Book Review: Limor Shifman - Memes in Digital Culture

The MIT Essential Knowledge Series is the perfect venue for a book on something like this, and in keeping with its subject there's a lot packed into a small space. Memes and viral content are ubiquitous, and even if individually each shared image is trivial, in the aggregate they present a fascinating window into a fundamental aspect of human communication that the internet has vastly amplified. Richard Dawkins was the first person to fix the idea of the meme in the public consciousness, proving his own point by giving the amorphous concept of an informational analogue to the gene a compact, shareable name and definition. However, Shifman mentions two main conceptual difficulties with a neat meme-gene analogy: firstly that people commonly refer to a "meme" as both an individual unit of selection (a gene) and as a unit of transmission (a virus), which are different things; and secondly that people are not helpless vectors for meme reproduction but are actors and meme creators themselves. This is not simple nitpicking, since borrowing terminology from evolutionary biology and epidemiology without care can only lead to confusion and obscure actual insights, and though the book's 2014 vintage makes it practically antediluvian in its examples, its careful analysis of both human and memetic behavior is perfectly undiminished by time.

In one sense, the internet is merely the latest transmission vector for memes, not truly a qualitatively different medium from the books, TV, audio recordings, etc that people used to use to share ideas in the before-time. But even in addition to offering, for the very first time ever, the tantalizing prospect of quantified empirical data on idea transmission, the internet has vastly transformed three key aspects of memes: longevity, fecundity, and copy fidelity. Pity our ancestors: one imagines a poor farming family in the depths of Nebraska in the 1800s, waiting anxiously on the porch of their sod-roofed homestead for a quarterly meme update from the big cities back East tucked neatly next to the Sears Roebuck catalog... no longer! Thanks to what Shifman terms the "hypermimetic logic" of the internet, in practical terms there are more memes than it is possible to humanly experience, with memes mutating and spawning new clades and taxa so rapidly that much of the attention-based economy of the modern internet is based on either exploiting something that's already viral or trying to manufacture something viral out of nothing.

But again, a meme is not a piece of viral content, although the two are often used synonymously (side note: "piece of viral content" is a clunky phrase and "virus" sounds bad; would something like "vireme" work better?). Exact definitions are tricky to make, and nearly impossible to sustain, but the basic idea is that a piece of viral content is a single thing that spreads to many people, while a meme is a "group of content units", each of which could be modified while the whole thing is recognizably descended from a single image, which could then also spread virally. Thus, to use one of her examples, the infamous "Leave Britney alone!" video is a piece of viral content, which then became a meme once people latched onto various aspects (the hair, the eyeliner, the bedsheet, the "Leave X aloooone!" plaint) to make their own parodies/imitations/tributes. Some viral items are rich enough in imagery and hooks to inspire vast families of memetic descendants, "Gangnam Style" being the current best example, but even relatively absurd and meaningless viral content like planking or "Kilroy was here" can be memeified, as the latter was referenced with hilariously nerdy panache as a Kilroy-shaped band-pass filter by Thomas Pynchon in his novel V. It's very rare for something to go viral without also getting memed, but there's still a useful distinction between a dendritic "founder-based meme" such as "Leave Britney alone!" that has a clear original viral basis, and a rhizomatic "egalitarian meme" like LOLcats, where theoretically there must have been a single original Ur-funny cat image but currently consists of an endless sea of comedically equal variations.

Of course, truly proving why one thing goes viral and spawns a million memes while something seemingly identical doesn't is a vast and almost certainly impossible goal. Christian Bauckhage had a 2011 paper titled "Insight Into Internet Memes", lamentably uncited here, which fit the popularity (using Google search results) of various memes into a few distributions, but refrained from entering the "why" debate, as it seems beyond the power of data science alone to explain "why" one meme's popularity fits a log-normal distribution and another's doesn't. Surely it must be some mix of inherent latent virality and the random luck of being in the right place at the right time shared by the right people, if we could ever break down the exact factors. Duncan Watts' superb book Everything Is Obvious (Once You Know the Answer) raised many of these same questions in the context of influence in social networks and how some people get to be so influential while most of the rest of us won't ever be, or how some songs become popular while others don't.

Shifman has a more sociological theory of virality, however, and as she says, successful viral content checks off as many of the 6 Ps as possible: 

• positivity
• provocation of high-intensity emotions
• packaging
• prestige
• positioning in social context
• participation, maybe the most important factor. 

This makes intuitive sense - boring irrelevant crap made by nobodies that no one ever sees is unlikely to go viral almost by definition - yet for all the countless YouTube videos out there that master some combination of relatable protagonists, flawed masculinity, humor, simplicity, repetitiveness, and whimsical content that Shifman identifies as crucial to viral success (see Mark O'Connell's excellent short book Epic Fail for a more poignant meditation on the appeal of watching people fail at something), there are even countless more videos out there that no one besides their creators will ever see. The numberless descendants of "Gangnam Style" might make it the Genghis Khan of memes, yet it too exists in a meme ecosystem where the overwhelming majority of similar K-pop music videos make barely a ripple on the international stage.

And it is on that international stage where the potential existence of a formula for virality and memetic success can be most clearly investigated. The somewhat ugly term "glocalization" refers to the ability of viral content to originate in one culture or context, spread to another, and then acquire memetic aspects specific to its new home. Shifman uses the example of "I upgraded Girlfriend 7.0 to Wife 1.0!" and its gender-reversed counterpart as examples of an incredibly hackneyed joke that is nevertheless extremely popular worldwide because it plays on such familiar stereotypes (although with some significant modifications for some cultures, such as Arabic or Japanese, which don't always have exactly the same cultural signifiers). Same with the American meme Successful Black Man becoming Humanist Ultra Orthodox Man in Israel. Political memes often resist translation, since although some complaints with government are universal, others - the Pepper-Spraying Cop at UC Davis, the "grass mud horse" of Chinese battles against government censors, satires of Benjamin Netanyahu or Nicolas Sarkozy's attempts to take credit or be present for everything - require just a bit too much cultural specificity to go global. Political meme urges are universal, but as Tip O'Neill might have said, all political memes are local.

The final chapter is an admirably humble call for further research on a number of memetic avenues, all of which I second:

• The politics of meme participation. With rare exceptions, the demographics of successful meme creators, influencers, and users looks remarkably similar to the demographics of just about every successful anything: young, well-off, straight white males front and center. If the internet theoretically gives everyone an equal voice, what gives?
• Memes as a language. People use memes to pithily communicate an incredibly broad variety of mental and emotional states to others, yet not all of them translate well. What's the general distribution of memetic communication on the "universal" to "extremely niche micro-sub-culture" spectrum, and why do particular memes fall where they do?
• Memes and political change. Some political memes become popular, and some are then successful, but these paths are neither universal nor uniformly paced - how should we interpret the varying success of memes like "I am the 99%" vs "repeal the death tax"? What about "Carthago delenda est" vs "54-40 or fight"?
• Viral and memetic success. Here's where research like Duncan Watts' or Christian Bauckhage's is important: a better typology of memes can help with the etiology. Once we know the "what"s, then the "how"s and "why"s of memes become much easier to study.

And yes, the book does end with a meme.