Welcome to Yet Another Blog Post About Politics, the MidEast, and The Internet, Part 2. I venture forth once more into this already oversaturated conversation for two reasons: one, I said I would, and two, this post is going to do something most others don’t, if you can believe it. Rather than making or debunking big claims about the net’s role or non-role in revolutions, I want to lay out a few evidential standards to start a conversation about how we might be able to test some of the major claims that’ve been making the rounds.
In the pre-internet era, doing this task justice would at the very least have required a treatment of scholarly-article length, or maybe even a short book. Attempting it in a blog post might therefore seem a bit foolhardy, but the blog medium offers a key advantage over the scholarly article and book, one which also compensates for my lack of comprehensive knowledge in this area: my very intelligent readers. My last post was fortunate enough to elicit some very high-profile praise, which subsequently attracted a level of global attention most academics seldom if ever enjoy. In the interests of not wasting this rare opportunity, I want to fully embrace the spirit of “publish, then filter” and nip consultation anxiety in the bud at the same time by asking everyone with relevant information on these issues to add their opinions, questions, and especially additional evidence in comments.
Those who read part 1 of this two-part series (thanks, by the way!) will be familiar with my four-category typology of claims about the net and revolutionary politics. As part 2, this blog post has two goals, the third and fourth of an analytical to-do list I introduced in part 1 (I really should have thought out the numbering better). In that post, I (1) introduced the typology and (2) teased out some of the key differences and similarities between each type; in this one, I plan to (3) articulate the kinds of evidence ideally required to substantiate each claim and (4) to determine how well the existing evidence supports each claim. This fourth goal, represented in the “evidence we have” columns, is what will need the most help—I’ll draw on everything I know, but since you almost certainly know things I don’t, you can help fill in the blanks and correct me wherever I misinterpret something.
And now, a few tables:
The net as dissident’s advantage (DA)
| Criteria req’d to test |
The evidence we have | The evidence we need |
| Broad efficacy: Digital tools must substantively contribute to multiple revolutionary ends. | Quotes and trace data are suggestive, but suffer from sample bias—how do we know these folks are representative? A systematic survey of participants asking about the significance of social media would eliminate this problem. | |
| Superiority: They must be significantly superior to non-digital substitutes such that their absence would place dissidents at a noticeable disadvantage. | Stats showing the degree of control the state exercises over traditional media. For example, the Mubarak regime directly controlled or licensed most print, TV, and radio outlets. | Extensive state control of traditional media plus trace data from social media are fairly convincing here, because they show that participants are using the latter but not the former. The role of small-scale media like flyers and of f2f conversation could be explored in a survey that asked dissidents to compare the roles of social media to other forms of communication. |
| Robustness against subversion: They must be significantly more useful for revolutionary than counter-revolutionary purposes. | News reports quoting participants who claim that authorities are IDing, tracking, and retaliating against them using information gleaned from social media (in addition to the above) | Assuming we can prove internet surveillance is happening on a wide scale (through official documentation of protocols, insider testimony, or convergent pieces of circumstantial evidence), we would then need to show that such actions are ineffective in stifling revolutionary activity. This appears to have been the case in Tunisia and Egypt, but China may be a different story. |
| Context independence: All of the above criteria must hold true in most cases most of the time. | News reports/anecdotes/trace data from recent revolutionary events; more rigorous empirical studies of older events | Ideally, we would run all of the above analyses for every revolution since the turn of the century. The more cases that can be shown not to exhibit at least one of the above criteria, the weaker the DA claim becomes. At present, we simply don’t have enough info to render a conclusive judgment. |
The net as public sphere platform (PS)
| Criteria req’d to test |
The evidence we have | The evidence we need |
Scale: Networked publics—ongoing, distributed digital communication among citizens—must engage some significant proportion of the population.
