In my previous entry Basic Memetics, I described the theory of the meme. Aside from this basic understanding of the concept of a meme, little else is agreed upon. Memetics is still a young science and much work is needed before it can make any predictions. Indeed, not even the very definition of a meme is agreed upon. Some are obviously too simple and broad to be of any use, such as Gabora’s definition: anything that can be an instant of experience.

Some are more complicated, such as Wilkins’s definition: the least unit of socio-cultural information relative to a selection process that has selection bias exceeding its endogenous tendency to change. If I understand that one correctly, which I probably don’t because it is so convoluted, it does not allow such a thing as memetic drift, which would be akin to genetic drift. Sometimes one version of a gene prevails in frequency over another not because it was “more fit” according to natural selection, but because it happened to start just a little bit ahead of the others and the more copies of something there is, the more copies it can make. Chance plays a large role in genetic evolution and there is some debate over whether natural selection or genetic drift plays the larger role. It may well be that memes experience a large amount of drift themselves. Measuring this phenomenon would be crucial to understanding memetic evolution (cultural change) and we would do good not to discount “drifting” memes as not being memes by defining them as only that which has a selection bias exceeding its endogenous capacity for change.

Incidentally, there is some dispute over whether all socio-cultural information can be broken down into “least units”. Some have suggested that meaning, which is different than information, carries the capacity to “contain itself”. Whether or not this is true, it cannot be disputed that at least some cultural information can be broken into units. These units are commonly called ideas, and we know that some ideas spread more than others. This is nothing other than memetic activity.

Some definitions are ridiculously complicated, such as Boyd’s: a bounded set of modulations on any carrier which when received by an animal has by virtue of its configuration a high probability of being transduced into neuronal group and synapse threshold representations which result in effectively identical replications as outputted modulations on other carriers. Nobody really knows how neurons and synapses hold thoughts, nor is it necessary to know in order to understand basic memetics. Don’t get me wrong; it may be helpful someday, but it shouldn’t be included in the definition.

Also, it is unclear how high a probability must be before it is considered a “high probability” and how similar modulations must be to be “effectively identical”. Until more study is done in the field on how memes work in real life, we won’t be able to answer this. We should be careful how we set our definitions so that we do not include too much or too little. To do that, we must know what is out there.

Other definitions include Dawkins’s: the basic unit of cultural transmission, or imitation. The problem with this definition is that to understand it, one must first know what is meant by cultural transmission, a new concept for many people. For most people, a phrase such as “basic unit of imitation” is meaningless without having a better context.

Plotkin’s definition is: the unit of cultural heredity analogous to the gene and the internal representation of knowledge. However, an analogy should never be part of a definition. Analogies are used for explaining things, with the understanding that all analogies break down someplace. If analogies did not break down under any circumstance, the two objects of discussion in question would then be identical; they would be the same object. Under different contexts, it may make more sense to make the meme analogous to something other than the gene (like a virus – see above). By including the gene, not just in the explanation, but also in the very definition of the meme itself, runs a high risk of causing confusion in those circumstances where the analogy breaks down. Also, few people would understand what is meant by cultural heredity.

Dennett’s definition is: a complex idea that forms itself into a distinct memorable unit and is spread by vehicles, which are physical manifestations of the meme. This definition is good, but it adds too much that is superfluous. A meme is an idea, yes, but one that spreads to other people. The acts of forming into a “distinct memorable unit” and manifesting “vehicles” for its spread are both prerequisites for spreading. It would be simpler to state that a meme spreads, and then let logic dictate what else must occur for that to be possible. Raindrops are defined as water that falls from the sky, not as masses of H₂O that form themselves into distinct units with mass great enough for gravity to overcome Brownian motion, gusts of wind, and electrostatic forces, and that is a good thing.

To all these possible definitions Daniel Noe has offered a counterproposal, his own definition: a meme is an idea that is spread from one person to another. This, I believe, is a definition that can be understood by virtually anybody. It is clear from everything I have ever read on the subject, that memes are thought of as ideas and the only difference between them and all other types of ideas is that they spread from one person to another. People know what ideas are. As for defining the word person, Noe believes it proper to refer to people as decision makers that handle meaning and decide, rather than as computers that merely handle information and react. For certain meme evolution simulations it is assumed that agents must decide to pass on a meme