Twitter  Thought Catalog acknowledges that giving contributors the ability to publish and tweet offensive content is risky nie.mn/1oRXz07  
Nieman Journalism Lab
Pushing to the future of journalism — A project of the Nieman Foundation at Harvard

Why hasn’t #OccupyWallStreet trended in New York?

A new SocialFlow analysis suggests that the movement’s growing popularity might actually have hurt its chances to trend.

#OccupyWallStreet, the most commonly used Twitter hashtag since the start of the 99 Percent movement, has trended in Vancouver, Portland, and San Francisco…but not in New York. #OccupyBoston has trended in numerous cities across the U.S., but never in Boston itself.

In a blog post analyzing Occupy Wall Street as a trending topic, SocialFlow‘s vice president of R&D, Gilad Lotan, explores those ironies, analyzing all the OWS-related terms that have trended on Twitter since the start of the movement, their volume of appearance in tweets, and the times and locations they’ve trended.

The core of his findings: The fact that Occupy Wall Street grew over time, steadily and consistently, actually impaired its ability to trend. Trending isn’t about volume alone; Twitter’s algorithm adapts over time, Lotan notes, based on the changing velocity of the usage of the given term in tweets. In other words, it rewards spikes over steadiness. If a topic — say, #WhatYouShouldKnowAboutMe — bursts and then fades, it’ll trend. On the other hand: “If we see a systematic rise in volume, but no clear spike, it is possible that the topic will never trend, as the algorithm takes into account historical appearances of a trend.”

Twitter, though mum on the specifics of its algorithm, confirmed that explanation. “[Trending Topics] are the most ‘breaking’ and reward discussions that are new to Twitter,” the company’s communications chief explained to BetaBeat, responding to accusations that Twitter has been purposely blocking #OccupyWallStreet from trending. “We are not blocking terms related to #OccupyWallStreet in any way, shape or form.”

There are other explanations, too: Trending Topics compete with others for user attention — and #OccupyWallStreet has been up against both #ThankYouSteve and, yep, #KimKWedding. And the lack of a single hashtag for the movement overall — “splinted conversations,” Lotan calls it — likely also impeded its ability to break through.

But what’s most interesting, to me, are the assumptions baked into the Trending Topics algorithm in the first place. On the one hand, it’s perfectly fair — in fact, it’s perfectly necessary — to define “trends” as brief ruptures of the ordinary. Spikes, you know, speak. But the algorithm’s assumption is also one that’s baked into the cultural algorithm of journalistic practice: We tend, as reporters and attention-conveners, to value newness over pretty much everything else.

Again, on the one hand, that’s absolutely appropriate — “the news,” after all — but on the other, the institutional obsession with newness often impedes journalists’ ability to address the biggest issues of the day — the economy, the environment, the effects of the digital transition on global culture — within conventional narrative frameworks. Just as #OccupyWallStreet, in Twitter’s algorithm, competes against #KimKWedding, we pit the long-term and the temporary against each other, forcing them to compete for people’s (and journalists’) attention. We accept that the slow-burn stories have to fight for space against the shocking, the spiking, the evanescent.

Which is unfortunate, since the most important topics for journalists to address are often the ones that are the opposite of “trending.”

                                   
What to read next
Peter Van Laer's Magic Scene with Self-portrait (better) via Shi-Chi Chiang
Caroline O'Donovan    Aug. 12, 2014
A new voice on social platforms helped Mother Jones beat its traffic records.
  • http://jonathanstray.com Jonathan Stray

    Here’s my favorite quote from that analysis:

    “What we’re seeing is an outcome of a purely algorithmic mechanism, with its built in biases, hence not always intuitive or logical.”

    More and more, algorithms are essential to the public sphere — there’s simply too much information to handle without them. But algorithms are not “objective,” as Google and others have sometimes claimed. They can be “fair” in the sense of treating all information sources exactly equivalently, but that just means they’re choosing what to highlight based on other criteria. There must be some criteria, otherwise there’s no way to decide what’s important.

    I suppose I draw three conclusions from this:

    1) Algorithms are editorial, and if journalists want to have a say in the construction of the public sphere they’re going to need to start designing algorithms.

    2) If algorithms increasingly mediate online discussion, how do we understand the rules? So far, important algorithms such as Google search, Facebook news feed, and Twitter trends are mostly opaque. For a discussion on the opacity of trending topics in particular, and how we more might transparently understand Twitter, see here.

    3) This might all seem like a scary transition, but I find one part of it very refreshing: algorithms are by nature concrete. Translating vague concepts like “objectivity” into code has forced us into some really healthy discussions of what such concepts actually mean.

  • http://www.niemanlab.org/ Megan Garber

    Great points. I loved that quote, too — and appreciated Lotan’s conclusion: that it might be time to think about adopting human-mediated algorithms (the cyborg approach adopted by Techmeme et al) more broadly. That raises a whole different set of problems and questions, of course…but, still, interesting to think about.

