Individual humans are a little like loose hairs on the mad dogs that are institutional conjunctures. Some drop out and are sucked into the vacuum cleaner of posterity. Some hang on and are damned or feted with the rest of their generation. The social constraints that drive these systems can reasonably be described and modelled in causal terms. Although there might seem to be some selective winnowing in effect that might eliminate certain types, each generation seems to re-create the full range of ideational morphs and predispositions. As with those polymorphic snails being preyed on by birds, there may be very real suffering and premature death, but no irreversible evolutionary change.
There is, however, one key respect in which the 'shaggy-dog' metaphor fails. Sometimes institutional constraints are weakened enough to allow an influential few to 'think the unthinkable' and shape opinion. Symmetry-breaking events can occur that activate synergetic multipliers, sweeping old causal constraints aside and allowing new dynamic patterns to evolve. These events cannot be said to cause the new synergetic conjuncture because no-one could have predicted how things would turn out. It only becomes clear with hindsight that they and the mythic histories woven around them played a pivotal role in shaping the emergent dynamic.
Causal constraints exist de facto. Their effect is to make system dynamics locally predictable, subject to some estimative uncertainty. However plesionic complexity also exists. In intervals of space-time where institutional constraints are weakened, time-symmetries can be broken, reorganising system dynamics from the bottom-up. The central thesis of this article has been that anthropologists need evolutionary models that can accommodate and, if possible, explain the dynamic balance between causal constraints and 'anti-causal' emergents in human activity systems.
Darwin's model of human evolution implies that 'tribes' can evolve by descent with modification. Whether he was aware of it or not, the tribe concept as developed in Descent was a hybrid structure - part plesionic system and part institution - a 'virtual agent' demanding fealty and constraining human action. This institutional model contrasts strikingly with 20th-century 'dual inheritance' theories, which suggest that it is not tribes, but behavioural memes or culturally embedded traditions that emerge, hijack human bodies and use them to colonise space and time. Stephen Shennan's variation of the dual inheritance model, for example, suggests that material culture traditions evolve by descent with modification (Shennan 2000).
One of the problems Shennan acknowledges with the dual inheritance model is that of knowing how qualitatively new, founder-cultures come into being. The Darwinian model resolves that problem by allowing that stable institutions constrain human actions and that these constraints are reflected, albeit imperfectly, in human behaviour. Unlike material culture traditions or behavioural memes, which are etic categories imposed on a study population by the scientific observer, institutions are emic structures that exist in the minds of the populations under study. We may not know what prehistoric institutions were called or how they were operationalised, but we have good ethnographic reason to believe they existed.
Those named institutions would have acquired spokespeople and rules of good conduct. They would have competed for affiliates and found ways of attracting or coercing individuals into compliance. They might even have entered into co-operative alliances that created a stable trade-off between the costs and benefits of institutional affiliation. If that trade-off were to evaporate, then individuals would abandon the institution, and its ability to constrain human behaviour would be compromised. Institutional collapse would unlock the adaptive potential of individual humans, allowing new traditions, institutions and behaviour patterns to emerge anti-causally.