What makes a meme


This piece originally appeared in The New Inquiry.

Biden-laughs and Ryan-abs, Big Birds and binders and bayonets: There is something fascinating when an event as stodgily ceremonial as the presidential campaign is run through the lulz-filter of social media, secreting a hallucination of phrases and images and videos and, of course, gifs. An army is at the ready to spin off a gag at every turn, to propagate the joke to maximum scope; digital arpeggiations of candidate goofs and campaign blunders are transmitted from host to host through a mere caress of the touch-sensitive screen. Watching debates with that second screen of fast-moving social media streams and text-input boxes begging our thoughts has positioned many of us as hunters for the most shareable, memeiest content, ready to pounce at something, anything, and in the process, changing the overall narrative of an event. We’ve developed a kind of meme literacy, a habit of intuiting in real time the potential virality of a speech act — to hear retweets inside words.

The New Inquiry

Retweets, reposts, reblogs, repins, and remixes lead to reporting. The Meme Election 2012 isn’t just a matter of what’s found in some sticky gif’d-out corner of Tumblr; it also dominates everyday Facebook feeds and news blogs. And because journalists are disproportionally connected digitally, popular memes also burrow into mainstream-media narratives as a measure of what has captured people’s attention. Whether you watched the conventions and debates on one screen or three, there’s a good chance you encountered discussion of Internet memes afterward.

The definition of meme can be debated, but the short of it is that a meme is a unit of culture, a parallel to the biological gene in Richard Dawkins’s original coinage. Many have since adapted the term to describe how cultural products pass virally from person to person by multiplying themselves throughout the social body. Technically, any shared image is a meme regardless of how viral it has become, but when we say meme, we generally mean a successful one.

Written By: Nathan Jurgenson
continue to source article at salon.com


  1. The definition of meme can be debated, but the short of it is that a meme is a unit of culture, a parallel to the biological gene in Richard Dawkins’s original coinage. Many have since adapted the term to describe how cultural products pass virally from person to person by multiplying themselves throughout the social body. Technically, any shared image is a meme regardless of how viral it has become, but when we say meme, we generally mean a successful one.

    If a meme exists, then it has to be coded in the neural networks of the brain, just as genes are coded in DNA molecules.  The difference between a gene and a meme is that the gene copies itself spontaneously, physically making the copy from raw materials in its immediate environment, whereas a meme by necessity must replicate via another medium such as speech, artistic products, writing, and behavioural repertoires such as dancing and posture.  Memetics also entails a selection process, whereby memes produce phenotypes that effect the rate of replication.  Meme phenotypes begin with brain activity triggered by it and continue through motor outputs to external products like cultural artefacts.

    The main difference between them is that genes act like recipes, influencing the phenotypes in subtle and complex ways, and you can’t map any one phenotype onto any one genotype.  In memetics, however, the process requires a blueprint-like structure in which any one product of the meme (say, a dance move or a totem) must be reverse-engineered by another brain and reconstructed.  In other words, genes are recipes for building proteins and protein products, whereas memes are blueprints for constructing behaviours.  An interesting side effect is that genes replicate and produce phenotypes independently, whereas memes produce phenotypes as a means of replication.  The upshot is that the memetic theory must rely on pre-existing interpretation and reverse-engineering mechanisms in the brain to sort out noise from the signal, coupled with an understanding of what the brain of the other person was trying to achieve.

    It also depends on the discretion of the rest of the system, whether or not it presses the right buttons to get the body to pass on the meme to someone else.  This isn’t necessarily to their benefit: the creationism meme is alive and well on this site despite being generally reviled as bad science, in fact BECAUSE it is so reviled.  Memes don’t have to be good or bad, only spreadable.

  2. One of the problems with the meme idea is one of competition.  Genes compete against alleles for resources available to them if they monopolize a spot on the genome; in other words, genes compete for a genetic locus.  It’s possible that memes compete for specific spots on memetic loci in the brain, as distinctive neural nets, but no one has yet managed to show that this is the case, and the compartmentalization of beliefs raises questions as to what exactly is competing with what.  In any case, there’s no evidence for this claim yet to vindicate the hypothesis.

    Another is the mutation rate.  Reverse-engineering is notorious because it’s often impossible to know what a device was for, and in any case there’s the age-old problem of getting lost in translation.  One way to “solve” this problem is to cheat and have the answers already built-in.  The cost of this is that the information must have come from innate concepts encoded in the genes, which lessens the autonomy of memes.  The rate of mutation of a meme per generation (from brain to brain), in any case, doesn’t seem to have been tested.