    Algorithms are certainly editorial — and I agree that journalists shouldn’t cede algorithmic curation to tech outfits that claim objectivity but propagate, inevitably, their own assumptions and interests. A lively algorithmic marketplace, in which consumers have some choice about the assumptions that determine the information they’re served, can only be a good thing. 

    But that’s the filtration side, the post-production side. And while it’s one solution to the overall problem of continued impact and relevance, journos-building-algorithms certainly isn’t the only solution. (One thing I didn’t mention in the post, because I figured it went without saying, is that its lack of widespread Twitter trending — its lack, in other words, of algorithmic endorsement — isn’t really hurting Occupy Wall Street’s ability to grow as a movement. Trending is an easy way to amplify a message or a meme; that doesn’t mean it’s crucial.) 

    What I’d love to see is for the algorithmic translation of values into code (a process that, I agree, has sparked incredibly healthy and valuable conversations) work the other way, as well: I’d love to see journalists start thinking about information culture overall as its own kind of “algorithm” — a system that renders assumptions into a process that, in turn, defines the information journalists produce, the information that feeds the mechanized algorithms. I’d love for us to start treating assumptions as just that: assumptions. Rather than as indelible truths, institutional morals, etc. I’d love to see us rigorously — analytically — examine the impact that our assumptions make on our information. And then I’d love to see us start tweaking our algorithm accordingly.

  • http://jonathanstray.com Jonathan Stray

    Absolutely agree that we need to look at “algorithms” involving people too — or “process” as it’s sometimes called. Our public sphere is  made of people and machines. Social software is designed to support social processes. So, we need to be co-designing the whole system: what the people do, in concert with what the machines do.

  • Bryan Murley

    Perhaps Twitter should have two algorithms – one for “popping up” and one for “continuing strength” or something – like a Klout score for hashtags.

  • http://www.niemanlab.org/ Megan Garber

    Co-designing. Yes. Love that.

  • http://www.niemanlab.org/ Megan Garber

    Great idea. I think we’re going to be seeing, increasingly, that kind of filtration — not just by algorithms, but of algorithms. Makes a lot of sense for Twitter. 

  • Anonymous

    Hi Megan, This Social Flow analysis is excellent. There are broader research questions around social media transparency, however. Just emailed you a link to http://theconversation.edu.au/did-twitter-censor-occupy-wall-street-3822

  • http://www.niemanlab.org/ Megan Garber

    Great, I’ll check it out. Thank you!

  • Anonymous

    Thx – be sure to check out comments from Sean @ Twitter, the first time they’ve officially responded re: #occupywallstreet

  • Anonymous

    ‘Which is unfortunate, since the most important topics for journalists to
    address are often the ones that are the opposite of “trending.”’  – my single biggest grief with mainstream media

  • Anonymous

    This is something we are debating currently with a new digital media platform about to launch….
     “the most important topics for journalists to address are often the ones that are the opposite of “trending”.  

    How important is it to be first?  The data is the story/news and drives the conversation.  But, personally, I want an opinion, point of view and perspective- not just algorithms and popularity, even if #KimKWedding is trending.

  • http://tarletongillespie.org Tarleton Gillespie

    For me, the interesting question is not whether Twitter is censoring its Trends list. The interesting question is, what do we think the Trends list is, what it represents and how it works, that we can presume to hold it accountable when we think it is “wrong?” What are these algorithms, and what do we want them to be?

    I took Lotan’s post as a starting point for a similar discussion, that might be interesting to all of you:

    Can an algorithm be wrong? Twitter Trends, the specter of censorship, and our faith in the algorithms around us 
    Oct 19, 2011

    originally posted on Culture Digitally: http://culturedigitally.org/2011/10/can-an-algorithm-be-wrong/
    reposted on Salon as “Our misplaced faith in Twitter Trends”: http://www.salon.com/2011/10/19/our_misplaced_faith_in_twitter_trends/

  • http://twitter.com/Optiaine Aine Farrell

    I really liked this analysis, but one of the conclusions is seriously flawed: study after study has shown that “hard” news gets significantly more coverage and consumers than soft news. The latest from Pew shows OWS falling right in line with that trend. http://www.people-press.org/2011/10/19/growing-attention-to-wall-street-protests/?src=prc-headline Media critics keep making these same claims about celebrities grabbing the spotlight, but it never seems to be true on closer inspection. 

  • Anonymous

    A parallel problem some of you might not have noticed, but I am sure some of this intelligent group has noticed; Utube is carefully controling how many hits a popular video can be credited with so as not to make most watched status. My favorite example is
    Bill Maher/Allen Grayson on occupy wallstreet. It is a most watch. First time I watched it they only admited to 15,000 watches and  5 days later only 25,000  watches ( sp )
    Help us make it viral