Retronyms and a Theory of Default Evolution

Ian Ayres

There was an interesting Safire Language column on retronyms, which is:

"a word or phrase created because an existing term that was once used alone
needs to be distinguished from a term referring to a new development."

Examples of retronyms include: real butter, tap water, snail mail. You can find a cool list of examples here. The concept got me to thinking about how defaults evolve over time.
Many of these retronyms start as unmodified nouns (whole milk was just milk) which didn't need a modifier because there was no alternative. When the new alternative comes into existence, we need a linguistic way of distinguishing between the two things. When the market share of the new thing is small, the default meaning of the unmodified thing almost certainly remains the same. When margarine was a new fangled invention, everyone knew that butter still meant real butter. But as the market share of the new thing grows at some point people will want to make sure that the person wants the old and not the new (I want to buy a convection oven).

And interesting empirical question is how high does the market share of the new have to be before people resort to retronyms. [There might even be a function that describes how the marketshare of the retronym vs unmodified designation varies with the marketshare of the new item.]

Most of the examples of retronyms are places where the default meaning of the unmodified word has not flipped to the new. We now have to specify "regular" or "decaf" coffee. As I look through the classic retronym list, we're often caught in a world without a default meaning for the unmodified noun. This is the kind of linguistic inefficiency that the Soviets would despise.

It would also be interesting to see if the market share of the new item because sufficiently large, does the default meaning of the unmodified word change. Will there come a time when "you've got mail" means email -- oops I guess it already does in some contexts.

Retronyms might teach us something descriptive about the evolution of legal defaults. Legal default rules are rules that govern if the parties fail to contract around them. So it strikes me that a retronym is a reiteration of an old default in situations where the speaker/contractual drafter is uncertain whether the default is still valid. As with retronyms, when two different sets of terms have substantial marketshares of use, will we find a failure to embrace either as a default. Will we see flipping of the default at the same tipping point.
Posted 8:46 PM by Ian Ayres [link]