The problem with undoification
A student in my “grammar class”: came to me yesterday evening with a clever question about our morpheme test case, undoification: a) since many derivational morphemes are limited in terms of the parts of speech they can modify, and b) since -ify can only be attached to nouns and adjectives, and c) since undo doesn’t look like either a noun or an adjective, how does the example work?
She is right that undo doesn’t feel like a noun or an adjective. For example, a test for whether something is a noun is whether you can precede it with a determiner (we’ll be learning about this next week, but basically that means an article like the or a, a demosntrative like this or that, or a genitive words or phrase like Bob’s); another is whether you can modify it with an adjective. A test for adjectives, on the other hand, is whether you can put them in the comparative or superlative (e.g. big, bigger, biggest) or qualify them with an intensifing adverb (e.g. really big).
Undo doesn’t work easily with any of these, except maybe being preceded by an adjective: the undo, the big undo, the really undo worker, the most undo worker. In our meeting yesterday, I suggested that maybe the problem was that it was just a bad example in that the stem was so unusual it required a large backstory to explain.
I’ve been thinking about this since, however, and think that I can provide both an unambiguous parallel to the example to show that it is correct and explain why this particular example is so uncomfortable.
First the parallel: recallifications. Recall is definitely a noun: you hear it all the time in relation to problems with cars. But recall is also clearly derived from a verb: a recall is when an auto company recalls its cars back to the dealer. Recallifications, presumably, are the name we could give to the process of recalling cars. Since recall (the verb form) and undo (the verb form) are very similar, we can see how this parallel supports the original example.
So why does recallification seem less troublesome than undoification? It isn’t because recallification is a more reasonable word: nobody would really use that to describe a recall, since recall works just as well.
The issue has to do with do, which functions in English in a special way: it has some lexical uses where it means roughly the same as French faire (i.e. perform, carry out, execute). But it most common use in English is as a pro-verb, i.e. a function word that works in relation to verbs the same way pronouns to with nouns.
We will cover this in the next two weeks, but you can see what I mean if you think of all the different ways we use do:
- He did wonderfully after the operation
- I did what you told me to do.
- Did she really go to the college?
- “I hate Mondays.” “Oh, do you really?”
- I do not like green eggs and ham.
- I do wish you would tell them!
In most of these examples, do is a function word rather than a lexical word—i.e. it is functioning as an auxiliary or it is functioning as a verbal proform. The only example in which it has a clearly lexical meaning of “perform, executive” is example 2 (and maybe example 1). The best example of its use as a proform is in example 4 where it is substituting for hate in the tag question.
So the unease Rachel felt about undoification was not that it doesn’t work, but that do is an unusually light verb, semantically speaking. Most of the time we use it, it doesn’t really have any lexical meaning, meaning it can be hard to think what an undoification might be without an elaborate back story. If we replace undo with a semantically heavier verb-derived noun like recall, the problem disappears.
