getting tired. I haven’t moved my head more than a centimeter in around twenty-five minutes, and the space is starting to close in on me. When the first frozen slide of text arrives on the screen for the rumination stage, I feel like I’ve been caught off guard. “Shit!” I say to myself. “Now I have to think of something.” For forty seconds of this $2 million machine’s time, I think of absolutely nothing worthwhile. I think about trying to think about something. If there is a cognitive version of flailing, this is what I do for the first scan.
But when the second round — the last run of the entire experiment — arrives, I’m prepared. I decide to let my brain do what had come naturally to it throughout the experiment. I’ve already started down the road of describing the experience in the scanner — why not take this last round and actually start working out the language? And so when the text flashes up on the screen, notifying me that the forty-second rumination period has begun, a sentence starts to take form in my head. I am writing.
The words I string together in the fMRI are roughly the same words you encountered a few paragraphs ago describing the resilience of the brain in the most uncomfortable of situations. The general idea arrived a few minutes earlier, but the exact phrasing originates in that last session. The specific sentence, of course, is incidental; what makes it interesting is that Joy Hirsch and her fMRI are watching as it forms in my head, as my brain pulls the words out of the nothingness and makes them into something fixed — sturdy enough to remain intact until I sit down at my computer several days later to type them.
A few days pass, and Joy sends an e-mail to let me know that the results are in. “You’re going to like this,” she writes temptingly. The next afternoon I take the A train up to 168th Street, and Joy and I sit down at a conference table to spend some quality time with my brain.
Joy has assembled a collection of about forty color printouts, each displaying four images of my brain at work. The images are overhead views, and each one is a “slice” of my brain, starting with the brain stem, at the very bottom, and ending with the tip of the cortex. For each stage of the experiment — there are four in total — the fMRI has captured twenty-five slices of my brain going about its business. That business takes the form of changes in blood flow to different regions; the scanner first looks at my brain during the “rest” periods, then during the “activity” period, and it records any salient differences between the two. These images let you see the areas that are relevant to a particular task, and shut out the background processing that the brain is always doing. My brain stem, for instance, was steadily plugging away maintaining my breathing pattern — along with many other mission-critical operations — but that area doesn’t light up on the scan images because those patterns didn’t change during the experiment.
Areas that do show noticeable changes appear on the images as a cluster of bright yellow pixels, fading out to orange and red at their peripheries. The images look strikingly like the Doppler radar images you see on the Weather Channel. (If you blur your eyes a little, you might think that yellow patch on the image was a thunderhead, not a brainstorm.) The image is projected over a grid with numbers running along each axis. The numbered grid and the slices create a three-dimensional system of coordinates, the latitude and longitude of neuromapping. The grid is made up of small cubes called “voxels,” and each voxel has a specific address.
Joy begins by laying down the twenty-five slices for stage one of our experiment, the dreaded checkerboard. The pattern of activity is immediately visible, even to my untutored eyes, mostly because there’s literally nothing going on in 95 percent of my brain. Only a thin band wrapping around the back of my head, roughly at ear level, glows yellow.
“We know that the flashing checkerboard is a very salient stimulus for just the visual processing areas of the brain,” she says. “And that’s exactly what’s happening here.”
She points to the yellow band: “This part of the brain is all primary visual cortex. What’s unique about this is that this activity doesn’t get out of the occipital lobe — and nothing goes on in the frontal lobes. Nothing. This is just as exclusively visual as you can get.” We both start to laugh. “Your brain is doing the minimal amount it has to do to sit there and look at that stupid checkerboard!”
Looking at those blank areas on my mental map reminds me of all the times that someone had gravely explained to me that we only use 10 percent of our brains, and then waxed rhapsodic about how smart we’d be if we could tap 100 percent. Of course we only use a small percentage of our brain at any given time — and it’s a good thing, too! Your brain has dozens of dedicated tools, most of which aren’t relevant to whatever it is you’re focusing on right now. If your visual cortex keeps kicking into overdrive as you’re trying to memorize a speech, the words won’t stay in your head as readily. Only using 10 percent of your brain is a sign of efficiency, not underachievement. Arguing that we’d be better off with 100 percent is like raving about how great Shakespeare would have been if he’d managed to use all twenty-six letters in each of his words, instead of a small fraction of the alphabet.
There’s something in Joy Hirsch spreading out the images on the table that brings to mind a tarot card reader, but there’s nothing mystical in her analysis. I find myself thinking, This person I barely know has ventured inside my head in a way that no one has ever ventured before. That’s why the hall-of-mirrors interpretation feels wrong to me. It’s not an endless simulation I’ve entered into here, but rather something that feels authentic, even intimate.
Thus far all the images we’ve examined have been composite sketches: each stage included two runs, and so the images are a combined look at activity over the two of them. But with the rumination round, I had asked Joy to look at the two runs separately, because I had fared so poorly the first time around and because in the final run of the day, I had managed to get my brain exactly where I’d wanted it to be for my forty seconds in the spotlight.
The images from those two sessions do not disappoint. In the first run, small spots of activity are scattered across my brain, mostly in red voxels (suggesting less activity than the yellow). There’s little shape or symmetry to the map; my brain looks cluttered. But in the second run, what jumps out at me immediately is how silent most of my brain appears. Only the language centers light up with any intensity, along with a sharp yellow rod at the center of my brain, extending up to the very top of my cranium. There’s very little visual activity, and almost nothing from the eye-movement regions.
