Widening Epicenters

Online culture is a fascinating thing. I've talked about it a lot, pointing out individual community behavior, the sizes and lifespans of online communities, community regulation, and community leadership. But now I want to ask a broader question: How has online culture changed over time?

Different question from the center of the Internet or the end of the Internet. (Image source here)

Once upon a time, I once drew a line between "old" and "new" internet culture, with Anonymous as the border between the two. 4chan used to be considered the epicenter of online culture on the Internet. Nowadays, that title might go to Reddit.

But that's just me talking about anecdotes. Can we demonstrate that there is such a thing as online epicenters? Where is it? Where has that title drifted to over the course of the Internet?

Part 1: The Word You Heard

First things first: What does the Internet say about Internet epicenters?

The Daily Dot, an online newspaper dedicated to Internet culture news, features many online hubs - YouTube, Reddit, Tumblr, and others - under its "communities" header. This tells us about the websites that a media outlet believe are worth the most attention. However, this only tells us what websites are particularly important today. We don't learn anything about what websites used to be important, which is important information if we want to track how the Internet's center of attention drifts.

Let's turn to Cracked, ever the surprising source for interesting information. Cracked made two skits in 2008 where they imagined relevant websites to be attendees of a party. They did a follow-up skit with the same premise in 2013. Let's look at the websites that Cracked chose to represent online culture, and how their choices changes over five years.

2008:

2013:

There are some pretty clear differences in what websites are featured. Between 2008 and 2013, Facebook has overtaken Myspace, and Reddit has overshadowed Digg.  2008 covered a lot more websites than 2013 did as well (35 vs 19), though 2008 had the advantage of featuring two videos. A lot of 2008's featured websites are still around today, but not so prominently noticed in our time online, either because they are services that we've come to take for granted (Mapquest, Paypal), or because we simply aren't as amused with them anymore (Urbandictionary, Wikipedia).

2008 Internet's websites were fairly distinct from one another, but some website attendees were somewhat "redundant". For example, AskJeeves and Google are both present, despite essentially having the same functionality.  2013 Internet's attendees are somehow even more redundant, portrayed as visibly circling around themselves and their web content. How meaningfully different is Buzzfeed's content from Tumblr or Reddit, for example? These videos give us the impression that the Internet in 2013 is more insular; the website diversity of 2008 seems to have given way to multiple redundant websites that recycle lots of content from one another.

But these are just the perspectives of one group of people. Plus, 2008 isn't even all that early in Internet culture. We need more data.

Back in time, through the information superhighway!

Part 2: Alexa Data

We have used Alexa before for analyzing trends online (being aware of its method and limits). I've also previously cited things using the Web Archive, a website devoted to taking "snapshots" of websites over time. The Web Archive stores records of websites as far back as the '90s, preserving the original website format and content. It is the perfect resource for digging up forgotten online pages that may not be readily accessible anymore.

So if Alexa is a website devoted to gauging the popularity of websites, and if Web Archive is a website devoted to archiving old websites, then we can hypothetically look for archived versions of Alexa to gauge the popularity of websites over time, right?

Unfortunately, Web Archive is not perfectly comprehensive, so some data on website rankings really are lost forever.  The Top 500 websites pages are usually the most consistently archived pages, so I looked through the top 500 list in various points of time that were available. These archives went as far back as 2002, a significant improvement to our old 2008-2013 mark.

Let's dive right into this data (and flail around in it, splashing data on our friends). Click any image to enlarge it.

With this first graph, we can observe the competition between social networks over the years. Friendster preceded them all, but Myspace topped the charts until about 2008. Facebook edged ahead after 2008, and the rest is history.

Friendster and Orkut seem to battle it out as the best second-tier social networking site until they both begin to fall off past 2010. Myspace, Friendster, Orkut, and Facebook all offered remarkably similar products, and ultimately the Western world chose Facebook. The only social networks that have stayed afloat since Facebook's rise to dominance are networks that offer specifically different functionality.

Twitter makes a huge jump from 2009 to 2010, now at the point where it directly competes with Facebook for public attention. LinkedIn hovers around sub-100 rankings for several years until finally catching on like wildfire in 2010. Perhaps this was a delayed response to the economic recession. Since Facebook is not built around the premise of professional networking, LinkedIn's functionality could offer a unique service in a time of great demand for it - job hunting assistance.

Perhaps now we should move on to a different sector of online interaction: blogs.

