High art is fine, but it isn’t enough. Dive bars, sports events, mountains, valleys, ranches, and beaches all go well with opera, lecture series, and theatre. I’ve learned a lot from hanging around the Oakland Coliseum parking lot on autumn Sundays. Concerts are terrific of course, and good places to blow off steam too. I’m taking my first meditation class next week; it takes a conscious effort to push down my own cynicism. Writing this book even got me to listen to jazz, and I think I figured out why I hate it.
One sure-fire way to both reduce idea prejudice and acquire greater answer resolution is to hold fewer opinions. Sure, our opinions help us distinguish signals from noise, but they’re essentially idea prejudice. They inhibit us from entertaining vast regions of idea space.
I feel very strongly that people should be less opinionated.
10
STARING AT A PICTURE WITHIN A PICTURE WITHIN A PICTURE
MOST OF THE IDEAS WE’VE SEEN HERE WILL SURVIVE to some extent and in some form as neuroscience matures. Neuroscience in 2016 is at a state similar to where physics was in the middle of the nineteenth century before electricity and magnetism had been combined into electromagnetism but after Newton, Leibniz, Gauss, Fourier, Laplace, Hamilton, and hundreds of others had developed the symbolic tools and techniques that would lead to the huge advances from 1850 to 1950.
Neuroscience is right around the point where Starla first saw the rainbow, the instant when she realized that there is more than on/off, light and dark. Neuroscience has just discovered its rainbow, and our perspectives on the brain, mind, and consciousness are forever altered.
Neuroscientists will soon decipher colors in that rainbow and then figure out the colors that seem to glow on the black-light posters of our minds.
Shards of the first-order, on/off observation that the left brain analyzes and the right brain creates have evolved to second-order, colored complexity. Now we have a concentrating, focused, but occasionally delusional left brain and a vigilant, wide-eyed, but occasionally depressed right brain. The progression from a conclusive statement to “well, there’s something to it, but you have to be careful” isn’t a flaw; it’s a method.
Science seeks the truth through a succession of improved understandings. By figuring out what and how, we try to unravel why. We may never get the exact answer, the stand-alone, wave-the-flag-from-the-top-of-the-hill truth, but as we peel away ignorance, we get ever closer to the truth until the distinction between theory and truth is akin to the distinction between the realities we reconstruct from our senses and the reality that’s out there, independent of our own experience.
10.1 LEFT-RIGHT, DOWN-UP, FRONT-BACK
We started this book by upgrading the roles of the left and right hemispheres, but the primary functional distinction of brain geometry is not the difference between left and right; it’s the hopping frog, affectionate puppy, and intellectual Feynman. This down-up construction of our brains follows the brainstem à limbic system à neocortex timeline of evolution. Still, the behaviors we inherit from the three are pyramidal: a heartbeat, a feeling, an idea. They build on each other but are so intimately interdependent, feeding backward and forward, amplifying and suppressing, that making distinctions can lead to mistakes. Don’t let the metaphor get in the way of the science.
We can play the same game in the longitudinal direction. Front-to-back thought within the neocortex makes associations. We haven’t spent much time on brain anatomy. If we had mapped out the anatomy, lent names to each region of the neocortex, each nodule of the limbic system, and the primary features of the brainstem, this book would be ten times longer and three times more boring. Plus, you would have had to drill me with flash cards so I could remember the details. So let me just state that there is no oversimplification for longitudinal thought. If pressed, I’d be tempted to say that sophistication increases from back to front, with sensory processing mostly in back, control of your limbs in the middle, and planning and goal-setting in the front. If you ignored an ocean of evidence to the contrary, you might even agree. Let’s not do that. There is no little person at the helm in your head, and every book that colors in a specific region for high-level processes is, at least at this point in the progress of neuroscience, full of shit. It’s true that you plan up front, and your feelings and self-awareness have been linked to several separate regions in the front third of your brain, like your left and right insula just above your ears and behind your temples, but the regions are interconnected with every other part of your neocortex and have no clear-cut boundaries.
Every region that’s related to how you assemble a unified self spends most of its time petting your inner puppy. Since your inner puppy creates most of the feelings on which you act, we could as accurately state that you are your puppy. In other words, you’re a dog. Some folks might not appreciate this oversimplification.
On the other hand, the concept of lateral thought, the source of humor and novelty, the font of genius, is a metaphor that might be grounded in literal, lateral brain geometry.
10.2 OBSESSIVE PATTERN PREDICTORS NEED LOTS OF EDUCATION
Our survival as a species, as well as the survival of most of the other species that share this planet, depends on our ability to solve the problems that come from ten billion people trying to live in harmony on one big, wet rock. Solving these problems—our tendency to fight wars, blow shit up, destroy fertile fields, muck up the atmosphere, you know the list—requires innovation. Not just scientific and technological innovations, but innovations of economy, diplomacy, politics, culture, and so on. The one thing we know for certain about innovation and creativity is that they emerge when we carry concepts from one field into another and then blend them into something completely new. In our heads, this happens through lateral associations of seemingly unrelated concepts that emerge as whole new ideas.
