Consciousness and the Social Brain

information
B
, the two chunks of information passed to downstream circuitry as
A
+
B
. That, at least, is the essence of the hypothesis.
    Synchrony and oscillations have since been found in the cat, monkey, mouse, rat, and human, and the synchrony does indeed increase in situations when the brain binds many different pieces of information into a perceptual whole. 4
    Recent work on oscillations in the monkey and human cortex suggests that periods of oscillation can open up between two distant regions of the cortex. 5 – 7 As the two cortical areas are active in synchronized oscillations, they become better able to transmit information from one to the other. The period of oscillation acts like a portal for efficient information transfer.
    Shortly after cortical oscillations were first discovered, in 1990, Crick and Koch 8 proposed one of the first mechanistic, neuronal theories of consciousness. They suggested that consciousness was linked to the synchronized neuronal activity that may bind information together across the brain. In their theory, consciousness arises as a result of complex, bound information that can be maintained over long enough intervals to enter short-term memory.
    Many scientists have since reiterated and emphasized the same general hypothesis: consciousness occurs when information is bound into complex units that span many disparate brain regions. 2 , 4 , 6 , 9 – 13
    A recent theory of this type, in which consciousness is a complex, bound set of information, was proposed by Tononi. 14 , 15 In Tononi’s integrated information hypothesis, as information becomes linked in increasingly complex ways, consciousness emerges. He describes the example of looking at a white screen. A simple machine equipped with a photodiode can register the information that the screen is bright. But the machine presumably has no conscious experience of whiteness. Why? In Tononi’s account, it is because the machine encodes impoverished information. A human has a vast linked set of visual information, including the luminance of the screen; the whiteness of the screen; the distinction between white and the many colors that are not present on the screen, between white and black, between white and red, between white and any other location in the vast informational space of the color wheel; information about the textureof the screen, about the borders of the screen, and so on. The sheer immensity of information encoded in the person’s brain about a white screen is starkly different from the simple, low-bit information of the machine with the photodiode. In Tononi’s hypothesis, this informational complexity in the human brain is why the human has a conscious experience and the machine does not. In that hypothesis, the conscious experience of whiteness is tantamount to the vast network of bound information in the human brain that is invoked by looking at the white screen. Consciousness is integrated information. As the amount of integrated information increases, consciousness emerges. As the integrated information is reduced, consciousness should fade.
    Tononi’s hypothesis is stripped of reference to the neuronal underpinning. Whether synchronous neuronal activity truly provides the mechanism is irrelevant to the deeper concept. Let the hardware be what it is, the essence of the theory is that information, bound in a rich enough interconnected network, becomes conscious.
    In the following sections I outline three potential difficulties with the integrated information theory and discuss how these difficulties might be resolved by the attention schema theory. My goal here is not to argue against the integrated information approach. I am not trying to knock it down and set up a new theory. On the contrary, I argue that the approach is valuable, and that some nagging problems with it disappear when it is considered from a new perspective and combined with ideas from a different source.
Integrated Information Is Not Always Conscious

    Most processing systems in the brain have no relationship to consciousness.
    For example, we move fluidly, easily, coordinating hundreds of muscles, forces, joint angles, and joint speeds, without any consciousness of exactly how. The computations that guide theintercoordinated nuance of muscle control are outside of awareness. Only very general aspects of movement control and movement goals ever reach consciousness. The highly integrated computations that control heart rate, blood vessel