Thursday 24 February 2011

Where Boyd and Richerson differ from memetics

One common criticism of memetics is that it adds nothing new to other approaches. When quizzed about which other approaches are intended, the work of Boyd and Richerson often crops up.

Richard Dawkins published on memes in 1976 - the same year that Boyd and Richerson first publlished on the topic. It seems a bit unfair to claim that memetics is duplicating their work.

Critics who cite Boyd and Richerson may not realise how comprehensively their work validates the main points of memetics.

I generally try to treat Boyd and Richerson as players on the same team as the meme enthusiasts. They at least recognise that culture evolves, coevolves with DNA, and can sometimes be harmful - the basics of memetics. They even explicitly endorse the "meme's eye view" (Not By Genes Alone, pages 153-154). Apparently they even drafted their "Not By Genes Alone" using "meme" terminology through-out - and then substituted in "cultural variant" before publication.

However, they occasionally make misguided critiques of memetics - and are sometimes cited by its critics (e.g. see here and here) - so I thought I would take a moment to explain one of the places where I think these researchers go wrong. Apart from rejecting the term "meme", and advocating use of their own "cultural variant" terminology, that is.


Please note: most of the rest of this article is now out of date. See the note at the bottom for details.


In 2001, Boyd and Richerson authored a paper entitled: "Culture is Part of Human Biology Why the Superorganic Concept Serves the Human Sciences Badly" where they lay out their philosophy:

Culture is a part of human biology, as much a part as bipedal locomotion or thick enamel on our molars.

Boyd and Richerson's identification of human culture as part of human biology seems to be a significant mistake to me.

In memetics, memes form their own cultural organisms, which do not interbreed with humans. Their relationships with humans are thus more like like those of symbiotic gut bacteria, pathogenic agents, or domesticated animals.

If you look at what is inherited, culture is not closely related to humans. Nor do cultural inheritance closely track the inheritance patterns of the human genome. Cultural entities are often mobile independently of humans, they can die largely independently of their human hosts and they engage in sexual recombination largely independently of humans. Their genotypes are often found outside the human body. Their phenotypes are often found outside the human body. I think these factors mean that modelling cultural entities as separate organisms makes sense - whereas modelling them as some kind of extension of humans does not.

Memetics allows classification of cultural entities into different sexual species - with limited meme flow between them. For example, Cobol programs tend to have sex with other Cobol programs while BASIC programs tend to have sex with other BASIC programs. However, if Cobol and BASIC are considered to be "part of human biology", it makes little sense to classify them into separate species.

Also, cultural adaptations benefit meme reproduction - while being deleterious to the human hosts involved. That makes a lot of sense under the interpretation that cultural entites represent separate critters that behave much like mutualists and pathogens - and very little sense under Boyd and Richerson's interpretation.

Similarly, memes compete with similar types of memes for some niches and not other ones - which makes sense in the context of an ecology containing different species occupying different niches - and not much sense otherwise.

Later in the same paper, they write:

The term coevolution classically derives from the interacting evolution of pairs of species like predators and prey, diseases and hosts, and mutualists. In the present case we imagine that our culture is something like a symbiont. It lives in the same body as our genes, but has a different life cycle and thus responds somewhat differently to evolutionary forces.

This is much better. Memes only spend some of their time in human bodies, and have what is usually a totally different life cycle, though. Not human: other. Also, there are many visitors with different reproductive cycles. Culture is not really a symbiont, though - it is a more like a collection of symbionts - mutualists and pathogens.

I think it is important to understand that we are dealing with a range of quite different "species" - some of which may not have our best interests at heart. Once memes have constructed computers, robots and nanotechnology, and are no longer so dependent on us for their reproduction, viewing them as "part of human biology" will become more obviously a mistake.

To make clear that Boyd and Richerson have not changed their mind about this in the mean time, here is a 2010 article, where Richerson again promotes the case that we should: "think of culture as part of human biology". This would be misleading IMO. Humans are also in a mutually beneficial symbiotic relationship with lettuces - but it would be confusing to claim that lettuces are part of human biology.

What would be OK is the idea that the capacity for supporting culture is part of human biology. If they were saying that, there would be no debate. However, that isn't what they actually wrote.


Update: I found some clarification about this issue in "The Origin and Evolution of Cultures", page. 4. They say:
Culture is part of human biology. The capacities that allow us to acquire culture are evolved components of human psychology, and the contents of cultures are deeply intertwined with many aspects of our biology.
The headline still seems misleading to me - but the subsequent clarification seems fine - I have no objection to that. This places Boyd and Richerson even closer to memetics than I had previously thought.

See also:

Most of the other remaining differences now appear to be differences in emphasis.

No comments:

Post a Comment