Original research by Emilija Jokubauskaite, Daniel Bach, Sal Hagen.
This blogpost is based on a panel talk given at the RESAW conference 2019 “The Web that Was”.
In academic and journalistic coverage, the infamous imageboard 4chan has often been treated as an amorphous mass. The reasons for this have to do with a combination of factors, one of which being the fact that anonymous posters to 4chan tend to think of themselves as part of a blob or mass (recall, for example, Anonymous’ slogan “we are legion”). This view of 4chan as a subculturally coherent mass, is however arguably shaped by the site’s technical affordances. As discussed by others (e.g. Bernstein et al. 2011) and within our own work (e.g. Hagen 2018), 4chan’s two most significant affordances are anonymity and ephemerality. What makes 4chan relatively unique as compared with most social media is that its users, or “anons”, have no portable reputational capital, hence the use of memes as a way to express in-group status. Furthermore, the imageboard has no consistent archiving function, the latter which contributes a sense of fleeting experiences. In the case of the most toxic (and lately, most popular) part on the site, /pol/ “Politically Incorrect” board, the blob-theory of 4chan has led some commentators to make overgeneralizations. The reasons for this are understandable as the board has been associated with the violent extremism of the alt-right. Yet it also runs the paradoxical risk of bestowing the board with an “almost godlike power” (Phillips et al. 2017). While this research does not argue that the radical right is not a cornerstone of /pol/, it does propose that “deblobifying” the board can lead to better insights into the present and past of 4chan and Internet culture more generally. To do so, we empirically approach 4chan/pol/ through looking at the subcultural practice of creating “general threads” and provide a methodology for a more granular distant reading of the board.
On ephemeral imageboards like 4chan or 8chan, creating and posting in “general threads” forms a way to maintain long-lasting thematic conversations in spite of the software’s automatic deletion of data. In practice, this concerns groups of users manually creating new threads with repeated opening texts on an overarching topic, for example the war in Syria. For some generals, this is done even before the previous post is deleted. For example, “/ptg/ President Trump general” is posted on /pol/ as often as every half an hour, allowing for specific discussions to continue and live updates to be included. Our previous research had mapped the significance of the “Pizzagate general” threads for the creation and early development of the now-infamous conspiracy theory (Tuters et al. 2018). We have also explored the organisation of the responsibilities of posting these opening posts in two of the most frequent general threads, namely “/sg/ Syria general” and “/ptg/ President Trump general” (Bach et al. 2018). In essence, general threads can be understood as a cultural practice that aims to counter the above-mentioned ephemerality of the imageboard’s software.
Zooming out from the discussed individual case studies, this blog post presents an overview, temporally mapping the most prominent themes on 4chan/pol/ through the general threads found between 2014 and 2019. Even though an “anon” on 4chan can hardly be considered via the normative idea of a democratic citizen, this research proposes that the different general threads can be regarded distinct issue publics, that is, their involvement in politics is “a practice that is occasioned by issues and dedicated to their articulation” (Marres 2007, 775).
The generals can be recognisable through the “OP subject”, i.e. the title of a thread, usually containing the word “general” and/or abbreviations in-between /slashes/ (example in Image 1, below). Needless to say, detecting such manually constructed posts that include typos and evolving titles, is not an easy task. Nonetheless, general-specific text queries can capture most of it. The dataset used spans from December 2013 to May 2019, comprising 4.837.299 opening posts in total (excluding the comments on these first posts). As noted, the majority of the general posts include either the word “general”, like “president trump general”, and/or an abbreviation between two slashes, like “/ptg/”. Therefore, we first extracted all of the posts that included more than 10 instances of the same abbreviation between slashes or more than 10 instances of the same words or phrases before the word “general” in the subject of the opening post. This provided us with 99.953 instances of general posts. After removing false positives, combining generals, and writing separate queries for each general title, we gathered separate datasets of 330 general threads, with anywhere between 10 and 28.543 opening posts in each. All general threads were then categorised thematically into 31 categories.
Compared to all opening posts, the proportion of general posts in the dataset comprises 2%, ranging between 0,13% and 3,4% per month. When the comments are taken into consideration, the percentage rises to 4.38% (ranging between 0.27% and 12%), showing that, on average, general threads have higher comment ratios than “regular” opening posts. The general posts are thus not representative of all of 4chan/pol. However, the fact that they are produced manually and recurrently speaks of certain importance of these issues. As such, this blog post maps out the more organised part of a rather unstructured online space.
What general threads are most prominent over time?
