Competition among memes in a world with limited attention

Scientific Reports, March 2012, by  L. Weng, A. Flammini, A. Vespignani, & F. Menczer

Scientific Reports 2, Article number: 335, doi:10.1038/srep00335

Received 19 September 2011, Accepted 08 March 2012, Published 29 March 2012

The wide adoption of social media has increased the competition among ideas for our finite attention. We employ a parsimonious agent-based model to study whether such a competition may affect the popularity of different memes, the diversity of information we are exposed to, and the fading of our collective interests for specific topics. Agents share messages on a social network but can only pay attention to a portion of the information they receive. In the emerging dynamics of information diffusion, a few memes go viral while most do not. The predictions of our model are consistent with empirical data from Twitter, a popular microblogging platform. Surprisingly, we can explain the massive heterogeneity in the popularity and persistence of memes as deriving from a combination of the competition for our limited attention and the structure of the social network, without the need to assume different intrinsic values among ideas.

Subject terms:

Statistics

Information theory and computation

Statistical physics, thermodynamics and nonlinear dynamics

Theoretical physics

At a glance

Figures

Introduction

Introduction

Results

Discussion

Methods

References

Acknowledgements

Author information

Comments

Ideas have formidable potential to impact public opinion, culture, policy, and profit1. The advent of social media2 has lowered the cost of information production and broadcasting, boosting the potential reach of each idea or meme3. However, the abundance of information to which we are exposed through online social networks and other socio-technical systems is exceeding our capacity to consume it. Ideas must compete for our scarce individual and collective attention. As a result, the dynamic of information is driven more than ever before by the economy of attention, first theorized by Simon4. Yet the processes that drive popularity in our limited-attention world are still largely unexplored5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15.

The availability of data from online social media has recently created unprecedented opportunities to explore human and social phenomena on a global scale16, 17. In this context one of the most challenging problems is the study of the competition dynamics of ideas, information, knowledge, and rumors. Understanding this problem is crucial in a broad range of settings, from viral marketing to scientific discovery acceleration. Aspects of competition for limited attention have been studied through news, movies, and topics posted on blogs and social media10, 11, 13. The popularity of news decreases with the number of competing items that are simultaneously available8, 18, 19.

However, even in the simplified settings of social media platforms, it is hard to disentangle the effects of limited attention from many concurrent factors, such as the structure of the underlying social network7, 13, the activity of users and the size of their potential audience19, the different degrees of influence of information spreaders20, the intrinsic quality of the information they spread21, the persistence of topics22, 23, and homophily24. To compound these difficulties, social networks that host information diffusion processes are not closed systems; exogenous factors like exposure to traditional media and their reports of world events play important roles in the popularity and lifetime of specific topics10, 25. Another example of our limited attention is the cognitive limit on the number of stable social relationships that we can sustain, as postulated by Dunbar26 and recently supported by analysis of Twitter data27.

We propose an agent-based model to study the role of the limited attention of individual users in the diffusion process, and in particular whether competition for our finite attention may affect meme popularity, diversity, and lifetime. Although competition among ideas has been implicitly assumed as a factor behind, e.g., the decay in interest toward news and movies28, 8, 10, to the best of our knowledge nobody has attempted to explicitly model the mechanisms of competition and how they shape the spread of information. In particular, we show that a simple model of competition on a social network, without any further assumptions about meme merit, user interests, or explicit exogenous factors, can account for the massive heterogeneity in meme popularity and persistence.

Results

Introduction

Results

Discussion

Methods

References

Acknowledgements

Author information

Comments

Here we outline a number of empirical findings that motivate both our question and the main assumptions behind our model. We then describe the proposed agent-based toy model of meme diffusion and compare its predictions with the empirical data. Finally we show that the social network structure and our finite attention are both key ingredients of the diffusion model, as their removal leads to results inconsistent with the empirical data.

We validate our model with data from Twitter, a micro-blogging platform that allows many millions of people to broadcast short messages through social connections. Users can “follow” interesting people, by which a directed social network is formed. Posts (“tweets”) appear on the screen of followers. People can forward (“retweet”) selected posts from their screen to their followers. Furthermore, users often mark their posts with topic labels (“hashtags”). Let us use these tags as operational proxies to identify memes. A retweet carries a meme from user to user. As a meme spreads in this way, it forms a cascade or diffusion network such as those illustrated in Fig. 1. We collected a sample of retweets that include one or more hashtags, produced by Twitter users over a specific period of time (see details in Methods section). This provides us with a quantitative framework to study the competition for attention in the wild.

Figure 1: Visualizations of meme diffusion networks for different topics.

Nodes represent Twitter users, and directed edges represent retweeted posts that carry the meme. The brightness of a node indicates the activity (number of retweets) of a user, and the weight of an edge reflects the number of retweets between two users. (a) The #Japan meme shows how news about the March 2011 earthquake propagated. (b) The #GOP tag stands for the US Republican Party and as many political memes, displays a strong polarization between people with opposing views. Memes related to the “Arab Spring” and in particular the 2011 uprisings in (c) #Egypt and (d) #Syria display characteristic hub users and strong connections, respectively.

