Wednesday, 25 August 2010

Memetics still misunderstood

One of the more interesting aspects of Sue Blackmore's latest piece on "Temes" in the New York Times blog is the comment section.

Not for the insightful comments - there didn't seem to be too many of those. Rather for the way commenters repeatedly slammed memetics and the whole idea of cultural evolution. For example:

If you want to coin a new term, that's fine; but just don't pretend it's an extension of evolutionary theory / natural selection; the constraints, characteristics, and initial conditions present in evolutionary theory vis-a-vis 'memetics,' are so different as to make a connection between the two inapplicable, not to mention empirically inappropriate.

When will these writers realize that analogy is not argument? One can draw an analogy between human brians and computers, between biological processes like evolution and internet. So what? These analogies neither explain nor predict anything.


Unfortunately the "meme" idea is based on an interesting analogy, not proper quantitative analysis. Perhaps the easiest way to see this is to realise that it does not provide new insights into hitherto mysterious facts, but merely re-describes them in metaphorical terms. For example, the nature of the copying process, and above all of the supposed "mistakes", is not specified.

...and so on.

Sue has a follow-up piece there as well - and the comments are much the same.

Memetics seems central to understanding the position of humans in the natural world - and yet it is one of the most ignored, misunderstood, and maligned areas of science.

Some heads need banging together, I figure.

Susan Blackmore's "temes"


"Temes" is a term coined by Susan Blackmore.

As I am making this video, Susan has recently published her latest piece on "Temes" in the New York Times, following up her previous material on the topic on TED in New Scientist and her own web site.

Her "temes" are named by analogy with memes and genes - and the term "teme" means - roughly - "technological meme".

Here is Susan at TED explaining the term:

The concept of "teme" - neatly illustrates some of the basic principles of cultural evolution.

* Inheritance: "teme" is descended from "meme" - which is in turn descended from "gene".

* Variation: "teme" is not an identical copy of "meme" - but differs from it in both spelling and meaning.

Lastly, there is the idea of "differential reproductive success". That raises the issue of what the fitness of the "teme" concept is likely to be.

It is hard to say for sure, but I think it would be somewhat surprising if very many people bought into this terminology.

The need for terminology to describe "temes" apparently arises mainly out of the idea that "memes" are defined as being ideas that are copied via imitation.

That definition apparently results in a range of unclassified cultural replicators - in the form of culture that is copied by machines - tape recorders, optical media, disc drives, solid state media - etc. There the copying does not happen via behavioural imitation, but rather via mechanical equipment.

Susan seems to be the world's foremost proponent of that definition of "meme". It apparently defines huge swathes of modern culture as being a type of replicator without a name. So - Susan feels the need of a new term more acutely than most.

One problem is that so few appear to have bought into Susan's definition of a meme in the first place. Certainly to me, memes are units of culture - and it doesn't seem too critical exactly how they are copied around. A definition of meme based on copying via imitation leaves out lots of animal traditions that are transmitted in other ways - and also leaves out a lot of modern human culture, which are copied by machines, without behavioural imitation.

I prefer an umbrella definition of the term "meme". Dividing culture which is transmitted via behavioural imitation from other types of culture using the definition of meme seems profoundly bad to me - it leaves us with no name for the category of all cultural replicators - and that category seems far more fundamental and significant than the concept of cultural replicators tha work via behavioural imitation.

To me - and as far as I can tell - to practically everyone else concerned - "temes" are just a kind of "meme". Susan said she used to refer to them as such as well.

It is no surprise that Susan is the one who is looking for new terminology in this area. From her point of view there are lots of unchristened replicators around. Also, she can't call the brain a "meme organ" and call computers "meme machines" - since she already used the term "meme machine" to refer to the brain.

A problem with attempting a classification scheme that distinguishes between human and mechanical copying is that a lot of human culture doesn't seem to mind very much exactly how it is copied. An idea can be copied by a human, copied by machines and then return to the human domain again. The idea that whether something is a "teme" or a "meme" depends on what agent happens to be copying it at the time seems rather awkward.

Bringing intelligent machines into the definition doesn't help. Yes, humans and intelligent machines are likely to operate on some different types of content - but there will be plenty of crossover initially - a series of shades of grey between human-readable content and machine-readable content. It seems as though there will be no hard line between "memes" and "temes" - making the area a challenging zone in which to make category divisions.

As far as terminology goes, it is bad enough that we have genes and memes - and genetics and memetics. "Memetics" is the "genetics" of ideas, and most of its concepts are ones that are shared with genetics. Terminology triplication - in the form of "Temetics" - seems like overkill to me.

It is true that there are some differences in dynamics between humans copying each others' ideas and machines doing so. Machines tend to speak to each other in their own computer languages, that are often challenging for humans to comprehend. So: the distinction has some virtues.

Susan Blackmore originally coined the term, and currently is one of the few people who talks about temes.

Everybody likes to invent terminology, and it is nice to contribute to the language. However, there are overheads associated with inventing new terminology. It creates jargon. I am not convinced that the term "temes" is more positive than negative.

Sorry Sue!