The blogging world has been consistently dominated by the Blogger/Blogspot domain since 2005, which might have something to do with Google buying them out in 2003 and some major site redesigns from 2004 to 2006. WordPress jumps into the public limelight around 2007, with Tumblr following in 2010. They currently co-exist, much like how we saw with Facebook, Twitter, and LinkedIn. Their features keep them differentiated: Wordpress offers downloadable software and independent hosting, Blogger enjoys Google's popularity (and knack for ease of use), and Tumblr's micro-blogging that is better suited for networking and rapid sharing of information.

However, it seems as though there used to be a different trio of popular blogging software in the mid-2000s: Blogger, LiveJournal, and Xanga. Xanga had a good run in the mid-2000s, but probably sputtered out as the online audience collectively realized that they shouldn't write bad poetry and post it on the Internet. Meanwhile, LiveJournal continues to exist, though it seems to be on the steady decline.

Why did WordPress and Tumblr overtake Xanga and LiveJournal? It could have to do with Wordpress' superior design over Xanga's defectiveness and Tumblr's faster reblogging/sharing capabilities, but a more definitive answer would require some research that isn't directly relevant to the point of this particular blog post. For now, we look at the next data set.

Here's some data that highlights the rapid changes in image hosting during the past three years. There were many image hosting websites through the '00s, among them Flickr and Photobucket. Notice how these services are losing ground to the likes of Instagram, Pinterest, and Imgur. Imgur is very intimately tied to Reddit, being the website's image host of choice. Instagram and Pinterest saddle a line between social networking and image hosting, giving them an extra push in appeal. Photobucket appears to be facing certain decline in use, though it remains to be seen if Flickr will continue its latest downward trend.

Before we get to the last graph, I want to mention that there was some other Alexa data that I didn't bother to graph out. Some interesting ones include the Drudge Report poking its head into the top 500 from 2004 to 2006, disappearing, and then coming back 2010 onward. Perhaps the first hump had to do with the 2004 presidential election, and then 2010 had to do with the rising Tea Party movement? YouTube and Wikipedia also consistently rank very highly, and Dailymotion has managed to stay in the top 100 since 2007.

But now, I attempt to sandwich all the rest of the online community data into one big mess of a graph:

DeviantART has been a consistent hit online, despite its strange, strange content. I was surprised that fanfiction.net - a website devoted to hosting user-submitted fan fiction - managed to get into the top 400 rankings from 2007 to 2009. You'll notice that The Pirate Bay - a website functioning as a cache for pirated materials - enjoyed high levels of popularity until they were shut down a few years ago.

InvisionFree, a website that hosts free forums, saw its rise and fall in popularity from 2006 to 2009, while Digg and Reddit started gaining popularity. Ebaumsworld briefly featured in the top 500 in 2005 and 2006, but might have had its "business" model threatened when other video-hosting and image-hosting services started appearing. Meanwhile, 4chan has a very brief stay in the top 500...in fall 2007.

Huh. That's funny. You'd think that a website as culturally hyped up as 4chan would be more prominent on this graph. And didn't their Time magazine stunt happen in 2009? Why can we only see evidence of 4chan's presence by a single data point in a year that wasn't even its most public?

And now we run into the problem with Alexa data: everything is a relative measure. It isn't necessarily that 4chan wasn't popular enough for the top 500 for all those other years; it's more that every other site was attracting more traffic than 4chan. With the numbers of people using the Internet growing steadily over the years, it isn't that fewer people view 4chan, it's that the new people are being directed to places other than 4chan. Even 4chan users have taken notice of how new sites like Reddit divert the attention of new online users away from them.

This is a huge problem for us, because it signifies that a lot of interesting online activity can still be happening below the top 500 mark. Using Alexa alone, we would think that the blips of data for some of these websites would represent the peaks of activity as well as impact, when we'd be wrong. In fact, 4chan is more popular than ever.

This limitation of the Alexa data is even more evident when we note what didn't get mentioned in the Alexa graphs: CollegeHumor, Funny or Die, Vimeo, SomethingAwful, and Kickstarter don't register at all. What's more, Alexa's 500 list is a representation of the globally most visited websites. Western Internet culture tends to interact very little with websites outside of the English-speaking Internet, which means that what seems popular on Alexa may not properly represent the state of Western Internet culture.

We must find an alternative source of data. Quantcast has been our alternative source for web rankings and served as a great resource for finding online racial demographics.  Unfortunately, Quantcast has WebArchive backups that only go as far back as 2008.


Part 3: Google Trends Data

Instead, we'll look into Google Trends, which will allow us to compare the magnitude of difference in popularity between two given search terms.