By looking at the themes of general posts (Image 2, below), issues on Trump, country politics, and wars or conflicts are most dominant. This proportion is impacted by the two most active general threads: “/ptg/ President Trump general” and “/sg/ Syria general”. Interestingly, the category of left-wing ideologies takes up 3.5% of all general posts, similar to religion and various right-wing ideologies. This, of course, does not argue that “/pol/acks” are leftists, but it does show that there is a small issue space where leftist thought is discussed and/or ridiculed.
More traffic = more generals?
When looking more closely at one of the previously mentioned categories, the post activity of each of the separate general threads can be mapped out chronologically. One example is shown in Image 3 – the category of left wing politics – comprised of 12 general threads each represented on a separate timeline. Each dot on a given timeline shows an opening post and the surrounding sphere is sized according to the number of comments it received.
Viewing all these visualisations together (Image 4, below), the majority of general thread activity seems to be condensed after mid-2015. At that time, a lot of new generals emerged and previous ones became more consistently active.
To examine this trend more closely, the line graph below (Image 5) shows absolute numbers of all opening posts on 4chan/pol and all general posts. We can see that while the overall number of posts on 4chan starts growing in 2014 and peaks at the end of 2016, the number of general posts starts increasing slightly later, around the end of 2015 and peaks later. Similarly, when looking at the proportion of general posts on the board every month (Image 6), up to a certain size of overall activity, the proportion of generals is insignificant (below 1%). However, when /pol/ grew, the proportion of generals also increases up to over 3 percent (squares in dark blue). Possibly, more traffic brought more ephemerality to the board, since frequent posting “speeds up” the pace of automatic archival, we can posit that the demand for generals as a means for continued conversation also grew. The increased frequency of general threads, however, stuck even after the peak in overall activity.
Rhythmic classification schemas
Apart from categorising the general threads in terms of the issues they deal with, we may also divide them according to the ways they counter ephemerality. In other words, we can look at their “rhythm”: frequency of the posts, lifetime of the threads, and the activity of the discussions in the comments. Four distinct rhythmic categories then emerge, which we’ll discuss here briefly.
Firstly, if we look to long lasting, consistently frequent and well-engaged with general threads (Image 7) — that is, generals that deal with issues of high interest that exist for years and are signified by strong and frequent engagement. Examples include /ptg/ President Trump general, /sg/ Syria General and /brit/pol general. The second category can be called live 4channing, denoting threads that have a short lifetime but a very frequent and high engagement. These mostly cover unfolding natural disasters, political events or are aimed at mobilisation, such as with the gilets jaunes in France (Image 8). Time-specific generals can be seen as a third category (Image 9). These have a short to medium lifetime, but differing frequencies and engagement — some may be more active than others, and some may disappear and reappear. These include generals on politics in specific countries or regions, politicians, celebrities or conspiracy theories. The last category includes issues of low interest — differing lengths of lifetime, but sporadic engagement, often inactive for long periods of time (Image 10).
Not surprisingly, some thematic categories correspond better to specific types of counter-ephemeral engagement. For instance, the disasters (Image 11) category is mostly centered around (natural) disasters happening for a limited amount of time — these generals are therefore very short and consistently frequent around the event. Other categories, for example religion (Image 12), political ideologies, country or regional politics present a bigger diversity in countering ephemerality and also often a longer lifetime.
We may then argue that according to how successful the general threads are in countering the ephemerality of the board, the previously mentioned four rhythmic categories can be placed on a spectrum, with the first one being the most and the fourth being the least successful. It can further be said that the most consistent generals present the most relevant issues for many anons, rather than simply the categories that have the largest general post counts or those general threads that have a lot of comments. It is the combination of these different metrics that point us to the issues that “spark a public into being” (Marres 2005) most effectively.
All in all, while general threads make up a rather small percentage of the overall posts on the board, their significance can be observed in the creation of (sub)cultures, as seen previously in the cases of Pizzagate or QAnon. This ongoing research into the overview of general threads has shown us that it is possible to empirically disentangle an anonymous ephemeral mass, or, in other words, “deblobify” it. In doing so, observed patterns in the practices of general-thread-use led us first to detect the dominant issue spaces — mainly right-wing and having to do with Trump, radical or country politics, conflicts and religion. On a more mediatic level, our findings allow us to argue that more traffic on the board produces more ephemerality, in turn, creating a need to counter it with specific cultural practices such as general threads. These copy-pasted opening posts may be looked at through the lens of their success in countering ephemerality, which is one of 4chan’s most characteristic affordances. As a methodological take-away, researchers of 4chan and similar boards may focus on studying general threads rather than individual metrics such as post or comment counts. In addition to providing a means by which to disentangle specific thematic conversations from the mass of anonymous chatter, the general threads on 4chan offer an exemplary case to study how user practices negotiate with and overcome limitations posed by specific platform affordances.