People anticipate the future. We do this not only consciously but subconsciously. I was first convinced of unconscious anticipation reading Hawkin’s book. McCrone’s more recent book provides more detail.

Hawkins’ subjective anecdote that convinced me is the familiarity of the adrenaline rush when, while walking in the dark, you foot falls where the floor is an inch lower than anticipated.

Knowing what will happen, or even what is likely to happen is so useful in ordinary life that it is difficult to explain its utility clearly enough to think about how earlier unconscious anticipation might have evolved as an adaptive function.

Unicellular organisms maintain a few bits of information derived from external signals and used to condition its behavior. Organisms with a complex nervous system maintain a more complex model of their immediate environment. A complex model is likely to attach what, where and when tags on elements of its supposed environment. Rather than posit an unordered set of such elements I will posit an array like construction where elements with similar tag values are adjacent. There are in turn several possible adjacency strategies for nature to use but I will suppose a 4D array which is not one of them. (This is in part to make it clear that I am not proposing physical brain models and also to have familiar mathematical terminology.)

The four dimensions are, of course three space dimensions and time. The senses provide most of the information that goes into our model but people can modify the model with conscious reflection. The senses produce only a meagre form of model which must be augmented by access to learned associations. We do not need to see a lion to add a lion to our model, sound or smell may suffice. Senses can inform our model even less for future times. For this conditioned reflexes will go far.

But why anticipation? — When was it first adaptive? When computer engineers try to design fast digital circuitry, such as computers, a common trick is to compute y earlier than its need if poosible so that f(x, y) may be produced more quickly after learning the value of x. Quickness is adaptive to both predators of prey. If we develop the ability to evaluate the convenience of some state of our environment then this ability might well be used to gauge possible future environments especially when there is some act now that we might take now to increase the likelihood of some desirable state. The above story assumes several future states to be imagined. I suspect that this was not an early development.

I started this note meaning to get to the following idea so I break and write down the idea without further preamble. Imagine a screen saver like app that produces some vaguely repetitious patterns. The task is to engage with some other measured activity while watching for unusual activity on the screen. The range of possible patterns makes this able to produce almost too much data.

I am especially intrigued by the out ability to program our senses to be sensitive to particular patterns, while our conscious mind goes off and does something different.