A major shortcoming of this method is that we're strictly looking at google searches for these websites, not visits to the website themselves. Someone Googling 4chan would register as a data point in these graphs, but someone going straight to 4chan.org would not be included. If people have their favorite websites bookmarked, then we probably won't get to see their impact on the data. Also factor in that websites like Facebook get a lot of business-related attention, which would inflate their search numbers.

That said, Google Trends goes as far back as 2004 and provides a much-needed dimension of analysis that Alexa cannot provide.

This first graph compares some important players in the older days of the Internet: YTMND, SomethingAwful, and General Mayhem. GenMay was a forum that had seen its best days far before 2004, though you can see a few spikes in activity around when the "Christopher Walken for President" campaign was in full swing. SomethingAwful, despite having some serious cultural clout on the Internet, has very meager numbers even in the mid-00s, though this is probably mostly caused by their paywall. YTMND sees its greatest spikes in activity in 2006, around when Eon8 was gathering momentum and when they waged their internet war on Ebaumsworld.

Ah, yes, Ebaumsworld! What's their traffic like, by comparison?

As we go from graph to graph, I'll leave a website from the previous graph so that you can have a sense of the changes in magnitude. In this graph, YTMND is your benchmark. Be careful! YTMND went from being colored red on the previous graph to blue on this one, due to the quirks of Google Trends.

We see that, compared to YTMND, Ebaumsworld was very active. Perhaps this makes sense in the context of that old Internet war: Ebaumsworld was the larger entity, staying popular by aggregating the content from smaller websites. Running contrary to that narrative is Newgrounds, a popular flash-hosting website that exceeded even Ebaumsworld's popularity. Interestingly, the peaks in popularity for Newgrounds and Ebaumsworld seem to happen around the same time that YTMND peaks in popularity, around 2006.

I also included some extra websites. HomeStarRunner was a flash-hosting website that housed a recurring cast of characters. College Humor is a comedy website that hosts skits and funny pictures. Both of these websites saw their greatest peak in activity before 2006, with HomeStarRunner's greatest peak possibly occurring before 2004. Their overall online popularity seems most comparable to YTMND.

Newgrounds was an interesting fixture in the earlier Internet because of its ability to host animations and short videos. Video hosting used to be a rare and decentralized thing on the Internet, with videos spread across websites like Ebaumsworld, Albino Black Sheep, and a bunch of others. That is, until...

...YouTube showed everything up. YouTube launches in 2005, and overtakes everything that preceded its video hosting services. Note that the window of data in that graph is from 2004 to 2006. If I had extended that window, then YouTube's numbers would have completely dwarfed the other figures. Moving on...

Here, Newgrounds is our website of reference from the previous graph, in blue.

A lot of other websites seem to be approximately on par with Newgrounds in popularity. LiveJournal peaks in 2005, then begins its slow and steady decline. Digg seems poised to get bigger up until mid-2007, where it loses steam and also begins to lose popularity. Fark, a news aggregation website, experiences a fairly consistent level of online attention until 2009, where it abruptly shoots up in activity, plateaus, and then also begins to lose its audience.

Notice that, by mid-2008, 4chan surpasses Newgrounds in popularity and goes completely unmatched. Strangely enough, 4chan's activity seems to peak around 2011, also defying our expectations that 4chan's greatest moments were in 2007-2009. It also doesn't seem to obviously agree the Alexa data, which reports 4chan's highest Alexa ranking in fall 2007.

This seems to confirm the idea that despite 4chan's continued accumulation of new users, other websites in the Alexa top 500 were simply accumulating users more quickly than 4chan. It also says something very crucial about 4chan's lifespan: it had not yet attracted its fullest audience when it was doing its most infamous things. It's safe to say that this graph defines very clear borders on when 4chan was the "big thing" on the Internet.

But still, we press on.

The torch gets passed from 4chan to Reddit in 2011. From that point on, Reddit's popularity surges at a rate much higher than that seen of its predecessors. Scroll back up and compare Reddit's graph to the first graph featuring SomethingAwful. When Pedogeddon took place in 2012, Reddit was already larger than 4chan had ever been.

Xanga and Gamefaqs are included here for some extra comparisons. Xanga was evidently very popular back in its heyday, with its peak in 2006 (what is it about that year?) exceeding 4chan's 2011 peak. Perhaps this is a good time to start comparing blogging platforms.

Reddit's success suddenly seems completely dwarfed as we zoom out to accommodate Tumblr. Xanga's natural peak seems absolutely squished as well. Even Wordpress gets left in the dust. Pinterest is interesting because of its extreme surge in popularity - check out that slope from 2011 to 2012. That's even faster growth than Tumblr. For whatever reason, Pinterest hits a ceiling, bounces down in popularity, and then maintains a more level amount of popularity.

