The Wisdom of Crowds – and Empathy

Hay Festival Downtown

Hay-on-Wye, Wales, May 2014

It was an interesting day this past Saturday.  It was WSU graduation.   Graduation for the longest time has always been held in Beasley Coliseum, the big arena on campus. And while I consistently attend the pre-ceremony line-up, I’m not much on going in and sitting in one of those tiny chairs for two hours.

As often happens in Pullman in May, it was cold and raining.  Not fun.  Usually, if the weather is good (meaning no rain, but usually cold) students line up outside, underneath the various disciplinary signs, and file in during the graduation procession, “Pomp and Circumstance” playing while the announcer declares the majors.  If it is raining, it’s a little more chaotic, but the same signs are used to organize students around the enclosed outside ring of the stadium.

This time, there were no signs. And not surprisingly, there was chaos. Because there was no directed ‘binning’ of graduates, everyone attempted to head for the main entrance.  It was what I call classic ‘crowding’ behavior, where people pressed toward the main entrance in the arena, pushing and shoving toward the one set of open doors, as well as adjacent to where students go to get their name cards, so when they show up on the big view screen above the crowd when they get their degree, their name is pronounced.  It was nuts.

I attempted to corral all the students that I knew — my goal, since I’m their capstone instructor, and the only faculty member that can consistently recite all their names — and hold them back from the rush.  It was unclear how there would be the standard line-up for the procession, but there was going to be no joy by pressing up closer to the main doors, which were also feeling an influx from the main outside doors to the Coliseum.

I found it was amazingly difficult to get the students that even knew me to hold back.  It was like there was an all-encompassing force driving those kids up where everyone else was.  Though no stampede occurred, you can see how people can easily get trampled in such events.  “This is not a wise crowd,” I thought.  Which got me thinking back to that catchphrase, which then led me to research who mainstreamed the popular phrase into the contemporary lexicon.

The answer is James Surowiecki, a writer for the New Yorker magazine, The Wisdom of Crowds.  Staring at the cover on, I think I’ve read it.  But I can’t remember!  Regardless, the Publisher’s Weekly review on the page sums up the four conditions Surowiecki lays out:

“Wise crowds” need (1) diversity of opinion; (2) independence of members from one another; (3) decentralization; and (4) a good method for aggregating opinions. The diversity brings in different information; independence keeps people from being swayed by a single opinion leader; people’s errors balance each other out; and including all opinions guarantees that the results are “smarter” than if a single expert had been in charge.

Not surprisingly, since according to Publisher’s Weekly, the book is based on behavioral economics and game theory,  there’s no mention of empathy.  Instead, we get that surface-level description of dynamics (back in Mario Kart again!) instead of a deeper understanding.  But it’s actually a pretty reasonable one.  We covered the value of diversity of opinion when we looked at Scott Page’s work.  Independence of members from one another is a little more challenging, because higher levels of empathy do indeed require differentiation of self from others.  At the same time, it’s a little tricky because this factor might also imply a lack of connected communication.  But when added to #3 — decentralization — now that’s an implied diss on the old Authoritarian v-Meme.  Not too bad.  And lastly, a good method for aggregating opinions — well, ‘good’ is a judgment word.  But we could also rephrase this to mapping to whatever knowledge structure we’re attempting to use.

That’s a little more tricky.  Good in the sense of ‘how many jelly beans in a jar’ would map down to the Authoritarian v-Meme fragmented information knowledge structure, and mixed up with the other three conditions, would imply an independent guess of said number of jelly beans. You would want as many people guessing how many jelly beans as possible, with some nod to ‘grounding’ — that people had some independent, calibrated ability to sense or measure the jelly beans in a jar optically.  For higher level knowledge structures (algorithms, heuristics, and multiple heuristics) you’re going to have to have a more complex process of building shared coherence.

These happen, of course, in all sorts of design reviews, codes and standards panels and so forth.  Though we don’t think of existence of current processes that produce all sorts of knowledge as collective intelligence exercises, they are.  And they exist at all levels.  Our world runs, quite literally on codes and standards that are created by expert staffs from all sorts of industries, for all sorts of situations.  Besides the ones for roads and buildings, which most of us are familiar with, there are codes and standards for literally every part of every operation in high-throughput manufacturing environments.  My students once did a project with one of my collaborators at an oil refinery on welding an external, strengthening patch onto a gas pipeline while gas was flowing! Needless to say, the algorithmic/Legalistic v-Meme part of the wisdom of crowds is well-covered.  Considering how many bridges collapse in the U.S., algorithmically based collective intelligence is doing pretty well.

We can keep going on up the Spiral in our Theory of Empathetic Evolution, and get guidance on Surowiecki’s four points.  Such collective, error-correcting behavior becomes emergent when one considers that many of the various institutions are already ‘good’ because of some level of mapping to those four principles.  OpenIDEO is most definitely a collective intelligence exercise that maps in the Theory of Empathetic Evolution’s Communitarian v-Meme.  And it also follows Surowiecki’s direction for diversity of voices, attempting to get community, as well as engineering representatives, and considering that it is crowd-sourced, it has decentralization and independence of voices as part of the mix.  That’s definitely going to be a solid approach for the neglected and unknown environments and the design solutions they demand, and the solution diversity contained therein. OpenIDEO ran a design collaboration recently on helping people from falling down less.  Considering the breadth of that problem space — utterly massive — it makes sense to go with the collective.

Surowiecki didn’t have access to our Theory of Empathetic Evolution.  Still, I think that his four points are valid excursions from Authority-based/Expert thinking for a large number of problems.  I’d argue, though, that understanding the underlying knowledge structure is a better bellwether of whether to poll a ton of people, vs. listen to a couple of smart guys or gals. And maybe the real key to whether you should trust experts or not, if we were to sum up in one fell swoop, is towards identification of the amount of metacognition and prediction of unknowns.  The short version — if it’s already known, your expert is your best bet.  But if there’s enough gray in the mix, go with the crowd.  Like “what’s the best pizza place in New York City?” Food critics, move aside for Yelp.

But back to that crushing crowd at graduation, indulging deeply in profound mirroring behavior, all headed toward oblivion in the nonexistent line-up for graduation.  Maybe the other lesson is to know when NOT to listen to the crowd.  And that’s pretty obvious — when it’s acting impulsively, with little or no empathy. Wisdom in crowds maps to wisdom in people and social systems in general.  And with wisdom, the more connection the better.  Empathy matters.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s