Crowd Control

If a crowd is heading one way down a street, what percentage of it needs to change direction before the whole crowd changes direction?

The answer is surprisingly precise: 6%. If 6% of people in a crowd change direction, the whole crowd changes direction.

And this is not just true of people. If 6% of a shoal of fish changes direction, so does the whole shoal; if 6% of a flock of starlings changes direction, so does the whole flock. And so on.

Crowd behaviour works to surprisingly simple rules. The quickfire evolutions of those starlings, in which they seem to change direction as though possessed of a single will, are governed by a very rudimentary algorithm, the inputs to which are merely the behaviour of the bird in front and each of the two birds to the side. No one starling is deciding anything, and no one starling is doing more than watching the starlings to the left, right, and in front of it.

I know how they feel. One of the great dangers of any group activity is group-think, and while this does not seriously affect a group of four, there is, inevitably, a move to consensus, particularly when groups are newly formed. In terms of total body-mass, 6% of a group of four is roughly two students’ brains. Enough to move the group in a given direction.

One of the hardest skills in any language, native or otherwise, is to say to a person you do not know very well, I do not understand, or I do not agree or quite simply I think you are wrong. We are social animals, and we tend to want to converge. But when a student does get the bit between his or her teeth, and stand up in defiance of the (almost always very genial) group opinion, then things get interesting.