“There’s a concept of efficiency that has emerged in the neuroimaging community in the last few years,” Joy says. “It’s basically that when there’s a task that the brain is having difficulty doing, the pattern looks very distributed, like this here.” She points to the cluttered image of run number 1. “This was not an efficient action — as opposed to here, where the specific tools of the brain are contributing in an efficient way to the task at hand.”
“You really look like you got your act together here.” She’s pointing to that bright yellow dot on the upper images of run number 2. “Here’s more evidence of that — look at this very focused medial frontal gyrus. This is one of the most distinguishing characteristics of this scan — this is a very high-level executive function of the brain, and you can see it running like a pole all the way down to the cingulate. I think that the medial frontal gyrus is important in coordinating different activities in the brain, reaching for the right tool at the right time. In this last scan, the entire structure — not just a part of it — is active.” In Joy’s phrasing, my language areas were perfectly “robust” during these inspired forty seconds, but they didn’t turn out to be the most interesting element of the image. It was the overall orchestration, the clarity of the pattern, that stood out, the lack of mental clutter.
What had I been hoping to find? I thought about this on the subway ride home. In the crudest sense, I suppose I thought that my skill at stitching words together in my head might turn out to have its own modulelike presence in the scan: a distinct patch of neurons devoted to imagining sentences. If the brain is filled with all these modular tools, then somehow it seems logical that tasks you’re good at should have some visible presence on the brain map. Sometimes this is the case: Einstein’s brain had unusually large inferior parietal lobes, which we think gave him his extraordinary spatio-logical skills. (He famously solved problems as images in his head weeks before he could turn them into working equations.) Such a skill most likely would have shown up directly on an fMRI: a person gifted in spatial intelligence shows more activity in regions of the brain dedicated to spatial processing.
But in my case, the scan revealed something quite different. (I’m no Einstein, as it turns out.) There was no special module. What caught Joy’s eye in the final rumination scan was not a specific region, but the overall pattern of brain activity. The tools in the toolbox weren’t particularly impressive, but the toolbox itself was well organized. In fact, the only specific region that seemed to be at all above average was the one responsible for coordinating activity in other regions. Perhaps the most telling thing about my brain map was what didn’t show up on the images: when I was focused, there was almost no activity in areas that weren’t related directly to the task at hand. Compare that to my episode of cognitive flailing in the first run of the rumination stage: on that scan, there’s hardly a discernible pattern. It’s mostly noise, and little signal.
I have no idea how replicable my fMRI results would be if I tried the exact experiment again, and it’s unclear whether that pattern of organization — with its strong medial frontal gyrus and its many silent regions — holds true for my brain generally, or just for this little snapshot. But I suspect there is a larger truth nestled in that last fMRI image, one that has begun to change the way I think about people I know. I suspect that the world of talent is made up of two kinds of brains: some that have specific modules that are unusually good at their job, and some that are unusually good at keeping all the different modules organized. Both types of brains come across to us as talented, as intelligent, but I think the types are different enough that you can learn to recognize them if you know what to look for. We all know people who have dazzling skills: they can sit down at a piano and pick out a tune they heard last week; they can calculate interest rate payments in their head; they can actually understand quantum mechanics. But we also know people whose brains seem gifted in a different way: no stunning, off-the-chart skills, but a general competence and efficiency, with very little noise complicating their signal.
My dad used to say to me during my high school years: “You’re not a rocket scientist, but you’re smart and you’ve got a lot of talent.” I used to bristle at the remark. (If I wanted to, maybe I could be a rocket scientist!) But now I think he was onto something. I’ve met rocket scientists — and astrophysicists, and programming wizards, and architectural geniuses — and I don’t possess anything like what they’ve got mentally. I don’t have their special gifts. But those fMRI images made me think that perhaps I have something else, a little less dazzling, but nothing to be ashamed of either. Maybe I have a well-orchestrated brain — with no world-famous soloists but a nice sound nonetheless. In a sense, this is what my dad had been trying to say, in slightly different language: I was talented in an orderly brain kind of way, not a supermodule kind of way.
It was only one experiment, but the machine had given me something that machines don’t normally deal out: a hunch about myself, and maybe a larger hunch about people in general. I’d been dreaming for more than a year of capturing my brain as it came up with an idea, and thanks to Joy and her uncanny device, I’d managed to catch precisely that glimpse. The results were mesmerizing and remarkably legible, even to my untrained eyes. But they didn’t provide unequivocal answers or magic bullets. They were more like clues. Seeing my brain come up with an idea had given me another, more interesting idea, one that still reverberates in my head as I write. Wouldn’t it be nice to have a scan of that?
- – - – - – - – - – - -
Copyright © 2004 Steven Johnson, reprinted with permission of Simon & Schuster.
Steven Johnson is the author of "Emergence: The Connected Lives of Ants, Brains, Cities, and Software" and "Interface Culture : How New Technology Transforms the Way We Create and Communicate."More Steven Johnson.
Like little stars.
World's best pie apple. Essential for Tarte Tatin. Has five prominent ribs.
So pretty. So early. So ephemeral. Tastes like strawberry candy (slightly).
My personal fave. Ultra-crisp. Graham cracker flavor. Should be famous. Isn't.
High flavored with notes of blood orange and allspice. Very rare.
Jefferson's favorite. The best all-purpose American apple.
New Hampshire's native son has a grizzled appearance and a strangely addictive curry flavor. Very, very rare.
Makes the best hard cider in America. Soon to be famous.
Freak seedling found in an Oregon field in the '60s has pink flesh and a fragrant strawberry snap. Makes a killer rose cider.
Ben Franklin's favorite. Queen Victoria's favorite. Only apple native to NYC.
Really does taste like pineapple.