Abstract

Raymond Kelly’s widely cited _Warless Societies and the Origin of War_ (University of Michigan Press, 2000) seeks to explain the origins of two central signatures of human society: war and segmented—i.e., multilevel—societies. Both, he argues, arose with the emergence of a social-substitutability principle, a rule that establishes a collective identity among a set of individuals such that any one member becomes equivalent to, and responsible for the actions of, the others. This principle emerged during the Holocene, when population increase gave rise to the first lethal ambushes. By its nature, ambush obscures attackers’ identities. Those attempting to retaliate for the ambush were therefore obliged to target members of the ambushers’ group indiscriminately—i.e., based on a social-substitutability principle. Kelly’s proposals draw welcome attention to a widespread, deeply influential, and unsettling human behavior, the disposition to hold everyone in a group culpable for the actions of a few, a proclivity that all too often results in mass slaughter. His general argument, however, is logically and empirically deficient, and cross-cultural evidence on ambush in contact-era New Guinea undermines his anonymity-of-ambush hypothesis. What then accounts for war and multilevel society? The New Guinea evidence strongly supports a contention that social-substitutability behavior arose not from offensive military action (i.e., ambush) but from the defensive military response to ambush. These findings render the social-substitutability argument’s unconventional definition of war superfluous, undermine its chronology for the emergence of war, and underwrite an alternative scenario for the origins of multilevel, segmented society.