Yes! Or so says Evgeny Morozov of Foreign Policy magazine’s net.effect blog.
I don’t agree. Or, maybe more correctly, I think he misses the point. If Twitter reveals that people are acting like a dumb herd of swine (which I’m not sure is the case), then so much the better! That’s an interesting and potentially valuable thing to know!
There are three fundamentally flawed assumptions in Mr. Morozov’s argument. The first is that Twitter serves as an important source of information about current events for most users; hence, the “panic” and “misinformation” found on Twitter is “dangerous.” But is that assumption warranted? Not entirely. Just from a logical basis, it is problematic to assume that the millions of people using Twitter use it for the same purpose. There is bound to be variability. And we certainly know that there are other uses of Twitter other than information seeking, namely as a tool of personal broadcast, a tool for distributed conversation, a tool for maintaining personal situational awareness among a group of friends, etc. So, before we panic about the supposedly dangerous panic supposedly caused by Twitter, we need to know if, among all the uses of Twitter, information seeking about current events is really one of the dominant uses.
The second flawed assumption (which relies on the first), is that when people use Twitter as a source of information about current events, they 1) believe everything they read there and 2) go no further. Again, there is no evidence presented to support this position. In many cases, Twitterers provide links to legitimate sources of information. Why should we not assume that Twitter serves an important re-broadcast function for legitimate information instead of assuming a priori that it is solely a generator of panic and misinformation? The eight examples of hysterical-sounding tweets (out of millions of tweets on the subject) that Mr. Morozov provides as “evidence” hardly suffice to prove that it is one way and not the other. It is most likely that the truth lies somewhere in the middle. Ultimately, what is needed is real research, not panicky, knee-jerk pronouncements about the dangers of Twitter-induced panics.
Third, Mr. Morozov assumes that the only possible value of Twitter to the outside observer or researcher is as a prediction tool, in this case as a tool for predicting the spread of the swine flu outbreak. And in this task, he says, Twitter fails. He writes,
Thus, Unlike basic internet search Ã¢â‚¬â€œ which has been already been nicely used by Google to track emerging flu epidemics Ã¢â‚¬â€œ Twitter seems to have introduced too much noise into the process
And yet the bottom line is that tracking the frequency of Twitter mentions of swine flu as a means of predicting anything thus becomes useless…
There are a couple of problems here. First is the assumption that prediction is the main possible use. Second is that we are mainly interested in predicting the spread of the disease. But what if prediction of disease spread were not the goal? What if the main value of Twitter were as a real-time measure of public attention or concern? In fact, this is the way that Mr. Morozov is using Twitter. And he is, in effect, concerned about the public attention and concern being devoted to and expressed about the swine flu, attention and concern about which he would likely not have been as aware (if at all), nor about which he would have been able to write in such self-righteous tones, without Twitter.
But the next problem with his claims about Twitter’s supposed inability to be used as a tool for disease surveillance and prediction is that he is by no means an expert in such matters, while at the same time people who are experts in such matters believe that Twitter does, in fact, hold great possibility in this area. In fact, one of the scientists who did the groundbreaking study on using Google search trends to track emerging epidemics (something Mr. Morozov approves of), also believes that Twitter (and other social media) could be used for the same or complementary purposes.
But beyond all that, I think he misses the point where evaluation of Twitter’s use as a tool for “outside” observers and researchers is concerned. He is approaching the case of swine flu on Twitter as a “first-order observer,” seeking to evaluate whether or not tweets about swine flu constitute legitimate information or misinformation. And when he concludes 1) that the majority of tweets are misinformation and 2) that this constitutes a danger (again, no evidence provided on this point), he renders a negative judgment on the entire service. But, from a “second-order” perspective, the value in this case is that we are able to detect sooner, and possibly in greater detail and with greater accuracy, the degree of misinformation that is capturing peoples’ collective attention, causing them concern, and getting spread about (assuming, of course, that he is correct that the majority of tweets are in fact spreading panic and misinformation). Is this not valuable information for public health officials?
One might respond by saying that the threat of Twitter-induced panic is greater than the benefit one might derive from being able to monitor the panic. First, again, I would point out that he has not provided any evidence of harm caused by the supposedly Twitter-induced swine flu panic. Second, there is no evidence that this supposedly Twitter-induced panic is greater than the panic that would have occured without Twitter. There would likely be panic even without Twitter. The difference this time is that we can monitor the panic and potentially intervene.
So, the real “bottom line,” as Morozov might say, is that that the assumptions upon which his argument is based are flawed; he has not provided enough (or any) evidence to support his argument; those who are more knowledgeable than him (and from whom he indirectly seeks support) have recently published work that contradicts his position; and his own use of Twitter in this case serves as an example of its potential value for outsider observers, researchers, and public health officials, contradicting his main argument.