Engineering a Meme – by “Paul C – YLT2012”

We are all familiar with memes and how rapidly they can explode online. But what exactly is it about a certain meme that elevates it from mere joke to a cultural staple? Is it possible to quantify the factors that constitute the success of a meme? For this project, I set out to do exactly that—through the attempted creation of a few memes on 4chan, I investigated the importance of certain factors that go into determining the success of a meme, such as shock value, inherent humor, current events relevance, and more. This discussion will be broken into three parts: memes I created, past successful memes, analysis and conclusions.

Disclaimer: The content and language in this report, particularly the replies of users on 4chan, may be offensive. This project is solely intended for academic purposes.

I: Memes created for this project

1. Spanish Shepard

Mass Effect is a popular role-playing game series set in space that already has a handful of memes associated with it due to humorous lines within the game itself. The game has no connection to Spanish culture, but a Spanish language version has been released. “Spanish Shepard” takes the face of the main character (Shepard), puts it on the body of “Blondie,” (played by Clint Eastwood in The Good, The Bad, The Ugly) and places him at the scene of the final level of Mass Effect.  The following factors were kept in mind:

  • Non sequitor humor (silliness)
  • Factual incorrectness (“Blondie” is actually American)
  • Bad Spanish (Google translated)
  • Relevance to an existing Mass Effect meme culture

“Spanish Shepard” was therefore meant to test the validity of “silly internet humor.”


More can be found at

Four threads were created  on 4chan’s video game board, “/v/,” over the course of two days. Two threads started by providing the image template and introducing it as “Spanish Shepard.” The other two started by reposting an existing example.


Reactions were mixed, with some appreciating the non sequitor humor, and perhaps surprisingly, some jumping on the factual incorrectness of the image.


Note the implicit understanding that this is a “meme,” despite the criticism.

Active references to the possibility of the birth of a new meme.

Correcting of Spanish.

Implicit suggestion that there is some sort of criterion for something to be a meme.

The full threads can be found in the archive linked at the end of this entry.


With these examples, we can reasonably place some weights on the importance of factors that went into “Spanish Shepard.”

  • Inherent/Non sequitor humor (silliness)

Roughly half of those that replied enjoyed the humor, though it had no direct relation or relevance to Mass Effect at all. The “ridiculousness” of “Spanish Shepard” approached the point where something was inherently humorous, regardless of relevance.

  • Factual incorrectness (“Blondie” is actually American)

Users were most critical of the fact that the premise of the image was incorrect—nothing about the image was actually Spanish. This suggests a framework of reason that goes into the success of a meme. Nonetheless, some users still found the image humorous, which suggests that enough inherent humor eclipses any factual issues.

  • Bad Spanish (Google translated)

This prompted only a couple replies, but there was no indication that perfectly correct Spanish would have made the meme more successful.

  • Relevance to an existing Mass Effect meme culture

This did not come up at all, which suggests that memes are independent of legacy.


“Spanish Shepard” can be characterized as a modest success—no independent propagations or reposts of the image were confirmed, but it prompted discussions of whether or not it had the potential to become a meme, as well as criticisms of why it couldn’t. It also provided a rough ranking of which factors are most important to a meme’s success, which will be explored further in the following examples.


2. Steph Jerbs

On November 4th, a user posted the following image on 4chan’s technology board, “/g/.”

The OP (original poster) took a picture of a portrait of a Steve Jobs tribute at his school to share with /g/, with the implicit understanding that the facial structure and body proportions were disfigured—prompting users to reply with the picture captioned with distorted Apple slogans, as if to emulate the speech of a person with autism.

The original thread died with only seven replies, and was not seen again. To clarify, I did not post any of the replies in the original post. I saw the thread and filed it away for later use as part of this project.

In December, I began reposting the image, with my own additions, as replies to various Apple-related threads on /g/.  The full image set can be found here.


The following factors were kept in mind:

  • Offensiveness – The image directly makes fun of the speech of autistic persons.
  • Relevance to existing anti-Apple culture on /g/
  • Revitalization of existing content
  • Accessibility of template

The results are somewhat surprising. Over the course of three days, I specifically targeted threads related to discussion of Apple products (numbering twelve in total), posting “Steph Jerbs” as replies, never starting a thread of my own. However, the image was largely ignored. One caveat to note is that I did not provide the template for the image. The goal was to see if the existing images would propagate on their own, not necessarily to encourage production of more macros.

Finally, three days after I stopped posting replies, I started a new thread providing the template for the image. One independent repost of an image I had posted earlier was confirmed. There was also one reference to the possibility of a new meme, and one other post suggesting the idea of the meme was not novel.

“Steph Jerbs” can therefore be characterized as a total failure, insofar as to say that only one other person was confirmed to having saved a “Steph Jerbs” image, with almost no discussion of the image’s potential as a meme.


We now go back to the factors that went into this “meme:”

  • Offensiveness – The image directly makes fun of the speech of autistic persons.