Full size image (815 KB)

Figures/tables index

Next

Limited attention

We first explore the competition among memes. In particular, we test the hypothesis that the attention of a user is somewhat independent from the overall diversity of information discussed in a given period. Let us quantify the breadth of attention of a user through Shannon entropy S = −Σi f(i) log f(i) where f(i) is the proportion of tweets generated by the user about meme i. Given a user who has posted n messages, her entropy can be as small as 0, if all of her posts are about the same meme; or as large as log n if she has posted a message about each of n different memes. We can measure the diversity of the information available in the system analogously, defining f(i) as the proportion of tweets about meme i across all users. Note that these entropy-based measures are subject to the limits of our operational definition of a meme; finer or coarser definitions would yield different values.

In Fig. 2 we compare the daily values of the system entropy to the corresponding average user entropy. The key observation here is that a user’s breadth of attention remains essentially constant irrespective of system diversity. This is a clear indication that the diversity of memes to which a user can pay attention is bound. With the continuous injection of new memes, this indirectly suggests that memes survive at the expense of others. We explicitly assume this in the information diffusion model presented later.

Figure 2: Plot of daily system entropy (solid red line) and average user breadth of attention (dashed blue line).Days in our observation period are ranked from low to high system entropy, therefore the latter is monotonously increasing.

Full size image (157 KB)

Previous

Figures/tables index

Next

User interests

It has been suggested that topical interests affect user behavior in social media29, 30. This is a potentially important ingredient in a model of meme diffusion, as an interesting meme may have a competitive advantage. Therefore we wish to explore whether user interests, as inferred from past behavior, are predictive of future behavior.

Let us consider every user in our dataset and any retweets they produce. When a user u emits a new retweet, we define her interests Iu as the set of all memes about which she has tweeted up to that moment. We also collect the set M0 of memes associated with the new retweet. The n most recent posts across all users prior to the new retweet are considered as a set of potential candidates that might have been retweeted, but were not. The corresponding sets of memes M1, M2, …, Mn are recorded (n = 10). We compute the similarity sim(M0, Iu), sim(M1, Iu), …, sim(Mn, Iu) between the user interests and the actual and candidate posts, and recover the conditional probability P(retweet(u, M)|sim(M, Iu)) that u retweets a post with memes M given the similarity between the memes and her user interests. We turn to the Maximum Information Path similarity measure31, 32 that considers shared memes but discounts the more common ones:

 

where x is a meme and f(x) the proportion of messages about x.

Fig. 3 shows that users are more likely to retweet memes about which they posted in the past (Pearson correlation coefficient ρ = 0.98). This suggests that memory is an important ingredient for a model of meme competition, and we explicitly take this aspect into account in the model presented below.

Figure 3: Relationship between the probability of retweeting a message and its similarity to the user interests, inferred from prior posting behavior.

 

Full size image (103 KB)

Previous

Figures/tables index

Next

Empirical regularities

In Fig. 4 we observe several regularities in the empirical data. We first consider meme lifetime, defined as the maximum number of consecutive time units in which posts about the meme are observed; meme popularity, defined as the number of users per day who tweet about a meme, measured over a given time period; and user activity, defined as the number of messages per day posted by a user, measured over a time period. These three quantities all display long-tailed distributions (Fig. 4(a,b,c)). The excellent collapse of the curves demonstrates that the distributions are robust even if measured over different time units or observed over different periods of time. We further measure the breadth of user attention, defined earlier through the meme entropy. Although the entropy distribution is peaked, some users have broad attention while others are very focused (Fig. 4(d)). This distribution is also robust with respect to different periods of time.

Figure 4: Empirical regularities in Twitter data.   (click here for graphic illustration)

 

(a) Probability distribution of the lifetime of a meme using hours (red circles), days (blue squares), and weeks (green triangles) as time units. In the plot, units are converted into hours. Since the distributions are well approximated by a power law, we can align the curves by rescaling the y-axis by λ–α, where λ is the ratio of the time units (e.g., λ = 24 for rescaling days into hours) and α ≈ 2.5 is the exponent of the power law (via maximum likelihood estimation33). This demonstrates that the shape of the lifetime distribution is not an artifact of the time unit chosen to define the lifetime. (b) Complementary cumulative probability distribution of the popularity of a meme, measured by the total number of users per day who have used that meme. This and the following measures were performed daily (filled red circles), weekly (filled blue squares), and monthly (filled green triangles). (c) Complementary cumulative probability distribution of user activity, measured by the number of messages per day posted by a user. (d) Probability distribution of breadth of user attention (entropy), based on the memes tweeted by a user. Note that the larger the number of posts produced, the smaller the non-zero entropy values recorded for users who focus on a small set of memes. This explains why the distributions for longer periods of time extend further to the left.

Full size image (278 KB)

Previous

Figures/tables index

Next

All of these empirical findings point to extremely heterogenous behaviors; some memes are extremely successful (popular and persistent), while the great majority die quickly. A small fraction of memes therefore account for the great majority of all posts. Likewise, a small fraction of users account for most of the traffic. These heterogeneities can in principle be attributed to a variety of causes. The broad distributions of meme popularity could result from a diversity in some intrinsic meme value, with “important” memes attracting more attention. Long-lived memes might be sustained exogenously by traditional media and real-world events. User activity and breadth of attention distributions could be a reflection of innate behavioral differences. What is, then, a minimal set of assumptions necessary to interpret this empirical data? One way to tackle this question is to start from a minimalist model of information spreading that assumes none of the above externalities. In particular we will explore to what extent the statistical features of memes and users can be accounted by the limited attention capacity of the users coupled with the heterogeneity of their social connections.