At this point, what constitutes online subculture begins to get blurry. Reddit, the so-called "front page of the Internet", is quickly outdone by Pinterest and Tumblr. While Tumblr certainly has a place in the canon of online subculture, can Tumblr users themselves be called a subculture anymore if Tumblr is this popular? To drive Tumblr's popularity home further, here's another one featuring the microblogging platform:

Despite DeviantART's consistent appearance in Alexa's top 500 lists, the online hub is completely overshadowed by other websites. Neopets - an online game that also frequently appeared in Alexa's list - also can't compare to the modern successes of Tumblr, Pinterest, and Instagram. Instagram threatens to overtake Tumblr in popularity, assuming that it doesn't immediately hit its peak.

But as we zoom out even further, we can find websites that even knock Tumblr down to size.

Now we enter into the realm of social networks. Friendster is the only social network whose popularity could not contend with Tumblr. Oddly enough, Friendster got its most searches in 2009, well after the peak of its Alexa ranking as a social networking website. Orkut enjoyed modest success until sharply declining in 2012. Compared to Twitter, however, Tumblr never stood a chance.

This graph also gives us a clear view into the popularity of Myspace in its prime. Myspace hit its peak in mid-2007, and was gathering even greater attention than Twitter is today. That should tell us about the number of people using the Internet back then. Even while websites like Newgrounds, YTMND, and others were peaking in 2006, MySpace towered impossibly high above them, representing the masses of people who weren't necessarily engaged in Internet subculture.

But even MySpace's impressive hump gets squished in comparison to Facebook. Facebook sits as the top social network in a time when more people are using the Internet than ever before. Nothing compares to Facebook's popularity. Not even porn.

Now scroll back up to the first graph, with YTMND, SomethingAwful, and GenMay. These sites were considered hotbeds of online memes. Look at how they compare to Facebook and the websites of today.


Part 4: The Strange Conclusion

Every iteration of online culture has been greater than the iteration before it. The number of people using the Internet has gradually been growing over the past 10 years. The highest peaks of yesteryear's Myspace do not compare to the highest peaks of today's Facebook. There are ebbs and flows to the relevance of websites, but the most important thing to note is that the waves of popularity get larger every time.

We can't speak for online trends before where our data extends, but there seems to have been a definite era from 2004 to 2007 when websites like Newgrounds, Ebaumsworld, YTMND, and others were the Internet subculture. Xanga, Livejournal, and DeviantART was a broader part of that subculture where people posted writings, poetry and artwork. MySpace was the "mainstream" Internet, firmly towering above the tiny websites but not privy to the interesting things happening among them.

Back then, memes weren't even called memes. YTMND called them fads. Fark called them Farkisms. Sure, some online in-jokes could get more public attention than others, but there wasn't any such thing as Internet-wide jokes. There wasn't any sense of a larger, global online presence.

4chan changed all that. 2007 to 2011 marked an online era where 4chan was the capital city of Internet subculture. It was in this era that the term "meme" started being used to describe what these online fads "really" were. Those memes spawned strange government videos and new business enterprises. So many memes were generated that people started feeling the need to keep track of them.

Couple that with the actions of Anonymous during this era, and you begin to see the emergence of an online "identity". Before this era, an online user was most strongly defined by what websites you go to. During this era, the smaller online hubs began deferring to 4chan's online subculture. The new distinction was whether you spend your time on the Internet, or if you spend most of it offline.

In 2011 and beyond, even that question was rendered invalid. In this post-4chan era where Presidents go on Reddit, almost everybody is on the Internet. It's not a question of whether you've heard of Reddit or Tumblr, or whether you "go" on them. You have gone on them. Everyone has. You might hang out on one site more than other sites, but at no point are you left unaware of what other websites are up to.

We all are, Jim. We all are.

There are ebbs and flows to the relevance of websites over the years, but the waves of popularity get larger every time a new website becomes relevant. If we compare them side by side, the popularity of contemporary Internet culture completely dwarfs the Internet culture of 2004-2007.

Online culture was so marginally small before the rise of 4chan that it doesn't even register on the graphs today. It's a shame that we can't look at data before 2004, because I would guess that we'd see one more "wave" of Internet culture in GenMay, LUE, and perhaps some others. Perhaps we could even go further beyond that, looking at Usenet and Gopher traffic. As popular as those online spaces may have been once, their numbers would be laughable today.

What's more, the line between Internet subculture and "mainstream" Internet is blurred to the point of incomprehension. There are clearly far fewer people visiting websites like 4chan and Reddit than there are people visiting social networking sites. Yet, the catchphrases and image macros that they create can freely propagate into the larger websites. Information flows freely into Reddit and Tumblr, filters into other websites, and then filters into even more websites.