Cutting arts programs in schools is as stupid as cutting science and mathematics programs. We need artists at the top of their games to help us understand and empathize, to feel the repercussions of political decisions. For scientists and engineers to help solve our technological problems, they need answer resolution fortified by literature, history, philosophy, arts, music, and social science. Fully educated people have a huge advantage no matter what challenges confront them.
We tend to innovate in layers of abstraction. Have you noticed how the progress of the World Wide Web follows the simplest model of abstraction layering? The first marketplace emerged about ten thou-sandyears ago so that communities could capitalize on specialization; eBay is nothing but an abstraction of the marketplace, a virtual flea market. Amazon is another layer of the department store. Facebook is a town square. You can predict everything that will come from the Internet by abstracting everything that came before it—I was serious about that Internet karaoke comment a couple of hundred pages back.
The occupants of this planet need the greatest possible answer resolution, the best tools for abstraction, and the widest and wisest perspectives.
10.3 THE NEUROSCIENCE ONION
When forced to answer what single point I hope to have made in these pages, I’d be tempted to say something about innovation, but that would make my marketing bullshit meter go off. The one theme that permeates neuroscience is that certain, seemingly obvious, aspects of the brain are not separable. Trying to separate talent and skill or any of the other dichotomies I’ve used as chapter titles is a loser’s game. The brain is a system of feedback layers, a picture within a picture that can’t be approximated by simple, general statements or “this” versus “that” arguments—it’s damn near always this and that.
10.3.1 Neuroscience has issues
Will special, unique, mirror neurons be discovered, or will the process of mirroring or mentalizing or whatever you want to call it turn out to be a more subtle process? Will high-level processes like judgment and awareness be nailed down to specific isolated regions of the neocortex, or will they be mapped across dozens to thousands of processors? How will “t
hought” be defined, and how many different categories of thought will be distinguished? We never even talked about sleep. It probably has something to do with memory formation. Will dreams turn out to come from our top-down processors trying to make sense of whatever boils up from the night-shift work of the hippocampus? Will consciousness turn out to be a threshold of complexity or a spectrum ranging from probably unconscious trees to primarily conscious dogs to whatever consciousness is experienced by sperm whales to the higher-level consciousness of some woman meditating on a hill in Marin County? And what does that mean for the consciousness of beehives or billions of networked computing devices?
The answers are limited by experimental techniques. The primary tools of neuroscience, EEG, fMRI, and PET scans, along with methods from psychology and the behavioral sciences, will experience an incremental improvement in the next decade. And the answers to many of these questions will be unveiled. For each answer, scientists will formulate dozens of new questions aimed at peeling off the next layer. Curiosity tends to generate its own job security.
Scientific conclusions can also be used for corruption and propaganda.
Science has earned a great deal of respect for its ability to unravel complex systems. However, it has also been responsible for giant prejudicial mistakes, like eugenics—the presumption that only wealthy “successful” people should procreate. In the late nineteenth century, as achievements in science flowed into civilization, Darwin’s theory of evolution was abused by politicians, businesses, institutions, and even scientists. Many of the European Arctic explorers believed that Inuit, Lapps, and Eskimos were a separate, inferior species.
The state of neuroscience now in the early twenty-first century could easily devolve into a similar pile of unjustified tripe if we’re not careful. Gross observations about differences between men and women, different cultures or ethnicities, and even income levels can easily be misconstrued from correlation to cause.
Lie-detection technology could continue on its infamous path. Most people believe that polygraphs can reliably tell whether or not someone is lying. They can’t, which is why polygraph results aren’t admissible evidence in courts. Polygraphs aren’t useless though. If you do a polygraph on someone who believes that they work, that person is far more likely to tell the truth when lying might be in their best interests.
We’re headed down the polygraph path with fMRI imaging. A slick neuro-snake-oil dealer could easily convince people that his super-expensive, superconducting fMRI machine can tell when someone’s lying: the more expensive, the more convincing; the more expensive, the greater the probability of corruption.
What if the dealer convinced you that her machine is 99 percent accurate? That should be good enough to sway a jury judging a case that’s otherwise supported by a pile of circumstantial evidence. Without going into a lecture on statistics (and I’m tempted!), let me paint the picture: Say we have one hundred people who never lie. The 99 percent accurate test means that the odds are 50 percent that it will call someone in that sample of one hundred honest people a liar. Now apply the test to ninety-nine honest people and one liar. The odds are almost 50 percent that the test will turn up two possible liars: a 99 percent chance that it will identify the real liar and nearly a 50 percent chance that it will falsely accuse an honest person—a coin flip.
If that’s good enough for you, I hope you’re not on my jury.
10.3.2 Neuroscience’s dark matter problems
It drives me crazy that we don’t have a functional definition of thought.
Ideas and perceptions are like black holes without a definition of gravity. Someday neuroscience will define thought as an unambiguous function in terms that allow independent observers to determine whether or not a thought has occurred. Right now, it’s possible to make a good guess when several different regions of the brain light up an fMRI or PET scan. These tests give us a hint of what the minimal, observable definition of a thought will look like. EEGs measure the currents that correlate to thoughts and can tell us if a brain is not thinking, but they don’t provide that unambiguous measure: “Ding! She just had a thought. Whoa, there’s another one. She’s a frickin’ genius!”