Model description

Our basic model assumes a frozen network of agents. An agent maintains a time-ordered list of posts, each about a specific meme. Multiple posts may be about the same meme. Users pay attention to these memes only. Asynchronously and with uniform probability, each agent can generate a post about a new meme or forward some of the posts from the list, transmitting the corresponding memes to neighboring agents. Neighbors in turn pay attention to a newly received meme by placing it at the top of their lists. To account for the empirical observation that past behavior affects what memes the user will spread in the future, we include a memory mechanism that allows agents to develop endogenous interests and focus. Finally, we model limited attention by allowing posts to survive in an agent’s list or memory only for a finite amount of time. When a post is forgotten, its associated meme become less represented. A meme is forgotten when the last post carrying that meme disappears from the user’s list or memory. Note that list and memory work like first-in-first-out rather than priority queues, as proposed in models of bursty human activity34. In the context of single-agent behavior, our memory mechanism is reminiscent of the classic Yule-Simon model∼\cite{yule-simon43, Cattuto3001200744}.

The retweet model we propose is illustrated in Fig. 5. Agents interact on a directed social network of friends/followers. Each user node is equipped with a screen where received memes are recorded, and a memory with records of posted memes. An edge from a friend to a follower indicates that the friend’s memes can be read on the follower’s screen (#x and #y in Fig. 5(a) appear on the screen in Fig. 5(b)). At each step, an agent is selected randomly to post memes to neighbors. The agent may post about a new meme with probability pn (#z in Fig. 5(b)). The posted meme immediately appears at the top of the memory. Otherwise, the agent reads posts about existing memes from the screen. Each post may attract the user’s attention with probability pr (the user pays attention to #x, #y in Fig. 5(c)). Then the agent either retweets the post (#x in Fig. 5(c)) with probability 1 − pm, or tweets about a meme chosen from memory (#v triggered by #y in Fig. 5(c)) with probability pm. Any post in memory has equal opportunities to be selected, therefore memes that appear more frequently in memory are more likely to be propagated (the memory has two posts about #v in Fig. 5(d)). To model limited user attention, both screen and memory have a finite capacity, which is the time in which a post remains in an agent’s screen or memory. For all agents, posts are removed after one time unit, which simulates a unit of real time, corresponding to Nu steps where Nu is the number of agents. If people use the system once weekly on average, the time unit corresponds to a week.

Figure 5: Illustration of the meme diffusion model.    (click here for graphic illustration)

Each user has a memory and a screen, both with limited size. (a) Memes are propagated along follower links. (b) The memes received by a user appear on the screen. With probability pn, the user posts a new meme, which is stored in memory. (c) Otherwise, with probability 1 – pn, the user scans the screen. Each meme x in the screen catches the user’s attention with probability pr. Then with probability pm a random meme from memory is triggered, or x is retweeted with probability 1 – pm. (d) All memes posted by the user are also stored in memory.

Full size image (373 KB)

Previous

Figures/tables index

Next

Simulation results

The model has three parameters: pn regulates the amount of novelty that enters the system (number of cascades), pr determines the overall retweet activity (size of cascades), and pm accounts for individual focus (diversity of user interests). We estimated all three directly from the empirical data (see Methods).

The social network underlying the meme diffusion process is a critical component of the model. To obtain a network of manageable size while preserving the structure of the actual social network, we sampled a directed graph with 105 nodes from the Twitter follower network (details in Methods). The nodes correspond to a subset of the users who generated the posts in our empirical data. To evaluate the predictions of our model, we compare them with empirical data that includes only the retweets of the same subset of users. To study the role played by the network structure in the meme diffusion process, we also simulated the model on a random Erdös-Rényi (ER) network with the same number of nodes and edges. As shown in Fig. 6, the model captures the main features of the empirical distributions of meme lifetime and popularity, user activity, and breadth of user attention. The comparison with the corresponding distributions generated using the ER network shows that in general, the heterogeneity of the observed quantities is greatly reduced when memes spread on a random network. This is not unexpected. Consider for example meme popularity (Fig. 6(b)); the real social network has a broad (scale free, not shown) distribution of degree, with a consistent number of hub users who have a large number of followers. Memes spread by these users are likely to achieve greater popularity. This does not happen in the ER network where the degree distribution is narrow (Poissonian). The difference observed in the distribution of breadth of user attention, for both low and high entropy values (Fig. 6(d)), may be explained by the heterogeneity in the number of friends. Users with few friends may have low breadth of attention while those with many friends are exposed to many memes and thus may exhibit greater entropy.

To study the role played by the network structure in the meme diffusion process, we simulate the model on the sampled follower network (solid black line) and a random network (dashed red line). Both networks have 105 nodes and about 3 × 106 edges. (a) The definition of lifetime uses the week as time unit. (b,c,d) Meme popularity, user activity, and user entropy data are based on weekly measures.

