The mental interferometer

Chapter 3 of Gregory Bateson's Mind and Nature: A Necessary Unity presents several ways that two sources of information can be combined to form a pattern -- and ways that such patterns can be perceived as information.

After running through binocular vision, time-delay contrasts, multiple sensory inputs and the like, there is a section on "beats and moiré phenomena", that contains this interesting passage:

Three principles are illustrated by these moiré phenomena: First, any two patterns may, if appropriately combined, generate a third. Second, any two of these three patterns could serve as a base for a description of the third. Third, the whole problem of defining what is meant by the word pattern can be approached through these phenomena. Do we, in fact, carry around with us ... samples of various sorts of regularity against which we can try the information (news of regular differences) that comes in from outside? Do we, for example, use our habits of what is called "dependency" to test the characteristics of other persons?
Do animals (and even plants) have characteristics such that in a given niche there is a testing of that niche by something like the moiré phenomenon? (Bateson 2002:75).

I wonder if certain social interactions may be characterized by a "natural rhythm" of this sort.

The usual model describes most social interactions as straightforward communications -- with one individual "polling" another and waiting for responses. An example would be Dunbar's model of grooming, in which individuals maintain their social relationships by grooming, which both transmits a beneficial message (i.e., I'm scratching your back), and gives an individual time to assess the attitude of the grooming recipient (i.e., will you be scratching mine anytime soon?).

The appeal of the moiré pattern as a mechanism is that an interferometer is both simple and sensitive. If a slight disjoint between social "rhythms" could be easily detected, then it might enable the acquisition of social information through more passive, everyday actions instead of constant interrogation of other individuals. This model would give a more nuanced depiction of social interactions, in which each interaction is a small but important part of an overall pattern -- instead of a few critical interactions bearing most of the importance, and most interactions are just "marking time" between the truly significant ones.

I suppose it's possible that a poll-response model might be such an interferometer-measurable pattern in some higher dimension space. On the other hand, it is quite possible that nothing about social interactions -- even repeated ones -- really corresponds to a simple "rhythm" that can be assessed easily.

But at least some kinds of social interactions would seem likely to lend themselves to this kind of pattern. For example, individuals have to eat and drink at regular intervals, and may want to arrange their eating or drinking either to coincide with or to avoid the times that certain other individuals eat or drink. In such instances, it would be quite obvious -- maybe even jarring -- if the schedules did not work in the predicted pattern.

Another relevant example would be escalating social tensions before a group fission. If bouts of threat displays and aggression followed some pattern of intensification, ultimately the coincidence of such bouts among many individuals might lead to fights, ejection of some individuals from a group, or fission. Each individual would have its own pattern of increasing aggression, manifested by outbursts -- but the outcome for the group depends on the intensification of these outbursts with respect to each other. Hence, a lower average level of aggression might have a greater impact on a larger group -- just because of the chances of multiple individuals having aggressive bouts at the same time would be much higher.


Bateson G. 2002. Mind and Nature: A Necessary Unity. Hampton Press, Cresskill NJ. Amazon