Here we have a comment by network member Ben Goldsworthy on the paper ‘The Impotence of Conventional Arms Control: Why Do International Regimes Fail When They Are Most Needed?’ by Lionel P. Fatton. It’s available here.
Here’s what Ben thought earlier in the year when he wrote some updates for the website that I’ve been putting up. If you’ve anything to add, feel free to pop it in the comment box at the bottom.
This paper argues that arms control regimes are fair-weather friends, effective only during pre-existing periods of mutual détente and collapsing as soon as tensions raise and military thinking begins to push out diplomacy in something of a vicious cycle. The authors explain this theoretically through the ‘phenomenon of contextual adaptation’ and practically through the examples of early 20th-century Japan and 21st-century Russia, and their experiences with the Washington System and the Treaty on CFE, respectively.
A couple more examples would have been appreciated in order to better support the case, as well as perhaps some where things were pulled back from the brink (if they exist) in order to facilitate an examination into why they turned out differently, but nonetheless the authors make a compelling case that rather undermines the utility of arms control instruments. With the US government currenly shut down (again) due to partisan intrasigence on both sides (again), I’ve been wondering if perhaps some sort of arms control agreement might be necessary in order to enshrine the depths to which neither party can sink to win political points at the expense of the nation. This paper suggests that idea is unlikely to go anywhere whilst both parties are at odds.
This phenomenon of contexual adaptation also suggests a risk of introducing autonomous machines into the conflict decision-making process. Just as the military’s sway in government increases as a result of their expertise being increasingly sought out in response to a perceived threat, an autonomous system (say, one made by the military) could be just as susceptible to seeing every problem as nail-like through its hammer-like lens. Without mandating that all autonomous weapons systems be programmed by Quakers, I’m not sure how we get around this, as any attempt to write a set of global standards to which all AWSes comply will be just as prone to adaptation.
Ben Goldsworthy, Lancaster University