Matt Goldenberg comments on Are the consequences of groups usually highly contingent on their details?