Sunday, February 10, 2008

Strong/weak verbs and meaning in English …

In a comment on a recent post, hh raises an excellent question:
there's an interesting directionality to the irregularity. I wonder whether, in causative/inchoative pairs where only one is irregular, it's always the inchoative?
Now, I'm not worthy to answer any questions about word forms, especially not those raised by someone who understands morphology, but out here in the blogosphere, qualifications apparently play no role in letting loose. So, despite my complete ignorance of the vast literature on this general area, where all relevant data have surely been fully analyzed, here goes …

In looking at causative/inchoative pairs, I wasn't seeing any variation in inflection that I could spot, so I tried from the other direction: Below is a list of English verbs where regular/irregular inflection is said to correlate with some difference in meaning.* I'll just list the strong forms, since the weak forms are pretty clear, unless there are two distinct forms beyond -ed.
  • bid, bade, bidden (vs. bid, bid, bid)
  • blow, blew, blown
  • cost, cost, cost
  • get, got, gotten (vs. got)
  • hang, hung, hung
  • heave hove, hove
  • knit, knit, knit
  • shine, shone, shone
  • slay, slew, slain
  • speed, sped, sped
  • stave, stove, stove
  • work, wrought, wrought (archaic, of course, as they note)
The Merriam-Webster's Dictionary of English Usage (see Jan's comment on the earlier post**) treats virtually all of these. Bid, bid, bid is said to be preferred for 'to make a bid', but "in other senses" bade is "most common". Regular staved seems restricted to 'to stave off', and they have slayed primarily in the theatrical sense of 'to be a great hit with'. I had to look up blowed to be sure that it was in the fixed phrases like 'I'll be blowed' (in addition to dialect), but even that sounds pretty marginal to me.

Beyond that, this list has some oddities, at least for my intuitions.
  • First, going back to the original example that triggered this, while shined/shone is in there, outshine is listed only as strong.
  • Second, I have no intuition at all about most of these, or didn't immediate consider them the same verb — hove is simply a nautical term or really archaic.
  • For some others they don't count, I do have a clear sense: lit works for various figurative uses where lighted just doesn't — like 'to spin the tires on a car' ("Earnhardt really lit'em up coming out of the pits"). The strong form seems like it might be better generally for causative uses.
The hanged/hung distinction is maybe the best known of these pairs. It's historically driven: English used to have a transitive/intransitive pair here, parallel to lay/lie, set/sit, that collapsed for this verb, as they have for many people in those other two examples. But MWDEU comes to the right conclusion:
The distinction between hanged and hung is not an especially useful one … . It is, however, a simple one and certainly easy to remember. Therein lies its popularity. If you make a point of observing the distinction in your writing, you will not thereby become a better writer, but you will spare yourself the annoyance of being corrected for having done something that is not wrong.
Beautiful! And right! There's the case for following lots of prescriptions. But in the end, I'm not seeing a clear correlation in this little dataset. But surely there's more data and this question has been answered somewhere that I just don't know about.

*Data from the English strong verb appendix to the Oxford Duden German Dictionary — the first place I found such a list.

**Don't miss Jan's column today on language and politics, here.


Anonymous said...

Judging by a quick look at the OED, I think sped/speeded may show at least a tendency in the direction predicted by Harley's Conjecture. Most of the examples they give with speeded are indeed causative, many in the passive voice (a typical early instance is "Eight Ships commanded by our Admiral were speeded out from Cadis," from 1678). Sped is also used frequently in the transitive as well as the intransitive sense, but I spotted only one intransitive speeded ("From the disastrous plain of Agincourt I speeded homewards"--1795). Meanwhile, there is one specific transitive sense ('to give a specified speed to (a machine)') for which all the OED's examples have speeded (as in "when an engine is speeded to run 300 revolutions per minute"--1889). This is a newer sense, of course, so the preference for the regular form could simply be due to that, rather than to the fact that it's causative.

Another verb I thought to look at was freeze. The strong forms are standard in all senses I know of, but I had a quick look at the first few examples of freezed in the Google corpus. Many of them seem to be about computers/applications "freezing" in the metaphorical sense--again, potentially reflecting a preference for regularity in a newer meaning. There were also several examples of freezed-dried, which is easy to confuse auditorially with the standard freeze-dried, and which makes sense morphosemantically to the extent that freeze-dry can be seen as a dvandva compound. (On the other hand, the end products of lyophilization are dehydrated rather than frozen, so maybe freeze-dry is really more tatpurusha-ish.) Setting aside these confounds, the remaining instances of freezed did seem to be mostly transitive; as with speeded, there seem to be lots of passive-voice examples (as in the headline Suspect doses of children's vaccine freezed, from

Mr. Verb said...

Thanks. MWDEU didn't have an entry on 'to speed' (they figure only 'sped' is standard?) and it wouldn't have occurred to me that 'to freeze' would have regular forms.

I wonder if any dialects support the Harley Conjecture more directly ... that might be an angle.

The Ridger, FCD said...

Surely any radically new use is likely to be regular, particularly if the verbal use is derived from a new nominal? That's pretty standard.

Also, did you leave the classic example of this (fly) off your list on purpose?

Mr. Verb said...

No, actually, it just wasn't in the list!

hh said...

hi all -- what a groovy thread! been out of the blogosphere for a bit so just chiming in now. these are such interesting cases!

I was thinking a bit when posting my original comment about Jonathan Bobaljik's recent result showing that change-of-state verbs based on adjectives ('degree' change-of-state verb) are always based on the comparative, irregular stem ('to better onself', not 'to good(en) oneself'). Of course, as Mr. Verb points out, the relevance of Pinker's note that denominal verbs are always regular is highly relevant too -- could support the notion that causatives are created on top of inchoative forms.

more anon! :) hh