Perhaps most surprisingly, offensiveness played no part in the success or failure of the image. 4chan is known for being a crass and offensive community, yet in the case of memes, the failure of “Steph Jerbs” would suggest that offensiveness is not an important factor in determining the popularity of a meme.

  • Relevance to existing anti-Apple culture on /g/

Though I limited my posts to only threads related to Apple, this also seemed to have little effect, suggesting that relevance to existing culture plays a minimal part—a conclusion that aligns with what we saw in “Spanish Shepard.”

  • Revitalization of existing content

This was included because, as we will see later, many currently popular memes enjoyed a period of hibernation between their first introduction and their eventual hit. However, “Steph Jerbs” would suggest that this is not all that important.

  • Accessibility of template

The template was provided from the get-go in “Spanish Shepard,” and was not provided in “Steph Jerbs.” However, neither produced any independent derivations, suggesting that the accessibility of the template is not a large factor in a meme’s success.

We now have a more refined idea of what makes a meme successful—the greatest importance seems to be its inherent humor, while details like relevance to existing culture and factual correctness seem to take a backseat to humor. Smaller details like accessibility have minimal effect that would require further study to illuminate, while offensiveness seems to have a much smaller effect than would be expected.


3. “Do you really have to…”

Adult swim, a late night television network, produces extremely eccentric video shorts for its viewers to enjoy in the early hours of the morning. One such example is the “Tim & Eric’s Show,” and one episode features a dwarf person singing on what appears to be the set of a talent show. Video here: The song’s catchphrase appears at 0:50, and the lyrics are: “Do you really have to pee in a girl’s mouth / to make babies?”

I took this phrase and attempted to turn it into a meme, with moderate success. The goal was to turn this into a stock phrase of the form “Do you really have to X to Y?” With that in mind, I prepared a list of derivations that drew on existing memes. Examples:

I then posted a thread on 4chan’s Random board, “/b/,” beginning with the original phrase and replying with the rest.

The following factors were kept in mind:

  • Generality: the phrase offers no constraints on the type of content it may be used for. It only has to be of the form “Do you really have to X to Y?”
  • Inherent humor/shock value: the original line is quite creepy and humorous
  • Accessibility of template: since it is just text, it is easily accessible and requires minimal effort to produce a derivation

There were additional challenges for this case in particular however, because I was posting on /b/:

  • The sheer volume of posts would make mine easily lost
  • /b/ now features ID hashes for each unique user in a specific thread, which would make it obvious that I was the only one posting
  • 444444444 GET was happening (where users all spam in an effort to try and “get” the post number 444444444)

Despite these challenges, the meme was a moderate success! I posted two threads in two days, and each received numerous independent derivations, totaling over 25 people who independently created new content based on my original post. There was also no mention of the phrase “becoming a new meme,” which, counterintuitively, is a good thing—the meme becomes organic, not a “forced meme.”

Now going back to the factors that went into this meme, we can see why each factor was important:

  • Generality: because the premise is free, there is virtually unlimited possibility for derivations
  • Inherent humor/shock value: the rhetorical nature of many of the derivations makes it generally humorous
  • Accessibility of template: this is closely related to generality, but the text-based nature of the meme requires extremely small effort on the user’s part to spread.
  • Added effect: the general nature of the meme made it easy to reference other existing memes within it, solidifying its potential as a new meme

What we can take away from “Do you really…” is that generality and ease of integration into the existing meme framework seems to be the most important factor in determining the success of a meme. There was no specific cultural relevance that accelerated its development, and the original context of the meme (peeing in a girl’s mouth to make babies) became less important as more derivations came in. The value of the meme is in its general, rhetoric nature, not in any specific circumstance.


II: Past Successful Memes

We can now take a look at other memes to see how these factors fit into their successes.

 Greentext Stories

This is perhaps the most general of the popular memes today, as it only requires the specified text to be prefixed by a “>” sign, highlighting the text as a green “quote” on 4chan. Because of its wide-spread use in other contexts (“>” is used in emails to designate quotations as well), it’s origin as a meme is hard to pinpoint, but its popularity can’t be understated. Generality here is the only factor, as there are no cultural references or inherent humor built into the “>” mechanism. Instead, the humor comes from the content of the story that comes within the greentext, which may vary.

The Ass Was Fat

The inherent humor in this meme stems from the shock value of having a children’s cartoon be used in such a vulgar way. There is also an aspect of generality to this in that the phrase can be used to refer to humorous situations in a general sense, as the link shows. Here again, cultural relevance is not an important factor.

That really rustled my jimmies

This meme is general in that it can be used as a reactionary post for a variety of situations. As the link shows, the phrase can also be modified and applied to other existing memes to facilitate its spread. There is great value in the humorous phrase “rustled jimmies,” but I would argue that the humor is secondary to its generality. It is the generality that enables the humor to spread easily to other situations, thus creating a meme. One thing to note is that like “Steph Jerbs,” “rustled jimmies” also went through a phase of hibernation between its first introduction and eventual success. But “Steph Jerbs” was highly specific, while “rustled jimmies” is not. Therefore this suggests that a hibernation period is not an important factor in a meme’s success.

