Despite the applicability of assemblage theory to extreme events, the relational ontology that assemblage thinkers employ makes it hard to ground the potential of artefacts to undergo substantial change. To better understand how artefacts can be unexpectedly destroyed, and thereby catch managers by surprise, this article draws on Graham Harman’s object-oriented ontology. This approach is used to explain how artefacts, as concrete objects, have the capacity both to cause and to exacerbate calamities. By contrast, assemblage theory is shown to provide insight into post-disaster organizing dilemmas. An analysis of artefact-based surprise at a disaster-prone New Zealand port illustrates the complementarity of the two approaches. Minimizing surprise was a major impetus behind attempts by port managers to make sizeable artefacts more disaster resilient. Yet there are limits to this process that stem from a gap revealed by the philosophical distinction between an artefact’s design plan and its maximum plan. Object-based surprise at the port also sparked tension between the competing imperatives of centralization and decentralization in post-disaster recovery. How this problem played out at the grassroots level is best explained, however, in assemblage-theoretic terms.