Full size image (264 KB)

Previous

Figures/tables index

Next

The second key ingredient of our model is the competition among memes for limited user attention. To evaluate the role of such a competition on the meme diffusion process, we simulated variations of the model with stronger or weaker competition. This was accomplished by tuning the length tw of the time window in which posts are retained in an agent’s screen or memory. A shorter time window (tw < 1) leads to less attention and thus increased competition, while a longer time window (tw > 1) allows for attention to more memes and thus less competition. As we can observe in Fig. 7, stronger competition (tw = 0.1) fails to reproduce the large observed number of long-lived memes (Fig. 7(a)). Weaker competition (tw = 5), on the other hand, cannot generate extremely popular memes (Fig. 7(b)) nor extremely active users (Fig. 7(c)).

Figure 7: Evaluation of model by comparison of simulations with empirical data (same panels and symbols as in Fig. 4).

To study the role of meme competition, we simulate the model on the sampled follower network with different levels of competition; posts are removed from screen and memory after tw time units. We compare the standard model (tw = 1, solid black line) against versions with less competition (tw = 5, dot-dashed magenta line) and more competition (tw = 0.1, dashed red line). (a) The definition of lifetime uses the week as time unit. (b,c,d) Meme popularity, user activity, and user entropy data are based on weekly measures.

Full size image (289 KB)

Previous

Figures/tables index

We also simulated our model without user interests, by setting pm = 0. The most noticeable difference in this case is the lack of highly focused individuals. Users have no memory of their past behavior, and can only pay attention to memes from their friends. As a result, the model fails to account for low entropy individuals (not shown but similar to the random network case in Fig. 6(d)).

Discussion

Introduction

Results

Discussion

Methods

References

Acknowledgements

Author information

Comments

The present findings demonstrate that the combination of social network structure and competition for finite user attention is a sufficient condition for the emergence of broad diversity in meme popularity, lifetime, and user activity. This is a remarkable result: one can account for the often-reported long-tailed distributions of topic popularity and lifetime7, 12, 14, 29 without having to assume exogenous factors such as intrinsic meme appeal, user influence, or external events. The only source of heterogeneity in our model is the social network; users differ in their audience size but not in the quality of their messages.

Our model is inspired by the long tradition that represents information spreading as an epidemic process, where infection is passed along the edges of the underlying social network35, 36, 37, 7, 28, 12.

In the context of social media, several authors explored the temporal evolution of popularity. Wu and Huberman8 studied the decay in news popularity. They showed that temporal patterns of collective attention are well described by a multiplicative process with a single novelty factor. While the decay in popularity is attributed to competition for attention, the underlying mechanism is not modeled explicitly. Crane and Sornette10 introduced a model to describe the exogenous and endogenous bursts of attention toward a video, by combining an epidemic spreading process with a forgetting mechanism. Hogg and Lerman38 proposed a stochastic model to predict the popularity of a news story via the intrinsic interest of the story and the rates at which users find it directly and through friends. These models describe the popularity of a single piece of information, and are therefore unsuitable to capture the competition for our collective attention among multiple simultaneous information epidemics. Although recent epidemiological models have started considering the simultaneous spread of competing strains39, 40, our framework is the first attempt to deal with a virtually unbounded number of new “epidemics” that are continuously injected into the system. A closer analogy to our approach is perhaps provided by neutral models of ecosystems, where individuals (posts) belonging to different species (memes) produce offspring in an environment (our collective attention) that can sustain only a limited number of individuals. At every generation, individuals belonging to new species enter the ecosystem while as many individuals die as needed to maintain the sustainability threshold41.

Since Simon’s seminal paper4, the economy of attention has been an enormously popular notion, yet it has always been assumed implicitly and never put to the test. Our model provides a first attempt to focus explicitly on mechanisms of competition, and to evaluate the quantitative effects of making attention more scarce or abundant.

Our results do not constitute a proof that exogenous features, like intrinsic values of memes, play no role in determining their popularity. However we have shown that at the statistical level it is not necessary to invoke external explanations for the observed global dynamics of memes. This appears as an arresting conclusion that makes information epidemics quite different from the basic modeling and conceptual framework of biological epidemics. While the intrinsic features of viruses and their adaptation to hosts are extremely relevant in determining the winning strains, in the information world the limited time and attention of human behavior are sufficient to generate a complex information landscape and define a wide range of different meme spreading patterns. This calls for a major revision of many concepts commonly used in the modeling and characterization of meme diffusion and opens the path to different frameworks for the analysis of competition among ideas and strategies for the optimization/suppression of their spread.

Methods

Introduction

Results

Discussion

Methods

References

Acknowledgements

Author information

Comments

The data analyzed in this paper was obtained through Twitter’s public APIs. We collected more than 120 millions retweets from October 2010 to January 2011, involving 12.5 million distinct users and 1.3 million hashtags. Each post contains information about who generated and who retweeted it. As expected in a social network, the follower graph has scale-free degree distributions.

Due to the size of the empirical follower network, we sampled a manageable subset for our simulations. The sampling procedure was a random walk with occasional restarts from random locations (teleportation factor 0.15). Though no sampling method is perfect, the modified random walk is efficient in terms of API queries and reproduces the salient topological features of the sampled network42. The sampled network has 105 nodes and about 3×106 edges. The empirical retweets generated by the users in the sample display trends similar to those from the entire dataset, therefore we expect the model predictions to be consistent not only with the sample but also with the full dataset.