This was observable even in the pre-4chan era of the Internet. Even if GenMay was a comparatively tiny community by 2005, its catchphrases were still being used by people within that era's online subculture. The conclusion here is that it only takes a few people to start cultural trends. Perhaps, one day, we'll even be able to map how they propagate FROM these obscure online hubs.

Unless that's already happened, perhaps on a more rudimentary level.

Take a look at websites like BuzzFeed, Upworthy, and ThoughtCatalog - websites with unoriginal and sensationally-titled articles that spoon-feeds "content" from around the web to its audience. These websites are designed less around creating content, and more about sharing content. They pop up in your Facebook feed and send you off to whatever list or video that got lifted from Reddit, Tumblr, or similar. Ebaumsworld was once disparaged for its flagrant theft of content; nowadays, Ebaumsworld's business model is the norm.

This. Goddamn. Website.

The attention economy is being tamed. People are seeking aggregation websites of aggregation websites. Information is curated and shaped many times over the course of its online lifespan, packaged and re-packaged so that people will click through the same regurgitated crap and think that it's something new. Content is cherry-picked and dressed up so that it plays into your personal echo chamber as much as possible, all for the sake of more ad clicks.

These are websites that care more about the consumption of their content, rather than the generation of new content. Internet culture is shifting away from the old model of having many small, almost tribe-like online hubs. Reddit has made most forums redundant. Facebook and Twitter comprise the vast majority of online traffic, and are likely to grow. The variety of old Internet is giving way to behemoth websites.

Perhaps one day, we won't have a need for online hubs at all. Maybe most people will just be spending time on three or four websites, and the online landscape naturally conforms to the same restrictive model that television and radio channels follow. Only, it won't be out of technological limitation that we fall into this model - it will be out of creative and intellectual inertia.

And don't think for a moment that the people that run these websites are particularly concerned about this scenario. They don't care. To them, the emergence of these larger, monopolizing websites are a good thing. It fits some twisted techno-utopian ideal that just happens to coincide with how much they personally profit. These behemoth websites fight each other for market dominance. When you read about Amazon's push for delivery drones and Google's response of buying a military robot-maker company, it's hard not to feel like you're reading about an arms race.

These enterprises are increasingly trying to exercise governance over the Internet, and it's working. The Internet is becoming more centralized, and nobody has made any real effort to fight against this.

Perhaps it's time for some regulation from our real government.

I'm told those are famous last words.

We need to establish a functional checks-and-balances system between online business, national government, and individuals on the Internet. Libertarian rhetoric online has kept government mostly unwanted online, while the businesses have swelled to massive proportions. The massive failure of the RIAA and MPAA in trying to curb the rate of music and movie piracy further demonstrates that government would have an easier time targeting businesses instead of individuals. It is the balance between government and business that needs to be most deeply rethought in the Internet age.

Let's just spitball some ideas: Copyright laws need to be amended so as to make certain intellectual properties - like memes - public domain. Novel ways to monitor the scope of a business's online presence need to be devised, and hard caps to the amount of information output for large online enterprises need to be implemented. Regulations on machine learning algorithms need to be done so that users can opt out of targeted advertising. Limits on ads per web page could be set. Limits on content replication could be set, perhaps enforced by algorithms that scan uploaded content and cross-check against other uploaded content.

Governments, in turn, should not be able to censor content on the Internet. That said, illegal activity online should be sufficient to mark those particular content creators for arrest. Free speech should be a protected principle online. If you're concerned about movements like Anonymous, then accommodate them instead of shutting them down. Incorporate national goals in fields like education into the digital infrastructure. Everyone and their mother already concern themselves with governmental censorship online, so these talking points are far less interesting than the limitations on businesses.

Perhaps there are others we can consider as well. We already know that regulation can produce better content when the enterprise is actually willing to regulate. It might be time for some new laws.

I do not know how this latest Internet "era" will continue to unfurl. The alarmist in me is concerned that the Internet has lost something valuable that it used to have. There feels like there is more white noise, and it seems more regularly spoon-fed to us. As though the Internet were getting more centralized, less diverse, more pandering to its constituents that make money, more out to encourage consumption over production.

But then, another part of me reminds me that I was 15 years old in 2005. The things that I thought were cool were probably really dumb. So, who knows? I could be seeing the old Internet through rose-tinted glasses. I'm also certainly not qualified to talk about law, or the state of government. Sometimes it's important to remember that I'm just someone who spends too much time on the Internet.