black-and-white alternations had hoodwinked the insects' nervous 

 systems into "believing" that they were registering the light-shade 

 variations of their natural habitats, these usually occurring on a scale 

 of several meters. This cunning study thereby established that a bee 

 measures distance by recording the amount of "visual flow" (one 

 form of sensory feedback) for a given amount of motion. This is 

 clearly comparable with the mechanism we discussed earlier in con- 

 nection with E. colt 



Cognition 



We can now return to my guess that science needs a sweepingly 

 different approach in its ambition to solve the mystery of conscious- 

 ness. E. coli and A. mellifera ligustica indicate that we need to turn 

 things around. I am going to suggest that for the human being too, 

 the relevant stimulus is the one associated with muscular movement, 

 while the relevant response is the sensory feedback from the sur- 

 roundings. I am thus proposing that our acts of cognition are always 

 related to muscular movements, though these may be merely covert 

 rather than overt — imagined rather than actually performed. 



Let's take a closer look at these acts of cognition. They appear to 

 be passive. We read the words on this page without seeming to 

 move. But the eyes are actually moving a great deal of the time, as 

 they scan the individual pieces of text. The importance of movement 

 is familiar to the blind person, who is forced to read by moving the 

 fingertips across the pattern of raised dots in a Braille text. A passive 

 variant of such touch-mediated reading, with someone else moving 

 the Braille text across the blind person's stationary finger, proves to 

 be impossible. Switching to another sensory mode, we may assume 

 that listening to someone speaking is a passive activity. Only when 

 asked to repeat that person's words do we realize that we are silently 

 mimicking what has been said. If we have not been paying attention, 

 and thus covertly setting up the muscular movements required for 

 articulation, we will not be able to recount what the other person 

 has been saying.. 



This is such a central issue that it ought to be illustrated further. 

 I am going to propose that we read three sentences aloud, even 



6 



