Wednesday 4 April 2012

Cultural epidemiology criticised

A recent article in The Economist criticises approaches to cultural evolution based on epidemiology.

The article is called: Social contagion - Conflicting ideas The article boils down to saying that disease organisms are bad, and that people want to avoid them - whereas many ideas are good - so people want to acquire them. On that basis it apparently advocates ditching models based on epidemiology.

The article concludes with:

But it suggests that ideas are sufficiently different from diseases that it might not be wise naively to apply models designed for one to probe the other. High time, then, for social psychologists to stop piggybacking on epidemiologists and work harder on their own models.

My assessment is that we need generalised epidemiology - or infodemiology - that deals with both deleterious and beneficial interactions with cultural symbionts. A generalised epidemiology is needed in both the organic and cultural realms - since both feature parasitism and mutualism.

The article's suggestion - that we should ditch epidemiological models and start again from scratch is misguided. It violates the basic principle of building on what you already have. Switching from parasitism to mutualism in basic models of contagion is just a case of switching a sign around. It is true that adaptations for avoiding symbiotes can look a bit different from adaptations for acquiring them - but that does not entail discarding all our existing models, terminology and framework. We do have some three decades of work in cultural evolution built on epidemiological foundations. Existing models don't make the half-baked predictions that author claims arise from an epidemiological perspective.

Copverage in Physorg puts a similar spin on the paper - saying:

Now however, researchers from Cornell University have shown that users adopting Facebook, tend to do so more predictably when receiving invitations from multiple sources, rather than a lot of requests from members of the same group, which implies that Facebook and its growth, does not actually compare with biological contagion at all.
The paper is this one: Structural diversity in social contagion by Johan Ugander, Lars Backstrom, Cameron Marlow, and Jon Kleinberg.

It has thirty one references, practicallly none of which are to the primary literature on cultural evolution. Three cites support their assertion that "traditional" models predict that ideas necessarily spread according to exposure to them - rather than to the extent that recipients want to be exposed to them.

Of course, the actual literature on cultural evolution has long been aware that contagiousness of a cultural entity depends on the properties of the entity, the properties of its hosts, and the structure of the host population (e.g. see Laland and Odling-Smee, 2000) - and it has long recognised both cultural parasitism and mutualism (e.g. see chapters 4 and 5 of Boyd and Richerson, 2005).

No comments:

Post a Comment