Abstract: In 2011 I joined a team of global security analysts at Lawrence Livermore National Laboratory (LLNL) to develop a systematic methodology for “information-driven” safeguards for the International Atomic Energy Agency (IAEA). This methodology would link the IAEA’s nuclear inspection schedules, within a given state, to a heterogenous body of information about the state. As the lone physicist in the group, I was quickly fashioned as the “quantitative guy”, who would produce quantitative frameworks to make information-driven safeguards “more objective”, and “less political”. But as my quantitative contributions earned the respect of my peers, I became leery of the role they might play in the political shift I saw afoot. If safeguards was to be information-driven, then whose information would “drive” the IAEA?
In this presentation, I tell the story of my experience as the “quant guy” at LLNL in order to explore a broader question: when is it useful to deploy quantitative methods of highly complex phenomena in matters of social consequence? I borrow from a distinction commonly made in the science studies literature - between disciplinary and mechanical objectivity - to argue that quantitative models serve us best when they are used not to “remove bias” or “increase rigor”, but to aid subjective judgement and facilitate communication.
About the Speaker: Chris Lawrence received his PhD in nuclear science at University of Michigan in 2014, and has published in the fields of nuclear detection and solid-state physics. His dissertation introduced novel neutron-spectroscopy techniques for the verification of warhead dismantlement. In 2011 he worked on nuclear safeguards policy issues in the Global Security Division at Lawrence Livermore National Laboratory. He is currently a Stanton Nuclear Security Fellow at CISAC, where he studies the creation and deployment of knowledge about nuclear programs and treaty compliance.