The parameter pn characterizes the probability of tweeting about a new meme. To estimate this parameter from the empirical data, we examine whether each hashtag has been observed in previous time units (weeks). The proportion of posts with new hashtags is approximately 0.45 ± 0.05. We thus set pn = 0.45 for all the simulations. For each simulation — standard model, model with underlying random network, and models with strong and weak competition — the parameter pr is tuned to capture the average number of posted memes per user per unit time (Table 1). Finally, the parameter pm represents the proportion of all memes tweeted by an individual that match the content of the memory. To estimate it from the empirical data, we compare each hashtag with those produced by a user in the previous time unit (week). Using the average value across all users (0.4 ± 0.01) we set pm = 0.4.

http://www.nature.com/srep/2012/120329/srep00335/full/srep00335.html

What Defines a Meme?

Our world is a place where information can behave like human genes and ideas can replicate, mutate and evolve
By James Gleick, Smithsonian magazine, May 2011,

What lies at the heart of every living thing is not a fire, not warm breath, not a ‘spark of life.’ It is information, words, instructions,” Richard Dawkins declared in 1986. Already one of the world’s foremost evolutionary biologists, he had caught the spirit of a new age. The cells of an organism are nodes in a richly interwoven communications network, transmitting and receiving, coding and decoding. Evolution itself embodies an ongoing exchange of information between organism and environment. “If you want to understand life,” Dawkins wrote, “don’t think about vibrant, throbbing gels and oozes, think about information technology.”
We have become surrounded by information technology; our furniture includes iPods and plasma displays, and our skills include texting and Googling. But our capacity to understand the role of information has been sorely taxed. “TMI,” we say. Stand back, however, and the past does come back into focus.
The rise of information theory aided and abetted a new view of life. The genetic code—no longer a mere metaphor—was being deciphered. Scientists spoke grandly of the biosphere: an entity composed of all the earth’s life-forms, teeming with information, replicating and evolving. And biologists, having absorbed the methods and vocabulary of communications science, went further to make their own contributions to the understanding of information itself.
Jacques Monod, the Parisian biologist who shared a Nobel Prize in 1965 for working out the role of messenger RNA in the transfer of genetic information, proposed an analogy: just as the biosphere stands above the world of nonliving matter, so an “abstract kingdom” rises above the biosphere. The denizens of this kingdom? Ideas.
“Ideas have retained some of the properties of organisms,” he wrote. “Like them, they tend to perpetuate their structure and to breed; they too can fuse, recombine, segregate their content; indeed they too can evolve, and in this evolution selection must surely play an important role.”
Ideas have “spreading power,” he noted—“infectivity, as it were”—and some more than others. An example of an infectious idea might be a religious ideology that gains sway over a large group of people. The American neurophysiologist Roger Sperry had put forward a similar notion several years earlier, arguing that ideas are “just as real” as the neurons they inhabit. Ideas have power, he said:
Ideas cause ideas and help evolve new ideas. They interact with each other and with other mental forces in the same brain, in neighboring brains, and thanks to global communication, in far distant, foreign brains. And they also interact with the external surroundings to produce in toto a burstwise advance in evolution that is far beyond anything to hit the evolutionary scene yet.
Monod added, “I shall not hazard a theory of the selection of ideas.” There was no need. Others were willing.
Dawkins made his own jump from the evolution of genes to the evolution of ideas. For him the starring role belongs to the replicator, and it scarcely matters whether replicators were made of nucleic acid. His rule is “All life evolves by the differential survival of replicating entities.” Wherever there is life, there must be replicators. Perhaps on other worlds replicators could arise in a silicon-based chemistry—or in no chemistry at all.
What would it mean for a replicator to exist without chemistry? “I think that a new kind of replicator has recently emerged on this very planet,” Dawkins proclaimed near the end of his first book, The Selfish Gene, in 1976. “It is staring us in the face. It is still in its infancy, still drifting clumsily about in its primeval soup, but already it is achieving evolutionary change at a rate that leaves the old gene panting far behind.” That “soup” is human culture; the vector of transmission is language, and the spawning ground is the brain.
For this bodiless replicator itself, Dawkins proposed a name. He called it the meme, and it became his most memorable invention, far more influential than his selfish genes or his later proselytizing against religiosity. “Memes propagate themselves in the meme pool by leaping from brain to brain via a process which, in the broad sense, can be called imitation,” he wrote. They compete with one another for limited resources: brain time or bandwidth. They compete most of all for attention. For example:
Ideas. Whether an idea arises uniquely or reappears many times, it may thrive in the meme pool or it may dwindle and vanish. The belief in God is an example Dawkins offers—an ancient idea, replicating itself not just in words but in music and art. The belief that Earth orbits the Sun is no less a meme, competing with others for survival. (Truth may be a helpful quality for a meme, but it is only one among many.)
Tunes. This tune has spread for centuries across several continents.
Catchphrases. One text snippet, “What hath God wrought?” appeared early and spread rapidly in more than one medium. Another, “Read my lips,” charted a peculiar path through late 20th-century America. “Survival of the fittest” is a meme that, like other memes, mutates wildly (“survival of the fattest”; “survival of the sickest”; “survival of the fakest”; “survival of the twittest”).