|
|
The key here is to discover the proportions of people who use the internet to communicate with others, as opposed to solely seeking out static information or purchasing goods and services. This is why internet penetration rates don’t cut it—PS is fundamentally about social uses of digital networks. Thus, we need Pew-quality usage data for every nation on the planet. We could use this to explore the required conditions (of size, depth of engagement, required topics) under which a country’s networked publics can successfully coordinate collective action. |
| Correlation: There must be a strong, non-spurious relationship between the strength (i.e. size plus activity) of a country’s networked public and the size and frequency of incidents of collective action aimed at political change. | None that I know of, but if you know differently, please enlighten me. Phil Howard apparently presents data of this type in his recent book The Digital Origins of Dictatorship and Democracy. (h/t to Clay Shirky) | Causation is probably impossible to establish definitively here. But we might be able to condense networked public strength down to a single continuous variable for insertion into an appropriate statistical model along with other possible influences (unemployment, per capita income, literacy rates, size of middle class, etc.) to get an idea of its relevance as a factor. |
The net as citizen journalism platform (CJ)
| Criteria req’d to test |
The evidence we have | The evidence we need |
| Scale: Citizen journalism must actually be occurring on some consequential scale during the revolution in question. | A wealth of trace data from Twitter, Facebook, Youtube, etc., but little has been thoroughly analyzed due to the recency of the events | This data would need to be analyzed for evidence of citizen journalism (using qualitative and quantitative methods) as well as other revolutionary purposes. |
| Efficacy: Citizen journalism must have some measurable impact on the outcome of the revolution. | Anecdotes from news articles, Twitter, etc. | Surveys could help here—if the majority of a representative sample claimed to have learned important facts from citizen-generated news, that would constitute strong evidence. The role of CJ should also be compared with that of traditional media, and this will likely differ from nation to nation due to differences in both the strength of the native press and foreign journalists’ access to protests sites. |
The net as revolutionary nicety (RN)
| Criteria req’d to test |
The evidence we have | The evidence we need |
| Low relevance: It must be shown that under most circumstances, uses of online tools do not significantly affect any consequential aspect of revolutionary politics. | The best way to substantiate this claim would probably be to start with a comprehensive list of revolutionary activities (this one and this one are good places to start) and then show that the internet only contributed to the more marginal activities. A key example is external attention, which is sometimes cited as the most significant function of online tools in revolutionary contexts. Due to its overly broad scope, the case for RN weakens as the cases for the other three types strengthen. Further, its validity is highly contingent upon the ongoing development and availability of online tools with major revolutionary applications. |
Theorizing the net and revolutions
I would like to close my little two-parter with two related observations about what people have been saying about the role of the net in revolutionary politics. First, one of the most obvious conclusions to draw from this exercise is that the bigger the claim, the bigger the evidential burden. People who wax optimistic about the net as dissident’s advantage find themselves in the position of having to show that it does in fact help more than hurt dissidents most of the time. In light of major differences between countries in variables such as net penetration, government use of net surveillance technologies, literacy rates, and military conscience, this is a difficult order to fill, and the passage of time will only make it more so. The same is true for the revolutionary-nicety crowd, who have the equally unenviable task of demonstrating the opposite. Big claims draw lots of attention, and that’s attractive to folks who are paid based on the appeal of their ideas, but I would argue based on the available data that technology and politics as a field is not overly amenable to broad theoretical generalization.
Now consider the much smaller evidential burdens of the middle two claims. Each is tractable enough to spur an active research agenda that can lead to interesting and valuable conclusions about net-enhanced revolutions. Of course, they won’t lead to eyeball-grabbing headlines like “Is X the Next Twitter Revolution?” or “Politics, Not Technology, Drives Revolutionary Change,” but they will leave readers with a better impression of how the internet is being used in various revolutionary contexts. (I thought of both of these headlines in less than a minute; small wonder that overworked newsrooms love these frames.) Ultimately, unless and until the empirical reality changes drastically, we must concede that the relationships between digital networks and politics are far too numerous and contradictory to summarize neatly in a headline-friendly phrase.
This point dovetails nicely with the second of my two conclusions. All of the claims I discuss above, along with many of the others you’ve likely been reading, can be rephrased as “the internet does/does not X” or “the internet’s role is X” in political revolutions. In the rush to make bold statements about the relationships between digital networks and politics, the roles of non-technological factors are too often either marginalized or ignored entirely. We would do well to bear in mind that not only is tech not a monocause, its influence is also highly context-dependent. So even though the notions of the public sphere and citizen journalism platforms represent improvements over their broader alternatives, their basic forms nevertheless decontextualize technology unduly. I should emphasize that I don’t see this as a fatal flaw—it’s quite simple to shift from “does the net do X?” to “under what circumstances does the net do or facilitate X?” Basic observations about technological properties are important, but they constitute only the beginning of a proper analysis of the role of any technology in any enterprise. In isolation, such observations can convey illusions of technology as an autonomous force, which I think most of us agree it’s not. Thus, cataloging the general social tendencies of the internet does not relieve us of the difficult task of empirically investigating how technologies are used to specific ends. Like politics, technology has a stubborn way of surprising even the most seasoned of experts. Armchair postulation is certainly fun and sometimes even profitable, but there’s been more than enough of that lately: let’s get to work.
As I mentioned earlier, please let me know where I’ve missed important considerations in comments. I will edit the post and give proper credit to the commenter in context.
UPDATE 3/03/11: A few lingering questions emerged in comments—
- Jeff asks, if I may paraphrase: How do digital technologies affect the speed of revolutions? This has actually been one of the most common claims about online politics since the late 90s: that it primarily helps the already politically invested complete their tasks faster. Few or no qualitative changes in the practice of politics are expected in this perspective. Empirically testing the speed of revolutions in high vs. low-ICT nations shouldn’t be too difficult, and as cases proliferate the answer should become clearer. Whether ICTs are enabling transformative political phenomena, on the other hand, depends on one’s personal threshold for transformativity.
