A little more on see and do

« previous post | next post »

Following up on had did (here) and have saw (here): a note on Richard Meade Bache; an I've saw sighting from the 20's (from John V. Burke); and (from Breffni O'Rourke) an observation about different verbs DO.

First, Bache, who noted (in 1869) that

Many persons who do not say "I done" and "I seen," do say, "I have saw."

My own copy of Vulgarisms And Other Errors Of Speech has arrived, and it seems, from internal evidence (publication in Philadelphia, references to the American publication The National Intelligencer, a footnote, p. 134, saying, "The word local is here used as referring to the whole of the United States"), that this particular Bache was in fact American. I don't at the moment have information about where he lived — potentially significant, since much later reports of PSP saw associate the form especially to the South and South Midlands of the U.S.

But Bache is probably a red herring. He doesn't actually cite anyone saying or writing "I have saw", but instead is reporting instances (from newspapers and other publications, one of them English) of what is in fact faulty parallelism, involving the coordination of a clearly perfect VP with a second VP that is past:

[from a correspondent] Hold a mirror so that Planet Jupiter may be reflected in it, when two of the satellites may be seen with the naked eye. [from the writer] We have tried it, and saw satellites of Jupiter 

— and family have arrived in Washington, and took up their headquarters for the winter

I have walked four or five miles, and although much fatigued, went to a dozen places

General Hawley had … beat down two little stone walls, and came in upon the right flank of the second line.

The point is that Bache is claiming these examples contain instances of (respectively) "We have saw …", "— and family have took …", "I have went …", and "General Hawley had came …" But they don't. 

[A digression on this point.

The phenomenon in question is often known as "reduced coordination" or "conjunction reduction", though such labels prejudice the way you think about examples; they encourage you to think that coordinations of non-clausal constituents are "really" (in some sense) coordinations of clausal constituents, with some repeated material omitted or "understood". However this idea is to be interpreted — and there are several different ways out there — in none of them is have seen the Queen a constituent in "We have been to London and seen the Queen", or have saw the Queen a constituent in "We have been to London and saw the Queen". (Ok, there are some non-standard views of constituency in which such analyses are possible. But they go way beyond the views we're looking at here.)

So there's an unexpressed assumption that coordination is "really" of clauses and everything else is parasitic on that. I'm sure this idea can be traced back to the Greek and Latin grammarians and logicians, where it's a natural concomitant of a focus on truth, deductive reasoning, and so on. But this assumption leads people to posit material that is not actually in sentences and to critique examples on that basis.]

I was initially struck by Bache's claim to have come across instances of PSP saw (and other verbs) in reasonably elevated writing and speaking. As far as I know, PSP<PST has always been non-standard (though widespread in some regions by some speakers at some times). I simply wouldn't have expected the sources Bache cites to actually have used have saw and the like.

But what were they doing? Looks to me like coordination of VPs in different tense/aspect constructions: a perfect-aspect [some would say a non-deictic past tense — see CGEL, chapter 3, section 5] VP with a simple past-tense VP. Is this by definition an instance of "faulty parallelism", to be excoriated on that basis?

Absolutely not. There is no general constraint against coordinating constituents with different values for some feature of form. Singular and plural NPs, for example, are easily coordinated:

A knife and two forks are provided in each setting.

And many coordinations of VPs differing in features of form are also unproblematic:

I went to Paris three years ago and am living there still.

I have gone to Paris many times and hope to return soon.

I have gone to Paris many times and was always fascinated by the city.

Some coordinations are more awkward than others — a matter of semantics/pragmatics, not syntax, I think — and handbooks often recommend against shifting tense, but when there's a problem it's a matter of shifting point of view, not of syntax. Some of Bache's examples might strike some readers as clunky, though in context I find most of them unproblematic. For example, the have tried … saw example seems entirely natural to me, on the understanding 'We have tried it, and [when we did] we saw satellites of Jupiter'. When there's a problem, it's a matter of sequence of verb forms.

As far as I can tell, writers of standard English didn't actually use have saw in the 19th century (and haven't used it in the 20th or 21st).

So much for putative-antique have saw. On to actual occurrences of have saw. Since PSP<PST is such a natural levelling, I would expect it to have a considerable history. So it's nice that John V. Burke was able to dredge up an example from the 1920's, in the vernacular of a Ring Lardner character. I'm giving the whole quote — from Ring Lardner, Jr. The Lardners: My Family Remembered (New York, Harper Colophon Books, 1977) — that Burke sent me, since it has entertainment value beyond the verb forms:

There are a number of signs to suggest a revival of interest in the work of Ring Lardner… During his lifetime and for thirty years after his death, he was practically unknown outside the English-speaking world; this year there are collections of his stories on sale in France, Italy, West and East Germany, Rumania, Norway, Czechoslovakia, Yugoslavia and a number of other countries. How full an appreciation of his work there can be in all these new languages, I am not competent to judge, but clearly some obstacles have not been surmounted:

The rhythm at least of 'And he gave her a look that you could pour on a waffle' is lost when it becomes 'E le diede un'occhiata che avreste potuto spalmarla benissimo su una fettina di pane.'

And the reader is not getting the same sense of the character speaking the line when 'I've saw outfielders tooken sick with a dizzy spell' is converted to 'J'ai vu des joueurs de l'exterieure avoir une attaque de vertige.'

The doubly marked tooken is delightful, and has been reported before. I have even written about it in a piece on child language ("A double regularity in the acquisition of English verb morphology", Papers in Linguistics 3.3.411-8 (1970)).

But the relevant bit is PSP saw. I assume that the PSP<PST levelling has a much longer history, but at the moment I don't have evidence on the matter.

Finally, Breffni O'Rourke on uses of DO. Scholars of English recognize that there are two ranges of DO — main verb ("activity") DO, as in:

I do what I can.
I did my homework.
What I did was turn the dial.

and auxiliary DO (also known as "supportive" DO), as in:

Did you read the assignment?
I do like butter on my bread.
I did not see that coming.

The two ranges of DO differ in meaning and in syntax, and are clearly now separate lexical items (despite their common history). But in standard English they continue to share their inflectional morphology. O'Rourke notes that in non-standard varieties with PST (<PSP) done, even the inflectional morphology is differentiated: only the main verb DO has this variant for the PST, while the auxiliary has did, as in standard English:

I done my homework.
What I done was turn the dial.

*Done you read the assignment?
*I done not see that coming.


Comments are closed.