Images. In Isaac Newton’s lifetime, no more than a few thousand people had any idea what he looked like, even though he was one of England’s most famous men. Yet now millions of people have quite a clear idea—based on replicas of copies of rather poorly painted portraits. Even more pervasive and indelible are the smile of Mona Lisa, The Scream of Edvard Munch and the silhouettes of various fictional extraterrestrials. These are memes, living a life of their own, independent of any physical reality. “This may not be what George Washington looked like then,” a tour guide was overheard saying of the Gilbert Stuart portrait at the Metropolitan Museum of Art, “but this is what he looks like now.” Exactly.
Memes emerge in brains and travel outward, establishing beachheads on paper and celluloid and silicon and anywhere else information can go. They are not to be thought of as elementary particles but as organisms. The number three is not a meme; nor is the color blue, nor any simple thought, any more than a single nucleotide can be a gene. Memes are complex units, distinct and memorable—units with staying power.
Also, an object is not a meme. The hula hoop is not a meme; it is made of plastic, not of bits. When this species of toy spread worldwide in a mad epidemic in 1958, it was the product, the physical manifestation, of a meme, or memes: the craving for hula hoops; the swaying, swinging, twirling skill set of hula-hooping. The hula hoop itself is a meme vehicle. So, for that matter, is each human hula hooper—a strikingly effective meme vehicle, in the sense neatly explained by the philosopher Daniel Dennett: “A wagon with spoked wheels carries not only grain or freight from place to place; it carries the brilliant idea of a wagon with spoked wheels from mind to mind.” Hula hoopers did that for the hula hoop’s memes—and in 1958 they found a new transmission vector, broadcast television, sending its messages immeasurably faster and farther than any wagon. The moving image of the hula hooper seduced new minds by hundreds, and then by thousands, and then by millions. The meme is not the dancer but the dance.
For most of our biological history memes existed fleetingly; their main mode of transmission was the one called “word of mouth.” Lately, however, they have managed to adhere in solid substance: clay tablets, cave walls, paper sheets. They achieve longevity through our pens and printing presses, magnetic tapes and optical disks. They spread via broadcast towers and digital networks. Memes may be stories, recipes, skills, legends or fashions. We copy them, one person at a time. Alternatively, in Dawkins’ meme-centered perspective, they copy themselves.
“I believe that, given the right conditions, replicators automatically band together to create systems, or machines, that carry them around and work to favor their continued replication,” he wrote. This was not to suggest that memes are conscious actors; only that they are entities with interests that can be furthered by natural selection. Their interests are not our interests. “A meme,” Dennett says, “is an information-packet with attitude.” When we speak of fighting for a principle or dying for an idea, we may be more literal than we know.
Tinker, tailor, soldier, sailor….Rhyme and rhythm help people remember bits of text. Or: rhyme and rhythm help bits of text get remembered. Rhyme and rhythm are qualities that aid a meme’s survival, just as strength and speed aid an animal’s. Patterned language has an evolutionary advantage. Rhyme, rhythm and reason—for reason, too, is a form of pattern. I was promised on a time to have reason for my rhyme; from that time unto this season, I received nor rhyme nor reason.
Like genes, memes have effects on the wide world beyond themselves. In some cases (the meme for making fire; for wearing clothes; for the resurrection of Jesus) the effects can be powerful indeed. As they broadcast their influence on the world, memes thus influence the conditions affecting their own chances of survival. The meme or memes comprising Morse code had strong positive feedback effects. Some memes have evident benefits for their human hosts (“Look before you leap,” knowledge of CPR, belief in hand washing before cooking), but memetic success and genetic success are not the same. Memes can replicate with impressive virulence while leaving swaths of collateral damage—patent medicines and psychic surgery, astrology and satanism, racist myths, superstitions and (a special case) computer viruses. In a way, these are the most interesting—the memes that thrive to their hosts’ detriment, such as the idea that suicide bombers will find their reward in heaven.
Memes could travel wordlessly even before language was born. Plain mimicry is enough to replicate knowledge—how to chip an arrowhead or start a fire. Among animals, chimpanzees and gorillas are known to acquire behaviors by imitation. Some species of songbirds learn their songs, or at least song variants, after hearing them from neighboring birds (or, more recently, from ornithologists with audio players). Birds develop song repertoires and song dialects—in short, they exhibit a birdsong culture that predates human culture by eons. These special cases notwithstanding, for most of human history memes and language have gone hand in glove. (Clichés are memes.) Language serves as culture’s first catalyst. It supersedes mere imitation, spreading knowledge by abstraction and encoding.
Perhaps the analogy with disease was inevitable. Before anyone understood anything of epidemiology, its language was applied to species of information. An emotion can be infectious, a tune catchy, a habit contagious. “From look to look, contagious through the crowd / The panic runs,” wrote the poet James Thomson in 1730. Lust, likewise, according to Milton: “Eve, whose eye darted contagious fire.” But only in the new millennium, in the time of global electronic transmission, has the identification become second nature. Ours is the age of virality: viral education, viral marketing, viral e-mail and video and networking. Researchers studying the Internet itself as a medium—crowdsourcing, collective attention, social networking and resource allocation—employ not only the language but also the mathematical principles of epidemiology.
