When is a crowd an echo chamber with no diversity of opinion? When is it a mob without direction or honor? When is wisdom too conventional?
I pondered this as I read a New York Times article about alleged gender discrimination and Wikipedia. The “encyclopedia that anyone can edit” gets 85% of its entries from men. Beyond issues of fairness, does this represent the wisdom of crowds (WOC) that is supposed to provide useful, even true, answers at Internet speeds? One of the premises of James Surowiecki’s book is that the contributors are diverse.
Wikipedia is relied on by over 40% of American adults, and it has become a prime example of the value of open approaches that reach past and exceed what the experts come up with. The impact of a possibly fatally flawed reference on discussion and decision-making is of some concern, but the wisdom of the wisdom of crowds approach also comes into question. This is especially true when the contributors are anonymous and when they may (in opposition to WOC assumptions) be colluding. I’ll raise some questions:
Are environments and cultures welcoming? On the terms of potential participants? One explanation for the gender disparity on Wikipedia is the harshness of the culture. I suspect this is the tip of the iceberg. When I have taught outside the U.S., I have frequently had to make allowances for how comfortably people join into discussion, ask questions, offer opinions and share experiences. Most approaches to WOC seem to assume everyone has equal comfort with participation, and I suspect that keeps some people with valuable knowledge and perspectives out.
Is ignorance a positive thing? Is a lack of interest good? Some of the bruited successes of WOC attract many participants of passion, deep knowledge and even prejudice. There can be value to being naïve and objective, but this can go too far when lines of investigation are derailed by misinformation, lies and attempts at humor.
How much are we excluding because of access? Here I wonder about more than the so-called “digital divide.” With more phones than people in 30 countries, physical access is becoming a non-issue. But what about literacy? Language? Learning modalities? We don’t even do a good job in lowering barriers to participation in classrooms. Do we do a good enough job to claim diversity online?
Is diversity of age, ethnic group, income, and other darlings of demographers enough, or do we need diversity on a deeper level?
Ultimately, I suspect we don’t fully understand what underlies WOC. Why it works when it does; where it will work and where it will fail; what the inherent limits are; and what necessary conditions must be met. We need a good taxonomy of situations so we can decide which will not work, which will need a mixture of approaches and which are no-brainer doable. We need simple tests of the crowd participating to determine if they have the diversity, independence and (possibly) expertise to merit our attention.
There are notable successes of WOC. Hollywood Stock Exchange is so good at predicting winners and losers that the studios pay for the results. And I still use Wikipedia as a starting point. But I am concerned that the failures of WOC approaches will ruin its reputation. I don’t expect it to live up to the hype, but I believe it has value that could easily fade if it is not critically examined soon.