- Jay Rosen helpfully distinguishes between headlines and substantive claims, noting that the former sometimes make promises authors never fulfill. I agree, which is part of why I based my claims in part 1 on pull quotes from the articles themselves rather than on headlines. But for this very reason I see breathless, utopian-sounding headlines as a problem. Of course anyone wishing to rebut an argument owes the target the courtesy of actually reading the full text of the argument itself, but I worry about the cumulative effect of lots of headlines that frame the issue in terms of Twitter Topples Dictators, whether pro or con. What impression do you think such a focus conveys to the casual reader who doesn’t follow this topic as closely as we do? Even if they manage to avoid the impression that Twitter Topples Dictators, they may come to believe that that’s what the debate is about, when it isn’t (or at least it shouldn’t be). So the onus is on us to find ways to frame analyses of digital-tech-and-democracy issues that don’t collapse the story into such oversimplified and misleading terms. In fact, I would suggest avoiding causal frames altogether, if for no other reason than that we simply don’t have enough evidence at present to say much about techno-political causality. One suggestion is simply to talk about how people seem to be using specific digital tools in revolutionary contexts, rather than portraying said tools as magical anti-dictator weapons.
For Scale in PS, the numbers of networked citizens is not sufficient – one also needs to know their geographic distribution, and their connectiveness with non-networked citizens.
Two scenarios:
1) 10 people on Twitter. All in the same city, all in the same college dorm, good friends with each other. No communication with other people. No Twitterati in other cities.
2) 10 people on Twitter. One each in 10 different cities. They communicate with each other via social networks continuously. Each is also a center of the local community of thousands of non-networked people using offline methods of communication.
In scenario 1, there are 10 buddies dreaming of revolution. In scenario 2, there are thousands of people in ten cities organizing revolution. In both, there are only 10 people on Twitter. Yet, the outcome is likely to be very different.
Thus, the ecology of the networkers: their spatial and temporal distribution, and their effectiveness in informing not just each other but many non-networked citizens, are important data one needs for this exercise.
One element I’ve been surprised to see not addressed are claims about technology and the speed of revolution: Egypt in 18 days versus Iran in one year. Or is this claim already commonly considered beyond reproach?
Thanks for doing this.
Additional thing to consider: the difference between a claim and a headline. In the research I did for this this post I found again and again that critics of cyber-utopianism were reacting to headlines rather than actual truth claims. Anyone who knows journalism well knows that frequently headlines insinuate claims the articles below them do not make. This is because the purpose of the headline is to arrest attention, to suggest something bold, new and interesting has happened, and of course, to get you to click.
But critics see headlines and the headlines become proof of the claim being staked, when in fact there is literally nothing there– that is, no truth claim that an author is prepared to defend. Often the impression left is very misleading. To get an idea of what I mean see this Mother Jones piece, “The Tunisia Twitter Revolution That Wasn’t,” an interview with cyber-skeptic Evgeny Morozov. The following is heard:
“Mother Jones: Andrew Sullivan and others have suggested the recent events in Tunisia were a ‘Twitter revolution.’ Even if this is an exaggeration, what’s wrong with emphasizing the role of social media played as an organizing tool?…”
The link goes to this post at Andrew Sullivan’s Daily Dish, where we find the headline: Could Tunisia Be The Next Twitter Revolution? Notice how the claim is already weakening, from “Sullivan and others have suggested” to the much more tentative, “Could this be…?” But then when we read the post, we find it is literally nothing but a reference to another post. “Ethan Zuckerman poses the question…” Sullivan hasn’t done anything but excerpt another writer. Clicking on that link takes us to this post, in which we read:
“Was social media the catalyst that helped frustration turn into protest, or helped protest spread from one corner of the country to another? It’s the kind of question that keeps scholars busy for years, as my colleague Henry Farrell wisely noted in a reaction to Malcolm Gladwell’s dismissal of the power of social media for protest. In the case of Tunisia, we need to understand whether information about the protests in Sidi Bouzid helped convince other Tunisians to take to the streets, and to understand how that information reached them – I’m far from ready to declare this a victory for social media, but I’m looking forward to studying it and understanding it better.”
In other words, Zuckerman’s not ready to call it anything yet, although he’s willing to raise the question of a revolution in which social media is catalyst. I found this pattern again and again. Either no attempt to document who’s calling it a Twitter revolution or this reliance on headlines as proof of a truth claim that melts away when you actually look for its author.
I think you should factor this into your analysis.
For some reason (which I do not fully understand) demand for observers who are claiming that social media gets the credit for recent uprisings is much greater than the supply of people actually making a drastic claim like that. Writers who wish to have the “cyber skeptic” title have a variety of strategies for coping with this imbalance. Most common: they simply cite no one at all. (I wrote about that in my post.) Second most common: use a headline as a truth claim. Third: cite an author who is raising the question, “how much influence does social media have?” and turn it into a much stronger statement, hoping we won’t check.
… And as I say that, Andrew Sullivan just wrote about this very thing.
What if Lady Gaga tried really hard to tweet Gaddafi’s head off?
David, for the Correlation column in the Public Sphere argument, I’d add Phil Howard’s _Digital Origins of Dictatorship and Democracy_, which does indeed conclude that ICT spread correlates strongly with increases in the appearance or entrenchment of democracy in the region (except, alas, in the nations that have significant oil wealth…)