One of the first to use the terms “viral text” and “viral sentences” seems to have been a reader of Dawkins named Stephen Walton of New York City, corresponding in 1981 with the cognitive scientist Douglas Hofstadter. Thinking logically—perhaps in the mode of a computer—Walton proposed simple self-replicating sentences along the lines of “Say me!” “Copy me!” and “If you copy me, I’ll grant you three wishes!” Hofstadter, then a columnist for Scientific American, found the term “viral text” itself to be even catchier.
Well, now, Walton’s own viral text, as you can see here before your eyes, has managed to commandeer the facilities of a very powerful host—an entire magazine and printing press and distribution service. It has leapt aboard and is now—even as you read this viral sentence—propagating itself madly throughout the ideosphere!
Hofstadter gaily declared himself infected by the meme meme.
One source of resistance—or at least unease—was the shoving of us humans toward the wings. It was bad enough to say that a person is merely a gene’s way of making more genes. Now humans are to be considered as vehicles for the propagation of memes, too. No one likes to be called a puppet. Dennett summed up the problem this way: “I don’t know about you, but I am not initially attracted by the idea of my brain as a sort of dung heap in which the larvae of other people’s ideas renew themselves, before sending out copies of themselves in an informational diaspora…. Who’s in charge, according to this vision—we or our memes?”
He answered his own question by reminding us that, like it or not, we are seldom “in charge” of our own minds. He might have quoted Freud; instead he quoted Mozart (or so he thought): “In the night when I cannot sleep, thoughts crowd into my mind…. Whence and how do they come? I do not know and I have nothing to do with it.”
Later Dennett was informed that this well-known quotation was not Mozart’s after all. It had taken on a life of its own; it was a fairly successful meme.
For anyone taken with the idea of memes, the landscape was changing faster than Dawkins had imagined possible in 1976, when he wrote, “The computers in which memes live are human brains.” By 1989, the time of the second edition of The Selfish Gene, having become an adept programmer himself, he had to amend that: “It was obviously predictable that manufactured electronic computers, too, would eventually play host to self-replicating patterns of information.” Information was passing from one computer to another “when their owners pass floppy discs around,” and he could see another phenomenon on the near horizon: computers connected in networks. “Many of them,” he wrote, “are literally wired up together in electronic mail exchange…. It is a perfect milieu for self-replicating programs to flourish.” Indeed, the Internet was in its birth throes. Not only did it provide memes with a nutrient-rich culture medium, it also gave wings to the idea of memes. Meme itself quickly became an Internet buzzword. Awareness of memes fostered their spread.
A notorious example of a meme that could not have emerged in pre-Internet culture was the phrase “jumped the shark.” Loopy self-reference characterized every phase of its existence. To jump the shark means to pass a peak of quality or popularity and begin an irreversible decline. The phrase was thought to have been used first in 1985 by a college student named Sean J. Connolly, in reference to an episode of the television series “Happy Days” in which the character Fonzie (Henry Winkler), on water skies, jumps over a shark. The origin of the phrase requires a certain amount of explanation without which it could not have been initially understood. Perhaps for that reason, there is no recorded usage until 1997, when Connolly’s roommate, Jon Hein, registered the domain name jumptheshark.com and created a web site devoted to its promotion. The web site soon featured a list of frequently asked questions:
Q. Did “jump the shark” originate from this web site, or did you create the site to capitalize on the phrase?
A. This site went up December 24, 1997, and gave birth to the phrase “jump the shark.” As the site continues to grow in popularity, the term has become more commonplace. The site is the chicken, the egg and now a Catch-22.
It spread to more traditional media in the next year; Maureen Dowd devoted a column to explaining it in the New York Times in 2001; in 2002 the same newspaper’s “On Language” columnist, William Safire, called it “the popular culture’s phrase of the year”; soon after that, people were using the phrase in speech and in print without self-consciousness—no quotation marks or explanation—and eventually, inevitably, various cultural observers asked, “Has ‘jump the shark’ jumped the shark?” Like any good meme, it spawned mutations. The “jumping the shark” entry in Wikipedia advised in 2009, “See also: jumping the couch; nuking the fridge.”
Is this science? In his 1983 column, Hofstadter proposed the obvious memetic label for such a discipline: memetics. The study of memes has attracted researchers from fields as far apart as computer science and microbiology. In bioinformatics, chain letters are an object of study. They are memes; they have evolutionary histories. The very purpose of a chain letter is replication; whatever else a chain letter may say, it embodies one message: Copy me. One student of chain-letter evolution, Daniel W. VanArsdale, listed many variants, in chain letters and even earlier texts: “Make seven copies of it exactly as it is written” (1902); “Copy this in full and send to nine friends” (1923); “And if any man shall take away from the words of the book of this prophecy, God shall take away his part out of the book of life” (Revelation 22:19). Chain letters flourished with the help of a new 19th-century technology: “carbonic paper,” sandwiched between sheets of writing paper in stacks. Then carbon paper made a symbiotic partnership with another technology, the typewriter. Viral outbreaks of chain letters occurred all through the early 20th century. Two subsequent technologies, when their use became widespread, provided orders-of-magnitude boosts in chain-letter fecundity: photocopying (c. 1950) and e-mail (c. 1995).
Inspired by a chance conversation on a hike in the Hong Kong mountains, information scientists Charles H. Bennett from IBM in New York and Ming Li and Bin Ma from Ontario, Canada, began an analysis of a set of chain letters collected during the photocopier era. They had 33, all variants of a single letter, with mutations in the form of misspellings, omissions and transposed words and phrases. “These letters have passed from host to host, mutating and evolving,” they reported in 2003.
