## On commensurability You might think that there's a common unit of exchange accross all kinds of value. That's to say, for any two state of affairs that are valuable, there's a right answer as to whether they are more, less or equally valuable. [[=Roger Crisp]] invites us to consider two lives: 1. The life of the composer, Joseph Haydn: rich, varied, mostly happy, dies at 77. 2. The life of an oyster: simple, always pleasant, but uniform, like a warm bath, dies at 20. Presumably, you'd take life (1) to be of greater value. Now: let's say we increase the lifespan of the oyster to a million years, or a billion years, or even longer. Is there some point at which you would regard (2) as more valuable than (1)? A straightforward proponent of the commensurability thesis is meant to say "yes" [^1] [^2]. Economists like to say that, in practice, we must act as if values are commensurable. We are constantly deciding between options: making tradeoffs between various goods—should I invest my time, money or other resource in X or Y? If you think that all values are commensurable, and you want to maximise value, you're likely to get very interested in efficiency. Of all the ways to produce valuable states of affairs, you'll want to know: which is the most efficient? If, by contrast, you're a maximiser who thinks that values are not commensurable, you're going to have to keep track of several variables at once (one for each kind of value), and decide whether to maximise the average, maximin, maximax, maxipok etc. When faced with tradeoffs between two valuable states of affairs, you'll have to make decisions using some procedure other than asking "which is most valuable?". #todo Does ordinal vs cardinal make a difference here? #todo Is maximising the problem here? Does the satisficer have an easier time? I once heard someone use commensurability to argue that value pluralism reduces to value monism in practice. The argument went roughly like this: 1. Of the many things you value, presumably one is cheapest to realise. 2. So to maximise total value, we should just maximise whichever is the cheapest. I don't think this works. It is not obviously unreasonabl to define a value function that requires diversity and/or particular kinds of balance between values. ## On incommensurability Several notable philosophers have denied commensurability. [[=Roger Crisp]] claims that John Stuart Mill was among them: > Mill is saying there are some pleasures such that a finite amount of them is better than an infinite amount of some other pleasures. E.g. Socrates spends a few years doing philosophy. That was so good for Socrates that any number of years as a contented pig would not be as good. [^1]: Echos of Parfit's Repugnant Conclusion here. [^2]: I'm not sure what to make of our intuitions about a case like this. As a matter of psychology, humans quickly become habituated to their situation, so they need a fairly varied set of experiences to generate pleasure. There's also the matter of physiological satiation. I suspect this kind of thing significantly explains our tendency to have pluralist intuitions and/or to value diversity of experience. --- ## Scratchpad Inbox: - https://plato.stanford.edu/entries/value-incommensurable/ - Parfit - Cowen: What Do We Learn from the Repugnant Conclusion? - Population Ethics ppl on commensurability - Broome? - Williamson etc on vagueness? Where do Bentham and Sidgwick come down on this? Agnes Callard is another notable proponent of incommensurability https://www.cato-unbound.org/2019/01/30/agnes-callard/human-lives-have-intrinsic-worth What about the theory / practice distinction? Is this just an epistemic issue? Cowen principle of roughness Small improvement argument: two things that are hard to compare, imagine they are equally valuable. Then make small improvement to one. Is the choice now easy? SSC commensurable ### [[=Ruth Chang]] Incommensurability means 'lacking a common unit of measurement', while incomparability means 'cannot be compared'. Incommensurability does not imply incomparability. You don't have to have a cardinal measure, they might still be ordinally comparable (i.e. one is better than the other). > "There may be no unit that measures the value of writing the Great American Novel and eating a slice of pizza, but the former could be better."