If a thought is defined as a pattern of neural associations, a minimum number of simultaneously active, coordinated neurons, a cusp of organization, then people can’t be the only thinkers on the planet.
When information theory has sufficiently merged with neuroscience that we can measure the difference between the whole and the sum of its parts—really measure it with integrated information—then we might have an observational definition of consciousness and levels of awareness.
Here’s another fun assault on what we’re trying to do: Glial cells make up half the cells in your brain. They include the stuff that forms myelin and the insulation around axons, but they also include astrocytes, cells that transmit action potentials but lack synapses. How do you think neuroscience would differ if astrocytes were easier to study than neurons, or if the first few discoveries in the field had revolved around astrocytes instead of neurons? Maybe we’d be discussing the “strange neurons that thread throughout the brain, probably serving as no more than a net that holds the entire structure together while astrocytes do all the work.”
10.3.3 Experimental difficulties
Functional magnetic resonance imaging, fMRI, has unleashed a torrent of observations about how brains work. To deny that fMRI plays a huge role in figuring out how we figure stuff out would be ridiculous, but no less ridiculous than failing to point out the technical limitations of fMRI.
Everything we’ve discovered about the mind can be traced back to electrical spikes propagating between neurons. So to figure out how everything works, all we have to do is track every one of these signals, map where they go and when—called a connectome, the neuroscience equivalent of genetics’ genome—and, voila, we’ll get a clue of why we do what we do, maybe even answer the ultimate question as to why some people listen to jazz and drink wine, while others listen to rock ‘n‘ roll and swill beer.
The multicolored graphics obtained through fMRI brain scans sure look like maps of what’s going on in there, don’t they? That’s kind of the problem with them: They look like what we’re after, but they’re not quite it.
The nuclear magnetic resonance imaged by fMRI scans is not directly caused by action potentials flowing from neuron to neuron. Instead of mapping the signals, fMRI measures water movement, which is to say, blood flow. Blood flows to cells that deplete their oxygen supply by burning energy. Regions of increased blood flow correlate to regions of increased electrical activity, but let’s make this perfectly clear: fMRIs do not measure neuron signal transmission or reception; they track variations in blood supply.
Brain scans are capable of distinguishing activity in neighboring regions that are a fraction of an inch (several millimeters) apart, about the width of a pea. The changes in blood flow that they track occur over a couple of seconds. Neuron cell bodies are about 0.030 millimeters (30 microns) across and axons are about 0.001 millimeters (1 micron) in diameter. Neurons receive and process a signal and then transmit their responses in thousandths of a second (milliseconds). Adding it all up, each pixel of an fMRI scan indicates the blood flow required to support the actions of tens of thousands of separate neurons over a time period when those neurons can receive, process, and respond to thousands of signals.
In other words, judging the activity of a brain based on fMRI scans is like determining traffic conditions by sniffing car exhaust once an hour.
That such coarse, imprecise measurements have unveiled such a trove of information indicates just how new this field is and should wave a cautionary flag: Do not conclude anything on the basis of fMRI scans alone.
10.3.4 Skepticism is warranted
Good scientists tread warily among new results. Physicists don’t accept a discovery unless the odds of random processes mimicking that discovery are worse than a million to one (5 sigma). In other word
s, if you report something new, but it’s possible for random processes to show the same signal more than once in every million trials, they will read your paper, stroke their chins, nod and take it in. They might even buy you a beer, but they won’t put out a press release until you get a stronger signal.
Typical neuroscience experiments are performed on too few people to draw statistically sound, decisive results. For example, in a test performed on sixteen people, the probability of mistaking a random fluctuation for a signal is about 25 percent, the statistical uncertainty; double the sample size to thirty-two, and the odds drop to 18 percent; double it again to sixty-four, and the odds drop to 12.5 percent. To get down to an uncertainty of 5 percent, you need a sample of four hundred. The results can be refined by applying rigorous statistical techniques to trends and correlations, but no matter how you add it up, these experiments have huge uncertainties. Now add the experimental bias caused by using university undergraduates—mostly healthy, well-educated people in the age range from eighteen to twenty-five— and the uncertainty goes up again. Most neuroscience results quote statistical significance but make no effort to estimate their entire experimental uncertainty. Estimating systematic bias is difficult and time-consuming, but it is certainly possible. The point is that neuroscience builds over many studies; don’t invest too much value in one or two individual investigations.
Neuroscience excites people; we all want to peer into the picture within a picture and catch a glimpse of the camera that actually takes the picture. We all want to know the secret of how we are who we are. I hope that we also want accurate answers.
People like to be right. Sometimes people like to be considered right even when they’re wrong. How many times have you argued with a blockhead long after his claims have been disproven? How many times have you been the blockhead?
The Left Brain Speaks, the Right Brain Laughs Page 23