Like a gene, their average length is about 2,000 characters. Like a potent virus, the letter threatens to kill you and induces you to pass it on to your “friends and associates”—some variation of this letter has probably reached millions of people. Like an inheritable trait, it promises benefits for you and the people you pass it on to. Like genomes, chain letters undergo natural selection and sometimes parts even get transferred between coexisting “species.”
Reaching beyond these appealing metaphors, the three researchers set out to use the letters as a “test bed” for algorithms used in evolutionary biology. The algorithms were designed to take the genomes of various modern creatures and work backward, by inference and deduction, to reconstruct their phylogeny—their evolutionary trees. If these mathematical methods worked with genes, the scientists suggested, they should work with chain letters, too. In both cases the researchers were able to verify mutation rates and relatedness measures.
Still, most of the elements of culture change and blur too easily to qualify as stable replicators. They are rarely as neatly fixed as a sequence of DNA. Dawkins himself emphasized that he had never imagined founding anything like a new science of memetics. A peer-reviewed Journal of Memetics came to life in 1997—published online, naturally—and then faded away after eight years partly spent in self-conscious debate over status, mission and terminology. Even compared with genes, memes are hard to mathematize or even to define rigorously. So the gene-meme analogy causes uneasiness and the genetics-memetics analogy even more.
Genes at least have a grounding in physical substance. Memes are abstract, intangible and unmeasurable. Genes replicate with near-perfect fidelity, and evolution depends on that: some variation is essential, but mutations need to be rare. Memes are seldom copied exactly; their boundaries are always fuzzy, and they mutate with a wild flexibility that would be fatal in biology. The term “meme” could be applied to a suspicious cornucopia of entities, from small to large. For Dennett, the first four notes of Beethoven’s Fifth Symphony (quoted above) were “clearly” a meme, along with Homer’s Odyssey (or at least the idea of the Odyssey), the wheel, anti-Semitism and writing. “Memes have not yet found their Watson and Crick,” said Dawkins; “they even lack their Mendel.”
Yet here they are. As the arc of information flow bends toward ever greater connectivity, memes evolve faster and spread farther. Their presence is felt if not seen in herd behavior, bank runs, informational cascades and financial bubbles. Diets rise and fall in popularity, their very names becoming catchphrases—the South Beach Diet and the Atkins Diet, the Scarsdale Diet, the Cookie Diet and the Drinking Man’s Diet all replicating according to a dynamic about which the science of nutrition has nothing to say. Medical practice, too, experiences “surgical fads” and “iatro-epidemics”—epidemics caused by fashions in treatment—like the iatro-epidemic of children’s tonsillectomies that swept the United States and parts of Europe in the mid-20th century. Some false memes spread with disingenuous assistance, like the apparently unkillable notion that Barack Obama was not born in Hawaii. And in cyberspace every new social network becomes a new incubator of memes. Making the rounds of Facebook in the summer and fall of 2010 was a classic in new garb:
Sometimes I Just Want to Copy Someone Else’s Status, Word for Word, and See If They Notice.
Then it mutated again, and in January 2011 Twitter saw an outbreak of:
One day I want to copy someone’s Tweet word for word and see if they notice.
By then one of the most popular of all Twitter hashtags (the “hashtag” being a genetic—or, rather, memetic—marker) was simply the word “#Viral.”
In the competition for space in our brains and in the culture, the effective combatants are the messages. The new, oblique, looping views of genes and memes have enriched us. They give us paradoxes to write on Möbius strips. “The human world is made of stories, not people,” writes the novelist David Mitchell. “The people the stories use to tell themselves are not to be blamed.” Margaret Atwood writes: “As with all knowledge, once you knew it, you couldn’t imagine how it was that you hadn’t known it before. Like stage magic, knowledge before you knew it took place before your very eyes, but you were looking elsewhere.” Nearing death, John Updike reflected on
A life poured into words—apparent waste intended to preserve the thing consumed.
Fred Dretske, a philosopher of mind and knowledge, wrote in 1981: “In the beginning there was information. The word came later.” He added this explanation: “The transition was achieved by the development of organisms with the capacity for selectively exploiting this information in order to survive and perpetuate their kind.” Now we might add, thanks to Dawkins, that the transition was achieved by the information itself, surviving and perpetuating its kind and selectively exploiting organisms.
Most of the biosphere cannot see the infosphere; it is invisible, a parallel universe humming with ghostly inhabitants. But they are not ghosts to us—not anymore. We humans, alone among the earth’s organic creatures, live in both worlds at once. It is as though, having long coexisted with the unseen, we have begun to develop the needed extrasensory perception. We are aware of the many species of information. We name their types sardonically, as though to reassure ourselves that we understand: urban myths and zombie lies. We keep them alive in air-conditioned server farms. But we cannot own them. When a jingle lingers in our ears, or a fad turns fashion upside down, or a hoax dominates the global chatter for months and vanishes as swiftly as it came, who is master and who is slave?

Adapted from The Information: A History, A Theory, A Flood, by James Gleick. Copyright © 2011 by James Gleick. Reprinted with the permission of the author.
James Gleick is the author of Chaos: Making a New Science, among other books. Illustrator Stuart Bradford lives in San Rafael, California.

Find this article at:

http://www.smithsonianmag.com/arts-culture/What-Defines-a-Meme.html