(e) The MNS is not a dedicated set of hardwired neural circuits; it may be constructed through associative learning. For instance, every time you move your hand, there is activation of motor-command neurons, with simultaneous activation of visual neurons by the appearance of the moving hand. By Hebb’s rule, such repeated coactivations will eventually result in the visual appearance itself triggering these motor neurons, so that they become mirror neurons.
I have two response to this criticism: First, even if the MNS is set up partially through learning, that wouldn’t diminish its importance. The question of how the system works is logically orthogonal to how it is set up (as already mentioned under point d above). Second, if this criticism were true, why wouldn’t all the motor-command neurons become mirror neurons through associative learning? Why only 20 percent? One way to settle this would be to see if there are touch mirror neurons for the back of your head that you have never seen. Since you don’t often touch the back of your head or see the back of it being touched, you aren’t likely to construct an internal mental model of the back of your head in order to deduce that it’s being touched. So you should have far fewer mirror neurons, if any, on this part of your body.
3. The basic idea of the coevolution between genes and culture isn’t new. Yet my claim that a sophisticated mirror-neuron system—conferring an ability to imitate complex actions—was a turning point in the emergence of civilization might be construed as an overstatement. So let’s see how the events may have played out.
Assume that a large population of early hominins (such as Homo erectus or early H. sapiens) had some degree of genetic variation in innate creative talent. If one rare individual through his or her special intellectual gifts had invented something useful, then without the concomitant emergence of sophisticated imitative ability among peers (which requires adopting the other’s point of view and “reading” that person’s intentions), the invention would have died with the inventor. But as soon as the ability to imitate emerged, such one-of-a-kind innovations (including “accidental” ones) would have spread rapidly through the population, both horizontally through kin and vertically through offspring. Then, if any new “innovative ability” mutation later appeared in another individual, she could instantly capitalize on the preexisting inventions in novel ways, leading to the selection and stabilization of the “innovatability” gene. The process would have spread exponentially, setting up an avalanche of innovations that transforms evolutionary change from Darwinian to Lamarckian, culminating in modern civilized humans. Thus the great leap forward was indeed propelled by genetically selected circuits, but ironically the circuits were specialized for learnability—that is, for liberating us from genes! Indeed, cultural diversity is so vast in modern humans that there is probably a greater difference in mental quality and behavior between a university professor and (say) a Texan cowboy (or president) than between the latter and early H. sapiens. Not only is the human brain phylogenetically unique as a whole, but the “brain” of each different culture is unique (through “nurture”)—much more so than in any other animal.
CHAPTER 5 WHERE IS STEVEN? THE RIDDLE OF AUTISM
1. Another way of testing the mirror-neuron hypothesis would be to see if autistic children do not show unconscious subvocalization when listening to others talking. (Laura Case and I are testing this.)
2. Many studies have confirmed my original observation (made with Lindsay Oberman, Eric Altschuler, and Jaime Pineda) of a dysfunctional mirror-neuron system (MNS) in autism (which we accomplished by using mu-wave suppression and fMRI). There is an fMRI study, however, claiming that in one specific brain region (the ventral premotor area, or Broca’s area), autistic children have normal mirror-neuron-like activity. Even if we accept this observation at face value (despite the inherent limitations of fMRI), my theoretical reasons for postulating such a dysfunction will still stand. More important, such observations highlight the fact that the MNS is composed of many far-flung subsystems in the brain that are interconnected for a common function: action and observation. (As an analogy, consider the lymphatic system of the body, which is distributed throughout the body but is functionally a distinct system.)
It is also possible that this part of the MNS itself is normal but its projections or recipient zones in the brain are abnormal. The net result would be the same kind of dysfunction that I originally suggested. In another analogy, consider the fact that diabetes is fundamentally a disturbance of carbohydrate metabolism; no one disputes that. While it is sometimes caused by damage to the pancreatic islet cells, causing a reduction of insulin and an elevation of blood glucose, it can also be caused by a reduction of insulin receptors on cell surfaces throughout the body. This would produce the same syndrome as diabetes without damage to the islets (for islets in the pancreas, think “mirror neurons in the brain’s premotor area called F5”), but the logic of the original argument is unaffected.
Having said all this, let me emphasize that the evidence for MNS dysfunction in autism is, at this point, compelling but not conclusive.
3. The treatments I have proposed for autism in this chapter were inspired in part by the mirror-neuron hypothesis. But their plausibility does not in itself depend on the hypothesis; they would be interesting to try anyway.
4. To further test the mirror-neuron hypothesis of autism, it would be interesting to monitor the activity of the mylohyoid muscle and vocal cords to determine whether autistic children do not show unconscious subvocalization when listening to others talking (unlike normal children, who do). This might provide an early diagnostic tool.
CHAPTER 6 THE POWER OF BABBLE: THE EVOLUTION OF LANGUAGE
1. This approach was pioneered by Brent Berlin. For cross-cultural studies similar to Berlin’s, see Nuckolls (1999).
2. The gestural theory of language origins is also supported by several other ingenious arguments. See Corballis (2009).
3. Even though Wernicke’s area was discovered more than a century ago, we know very little about how it works. One of our main questions in this chapter has been, What aspects of thought require Wernicke’s language area? In collaboration with Laura Case, Shai Azoulai, and Elizabeth Seckel, I examined two patients (LC and KC) on whom I did several experiments (in addition to the ones described in the chapter); here is a brief description of these and other casual observations that are revealing:
(a) LC was shown two boxes: one with a cookie, one without. A student volunteer entered the room and looked at each box expectantly, hoping to open the one with the cookie. I had previously winked to the patient, gesturing him to “lie.” Without hesitation LC pointed out the empty box to the student. (KC responded to this situation the same way.) This experiment shows you don’t need language for a theory-of-mind task.