POSIWID, deepities and scissor statements

A response to Scott Alexander, with whom I largely agree

I apologise for the long interval since my last post. As I mentioned in a comment the other day, for the last few months I’ve been working on a long-form writing project which consumes most of the mental bandwidth I have to spare when I’m not in my day job, and have been neglecting the blog in consequence. However, today I had an idea for a short post which I think I can bash out quickly.

Last week, Scott Alexander published an article called “Come On, Obviously The Purpose Of A System Is Not What It Does” followed by “Highlights From The Comments On POSIWID” today. I recommend reading both first, but if you’d rather not I will attempt to summarise Scott’s thesis under the “POSIWID” section.

If you know what POSIWID, deepities and scissor statements mean, feel free to skip down to “POSIWID is a deepity” (spoiler alert for the meat of my argument), in which I offer my own analysis of the phrase.

POSIWID

POSIWID is an acronym standing for “The purpose of a system is what it does”, coined by the management consultant Stafford Beer. As near as I understand it, Beer was hired by companies to audit their existing business processes and suggest improvements. When he pointed out that a given business process or system was producing undesirable results, the C-suite executives would sometimes defend the process by pointing to the desirable purpose the system was intended to accomplish. Beer would retort “the purpose of a system is what it does”: in other words, regardless of what purpose the system was intended to accomplish, the executives must take ownership of what the system is actually doing and what results it is actually producing.

Scott’s recent posts concerned his disagreement with how the phrase is often used in political discussions, such as by progressives who assert that the real purpose of police services is to oppress, imprison and murder black people (and stopping crime is just an incidental positive externality); or conversely, by conservatives who assert that the real purpose of non-profits designed to combat homelessness is actually to exacerbate homelessness: if homelessness were to end, they’d be out of a job! Scott argues that this framing is needlessly hostile, cynical and paranoid; instead, it is more productive to model organisations as having goals that they are trying to accomplish in earnest, but pursuing these goals sometimes incurs undesirable but unavoidable side effects (e.g. carbon emissions, medical mistakes); or the organisation is prevented from accomplishing their goals to their full extent due to factors outside of their control (e.g. budgetary limitations, competing organisations).