Another point to take away from these successful memes is that none of them are particularly offensive in any specific way. In order to further explore the concept of offensiveness in memes, it is worthwhile to analyze the memes that came out of the Connecticut Sandy Hook elementary school shooting.

Adam Lanza

Though distasteful, we can gain useful insights into meme culture by viewing the Sandy Hook shootings through the lens of what factors go into making a meme successful. In the days after the shooting, /b/ was filled with numerous threads making jokes about the shooting, with many striving for meme-status. It would take far too long to systematically analyze the failures of each one, but suffice it to say, none of them really made it past a couple of days. This is evident in the volume of threads and posts relating to the shooting on /b/: after the first two days following the shooting, the number of threads relating to the event dropped significantly. Twenty-four of these threads are saved in the archive available at the bottom of this post.

However, one picture appeared with regular consistency in each of the threads, and continues to do so: a picture of a young Adam Lanza in a blue polo shirt in what is presumably his home, looking at the camera with an eerie grin on his face. More so than any elaborate joke or copypasta, this image appears to have survived as a general and ubiquitous meme-characterization of the event.

In my view, this picture is not offensive—it merely depicts the person responsible for the shootings. To use the terms of this analysis, the picture is “general” in the sense that it can be used as a general representation of the event. This result is similar to the Aurora Colorado shooting, after which many memes initially appeared, but ultimately died out. In the end, the one meme that emerged from the Aurora shooting was a picture of the shooter James Holmes sitting in a courthouse—a general classification of the event.

The fact that the resulting memes from these events were general pictures, and not more offensive copypastas (of which there are plenty), speaks volumes about the importance of offensiveness in determining a meme’s success—that is to say, offensiveness is not an important factor.


III: Analysis & Conclusion

Throughout this report I have introduced some possible factors of what makes a meme successful. Keeping in mind that this project only produced three “memes,” these results are meant to be broad classifications, and readers are encouraged to take them as merely suggestions, not conclusions. However, combined with the analysis of past successful memes, I am reasonably confident that this is on the right track.

Factors (in proposed order of importance):

  • Generality
  • Inherent humor
  • Cultural/current events relevance
  • Accessibility of template
  • Factual correctness (where applicable)
  • Revitalization of existing content
  • Offensiveness

Our examples have shown that the most successful and popular memes all share an element of generality, in that they can easily be applied to a number of situations. Following this is the inherent humor, which is harder to quantify but relatively easy to identify—we know it when we see it.

This suggests that for a meme to succeed, it must first be general. This can be thought of as its “raw potential.” Then follows the humor index—after a certain threshold is reached, the meme explodes. Other factors like cultural relevance (the shootings) may also cause the meme to take off, but their thresholds are placed much higher.

I envision the metric for measuring a meme’s success to be something like this:

The red curve represents the meme’s raw potential, that is, the generality. The more general it is, the greater chance it has to take off. The horizontal green line represents the humor threshold—once this level is met, the meme can take off, but until such condition is met, it cannot gain momentum. The bounds are normalized at 0 and 1—the most successful meme in history would asymptotically reach 1—memes that are both insanely humorous and incredibly general. Rickrolling might place highly on this metric, in which case we can specify an arbitrary second horizontal line, and say that once a meme’s success value crosses that line, it is able to “break out” of the internet and into the real world.

If we were to place current memes onto this graph, most would sit on the plateau right near the humor threshold—mediocre in quality, local in scope. “Spanish Shepard” might be on this plateau, while “Steph Jerbs” would be in the pit close to zero, while “Do you really…” might place just above the humor threshold—only time will tell if it truly takes off.

Again, this model is meant to be a rough way to graphically organize the successes and failures of memes. The exact mathematics of this model would need a more in-depth study and quantification of the humor and generality of memes, but an exponential function to represent the generality, combined with a polynomial factor to include the humor threshold, might do the trick. Some might say that it is futile to attempt to quantify such intangibles, but economists have spent decades trying to quantify utility, so why not the success of memes? Generalizing with models can be a useful tool to eventually understanding the sociology of memes.

Memes appear all around us online as ubiquitous jokes that seem to have no specific origin or reason as to why they are popular. Through this project, I have attempted to illuminate some of the factors that determine a meme’s success, and the results are simpler than one might expect—humor and generality. Things like cultural/current events relevance and offensiveness might seem relevant at first, but I contend that they take a back seat to more important traits (and in the case of offensiveness, no seat at all). In this we see that online memes are eerily similar to biological ideas of evolution and spread. An analogy might be made between memes and viruses—their most important traits are that they are able to spread, supplemented by a baseline value for effectiveness (in a meme’s case, humor).

Once again, this project was meant to be a broad introduction to the quantification of memes, and what kind of metric one might use to do so. The limited sample size hinders the validity of the results, but my hope is that this will provide a stepping stone to those that wish to pursue this idea further.

Paul Chung

Zip file of source threads:

Edited for size of pictures. Added archive link. Fixed graph axis error.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s