    The third problem is simply the law of parsimony: memetics requires complexities that the data currently doesn’t need.  Mainstream psychology and neuroscience has not been advanced by it, and a lot of cultural practices are better recognised as products of psychology than as self-replicating ideas.  The change of ideas and of cultural products can be explained by brains interacting with other brains in complex ways by following open-ended algorithms installed by genetic evolution, coupled with the oddities and quirks of chaos theory.  Brains follow instructions which manifest as behaviours which effect other people’s brains, which follow instructions which manifest as behaviours etc.

    The fourth problem is giving memes too much causal credit.  Ideas do not replicate themselves.  Instead, they need accompanying machinery to do it for them, and that’s put there by genes that don’t in any case pass on whatever an individual learns during its lifetime.  An analogy is with paper and photocopiers.  If paper copied itself, it would be very different to the photocopier copying it.  In the first, the paper is the object of interest.  In the second, the photocopier.  Memetics strikes me as being unhelpful because it places interest in the ideas, which don’t do anything, at the expense of the mechanisms in the brain that turn a dormant idea into a behaviour, create a product, and then reverse-engineer the information obtained from the product and from a mental cheat sheet into a copy of the idea itself.  Any discretion and “selection” of ideas must happen here because the mechanisms are the main causal agents.  They do require ideas to have certain properties, hence the changing personal convictions and the larger trends of cultural climates, but those properties are relevant only to the discretions of the mechanism that already exists, so we end up being more interested in the mechanism in any case.

    In short, memetics is a valid hypothesis, but it’ll be a long time yet before it becomes established scientific theory, if that ever happens.  In the meantime, it might spare us some confusion if we refrain from treating it as though it was already proven.

  3. Anthrapologist Scott Atran has an interesting article called The Trouble with Memes where he argues that the meme isn’t really a useful metaphor for analyzing culture.  The article can be found here: http://sitemaker.umich.edu/sat

    His argument is that cognitive science research data shows that human memory and thinking don’t behave the way the meme model would suggest. He argues that the meme (as an independent entity separate from a specific individual agent) will go the way of the solar system model of the atom, something useful to generate questions but not as an actual model of how the system works.

  4. An excellent article which explores two important aspects of the emerging medium, the Net.

    Political parties are clearly struggling to come to terms with a medium where the citizens are no longer merely passive ‘consumers’. Something the article fails to explore is the way that their subsequent frustration is becoming a feedback loop of anti-Net legislation.

    The article does a good job of exploding the myth of political ‘debate’ – and shows how the Net has become a great lens to focus on the vacuity of a series of prepared mini-statements designed to counter likely opposition approaches to major subjects. Am I the only one who believes that modern political candidates only project authenticity when they go off message? For me it is the only time we see the real person and we can gather evidence for their credibility, skill and intellect. It seems to me this is a rhetorical question because, as the article points out, the focus of the connected is right there.

    The article is also interesting in that it is prima facie evidence of the evolution of meme – both the specific meme of meme, and the ‘menome’ (analogous to genome) of English language culture.

    I am less convinced on the Author’s claim to have discovered that the analogy of memes is an accurate description of what is going on – particularly in social media. For that to be true, we would have replication (tick box), a meme-changed and circumstance-changed environment (tick box), meme mutation and/or recombination (hmm, definitely a hole here), competition of memes (true up to a point – but the analogy breaks down when the memes don’t interact to the extent that they don’t change each other, which is particularly true of partisan politics) varied offspring (most political memes seem to lead short, sharp, monotonic lives inconsistent with the biological metaphor) and varied levels of survival (the real meme survival-to-reproduction rate seems particularly hard to define).

    There would also appear to be a particular problem with missing links. Although the digital records of old memes, and even memeplexes, are lying around like piles of fossils – there are many claims of links from past to present (or long past to recent past) that remain to be demonstrated.

    Does the meme model also break the second law of thermo-dynamics?


  5. Like it or not, memes have gone mainstream.  Viral, even.  There are those among us who want to nail down the correlations or otherwise between the concept of “meme” and the concept of “gene”, but in any case the memes – whatever their precise definition – are out there.

    As for 2nd law of thermodynamics, how does that apply to information?  Or was that just a throwaway rhetorical?

  6. The concepts of “meme” and “gene”seem to have similarities because they are probably two instances of the same underlying fundamental principle of the universe: Self organization.  Relatively simple parts following simple rules can organize themselves into complex systems that make new things happen. They can do this because everything is in the same pot, so to speak, and causes are not isolated from effects.  Effects can influence causes through feedback loops until a dynamic equilibrium is reached, resulting in something new.  The structure of the complex system determines what will happen and thus defines the meaning of the information in the system.

Leave a Reply