I wrote a post earlier in the week about how the claim in The New Yorker that brainstorming "doesn't work" is an oversimplification. I gave various reasons: Most of this research is done with novices rather than skilled brainstormers, only looks at one measure (quantity), and ignores how brainstorming is done and the impact it has in real organizations. As I have been thinking about this research a bit more and of the brainstorming that Andy Hargadon and I studied at IDEO years ago, that I see at the Stanford d.school, and especially, that I've seen in recent weeks in some very skilled groups I have seen in action, something struck me:
The comparison between group and individual brainstorming that underlines this research is false, or at least irrelevant, because both happen at once when skilled practioners do it.
When a skilled facilitator calls a brainstorm, he or she usually gives the topic in advance and asks members of the group to do some individual thinking about it before the gathering; for example, I once went to a brainstorm at IDEO on how to give an itchless haircut. I dutifully went to a stylist and asked her to give me an an itchless haircut She did things like wrapped my neck really tight with the top of the smock and put a bunch of talcum powder on my neck. So I came prepared to add some ideas. The funniest part was one designer who tried to talk his barber into giving him a haircut while he hung upside down. It was a crazy idea, but the notion of using gravity to solve other design problems was not -- so having this story in the IDEO culture was useful.
In addition to the routine practice of encouraging solo idea generation before the group meets (and most relevant to the research) is that if you watch skilled teams, there is a blend of individual and collective idea generation going on most of the time DURING the brainstorming session. Typically, in a group of say 6 or 7 brainstormers, you will have 2 or 3 people talking about the idea that is in play at the moment -- one written on a post-it, written on the board, illustrated with a drawing, or a quick prototype. Meanwhile, the other 4 or 5 people are half listening, writing ideas on post-its, drawing, building something, or semi-tuning out and just thinking about how to mix their ideas with with those they are hearing and seeing around them.
There is a method called "brainwriting" where members write ideas on slips of paper, then pass their ideas to each other, and generate new ideas in response to others -- all in silence. As least one experiment shows that brainwriting enables people to develop more and apparently better ideas compared to brainstorming alone. This research is interesting in that, when you watch the best brainstorming groups, although they don't work in silence (the solo brainstorming happens before they meet in many cases ... and since they are working in ongoing projects, they have time for individual silent contemplation afterwards as well), people are constantly switching between "solo" mode to generate ideas and "social" mode to share their ideas, listen to others, and build on the ideas of others
Real groups do "brainstorming" in much messier ways than it is sliced-up in psychological experiments, but the headline here is that in practice, if you watch how the pros do it, it entails a blend of individual and group idea generation -- even during group gatherings. This insight is, I think, important because skilled brainstormers are constantly switching between "solo" and "social" mode and the best facilitators -- I think of people like Perry Klebahn and Jeremy Utley at the Stanford d.school -- constantly take steps to help brainstormers switch back and forth between these modes in the moment.
Again, I don't want to defend brainstorming too strongly because there may well be better methods for facilitating idea generation and creativity in general . As I said last time, I do believe that teaching groups how to fight well is probably more important than teaching them how to brainstorm (and a lot harder) if you want to spark creativity, a point made well in The New Yorker story. I also believe -- and can show you evidence, notably from the late Robert Zajonc -- that it is impossible for human-beings to withhold judgement about anything they encounter (despite instructions to do so during brainstorming). But I confess to be annoyed by the conclusion that "brainstorming doesn't work" because it is based on research that is largely irrelevant to how it is actually done in teams and organizations that use it routinely.