What’s in a Meme

Today, for anyone working in online (and increasingly offline) media, the term “meme” is well known, and everyone is desperate to get theirs to go “viral”. Behind these two terms lie quite a lot of interesting thinking, which is far less well known.


The term Meme was coined by Richard Dawkins in his book The Selfish Gene, in 1976 and was expanded on in The Extended Phenotype. In short, he hypothesized that human evolution has moved beyond genes, and we are also evolve by passing our ideas and culture from person to person, and from generation to generation.  Thus he coined the term meme- a mental gene – as  the replication unit by which these were passed on.

He defined a meme as the smallest unit by which an idea, behavior, or style spreads from person to person within a culture

He also hypothhesized that, as with genes, memes travelled in larger groups of connected memes – meme complexes, or complex memes, susually called memeplexes – and these memeplexes also altered over time and were influenced by evolutionary pressures


Assuming culture is transmissable from mind to mind, and evolves in an evolutionary manner. Life-forms can transmit information both vertically (from parent to child, via replication of genes) and horizontally (through viruses and other means). Memes can replicate vertically or horizontally within a single biological generation. Memes, analogously to genes, vary in their aptitude to replicate; successful memes remain and spread, whereas unfit ones stall and are forgotten. Thus memes that prove more effective at replicating and surviving are selected in the meme pool

These conditions drive a number of properties that Memes need to adopt to survive

  • To succeed, Memes first need retention. The longer a meme stays in its hosts, the higher its chances of propagation are. When a host uses a meme, the meme’s life is extended. The re-use of the Humans’ neural to host a different meme is the greatest threat to a tenant meme.
  • A meme which increases the longevity of its hosts will generally survive longer. On the contrary, a meme which shortens the longevity of its hosts will tend to disappear faster.
  • Memes also need transmission. As hosts are mortal, retention is not sufficient to perpetuate a meme in the long term.
  • Memes reproduce by copying from a nervous system to another one, either by communication or imitation. Imitation often involves the copying of an observed behaviour of another individual. Communication may be direct or indirect, where memes transmit from one individual to another through a copy recorded in an inanimate source, such as a book or a musical score.
  • But, Memes can also lie dormant for long periods of time and then be taken up again. Physical methods of storage massively increase this possibility.
  • Memeplexes also play a part in transmitting Memes.  Memeplexes comprise groups of memes that replicate together and coadapt (such as  cultural or political doctrines). Memes that fit within a successful memeplex may gain acceptance by “piggybacking” on the success of the memeplex. New Memes can also find transmission by joining the Mremeplex.

Some theorists differentiate between internal or external memes (i-memes or e-memes) depend on whether they are transmitted directly (human to human) or idirectly (via inanimate object).

Also, unlike genetic evolution, memetic evolution can show both Darwinian and Lamarckian traits. (Lamarckian inheritance is when a host aspires to replicate the given meme through inference rather than by exactly copying it) . In short the Darwinian mode is “copying the instructions” and the Lamarckian mode is “copying the product”

As one would expect, there were arguments about whether such a thing exists (and still are – memes land smack bang in the centre of the nature vs nurture bunfights), whether it is truly evolutionary and so on, but for ther purposes of explaining and calculating -a nd most critically, predicting –  the spread of ideas in populations it worked better than anything else at the time (and still does, in the main), so was adapted by various disciplines as a “good enough for now” approach to try and put some rigour into looking at cultural transfer and trends.

All very interesting, but so what?

Here’s the kicker. Social phenomena such as crazes, fads, hysteria, copycat crime or suicide exemplify can be seen as having a behaviour similar to a contagious disease. Cometh the online world, this was much more observabe in real time, and discretely measurable – especially ins Socialmedia. And contagious diseases have a whole body of theory – and most critically, maths – to predict their behaviours.

So you now have an idea (a meme…)  that says there is a unit that describes cutural transmission, it  can be measured online, and there is a body of maths raedy and waiting to start measuring it by. (Hence, predictably the the term “going viral” sprang up for memes that became truly contagious ) .

Which of course is where we came in…..Evolutionary mathematics,  meet Web science!

Of Maths and Memes

In discussing our start in the post “Finding the Zeitgeist” we noted that:

To anyone familiar with the online world (which we were being asked to search), the Zeitgeist looks a lot like an internet meme.

And it turns out this is just one application of memetic mathematics. As Wikipedia notes, the use of various types of Memetic algorithms have found applications in areas as widely dispersed as:

….training of artificial neural networks, pattern recognition, robotic motion planning, beam orientation, circuit design, electric service restoration, medical expert systems, single machine scheduling, automatic timetabling (notably, the timetable for the NHL), manpower scheduling, nurse rostering optimisation, processor allocation, maintenance scheduling (for example, of an electric distribution network), multidimensional knapsack problem, VLSI design, clustering of gene expression profiles, feature/gene selection, and multi-class, multi-objective feature selection.

We have found that applications in online media listening can be used in optimising systems used by clients for marketing, customer service and experience, innovation and  competitor analysis, in as well as trend measuring and insights from performing similar tasks on the metadata the social graph includes.