Andy Clark, a philosopher of the mind, has entered a useful essay in the NY Times online commentary section: "Do thrifty brains make better minds?"
"Thrifty" in the headline refers to efficiency of information processing. That's a departure from the standard anthropological version of the story, in which "expensive brains" are optimized for energy efficiency. These ideas are not mutually exclusive: a strategy toward bit-saving might well reduce the neural overhead, so to speak. But a brain that follows a strategy of greatest information efficiency might in some respects be more energetically expensive. More important, an evolutionary process that results in a brain with high information efficiency might follow a very different pathway than a process that would give rise to high energy efficiency.
Clark considers the philosophical implications of this "thrifty" model of neural processing, particularly as applied to the relative roles of perception and cognition:
All this, if true, has much more than merely engineering significance. For it suggests that perception may best be seen as what has sometimes been described as a process of controlled hallucination (Ramesh Jain) in which we (or rather, various parts of our brains) try to predict what is out there, using the incoming signal more as a means of tuning and nuancing the predictions rather than as a rich (and bandwidth-costly) encoding of the state of the world. This in turn underlines the surprising extent to which the structure of our expectations (both conscious and non-conscious) may quite literally be determining much of what we see, hear and feel.
Clark does not really touch on the evolutionary constraints that affected brain evolution. He discusses perception and cognition as related engineering problems for which efficient information encoding is the principal constraint. From this point of view, certain well-known perceptual illusions (he uses the "hollow-face illusion" as an example) make great sense.
It may be more useful to rephrase the headline. Thrifty brains may not make better minds, but they do yield a certain kind of mind. There are some things about which it is better not to be fooled. In a world where the brain evolved under natural selection, we should expect some kinds of perception to be more subject to mental abbreviation and shorthand than others. Illusions give us not only insight into how our brains work, but also how they evolved.
Meanwhile, human minds include much information that will not be found in other primates. This includes at least one modality of information (language) not found elsewhere in nature. It seems unlikely that our brains should have been optimized for processing this kind of information in the limited time available. The kinds of tricks visual perception uses to make visual processing more efficient may be analogous to "verbal illusions" in language processing, and maybe there is some evidence there about the pathway taken by language evolution. For a new perceptual modality to come into our population de novo, bootstrapping itself in every growing child, I expect that many steps along that pathway were determined by limitations and constraints.
What we perceive today as elegant, natural selection created as simply as gravity creates a river. The water will flow downhill, every other parameter is free.