Thomas Mills Hinkle asks
Just now, my wife asked the following "Would you mind go checking on the laundry?" The (to me) error "go checking" made me think: what is the deal with go X constructions in English?
The context makes it clear that Thomas means X=Verb (as opposed to things like "crazy" or "fishing" or "missing"). Consulting his intuitions, he observes that:
Go X works in:
* Imperatives: "Go check the laundry".
* Combination with auxiliary verbs: "I'll go check", "You should go check", "You might go check..."
* Infinitive phrases: "I like to go check the laundry".
Go X sounds slightly dodgy with:
* "do": "Did you go check the laundry"... "I did go check the laundry"
Go X is unacceptable with:
* Present: *"He go checks", *"He goes check"
* Progressives: *"He's go checking", *"He's going check"
* Gerund phrases: *"I wouldn't mind go checking the laundry", *"I wouldn't mind going check the laundry"
My intuitions are pretty much the same, except that "Did you go check the laundry?" seems fine to me.
Go V-ing works when go is an infinitive or imperative and V-ing is the complement of go, as in the common phrase "don't go looking for trouble". But as Thomas observes, go V as a whole doesn't have any gerund-participle form, and attempts to use go V-ing (or going V, or going V-ing) for this purpose don't work.
Except that for some people, this is apparently not true.
Looking on the web, I find a certain amount of stuff like this:
"He never came out of his room but breakfast should be ready soon.Would you mind go getting him?"Kurmam asked politly.
“That’s my hope,” Barbara replied. “Dinah, would you mind go getting the wheel chair for me. I’m getting tired.”
"Would you mind go getting the water? I need to show Mom this, not just tell her."
"And you said that you wanted a bookbag with the Gryffindor logo? I'm pretty sure we have it over there--" she pointed to a corner in the shop "--would you mind go getting it please?"
"Oh, well good job, Nate! Hey, would you mind go getting your sister?"Sean: Hey, would you mind go checking the microphones one more time?
because im such a lame loser no one seems to be commenting my icons i posted, and you know, i just lvoe comments, so would you mind go checking them out?I really don't mind go looking, if you want to.
At least some of these seem to be genuine instances of constructions that some native English speakers find OK, not typos or other mistakes. (There are a smaller number of examples like "I wouldn't mind going get coffee beforehand", but these seem most plausibly to come from leaving out "to" by mistake.)
And it's also apparently not true for everyone that go V is impossible with inflected forms of V. At least, there are lots of examples of "went V" out there. Some of them may be typos for "went to V", but this set of selections from one sequence of Xanga entries shows that there are apparently some people for whom went V is routine:
wE tAlKeD tO jOe bLoWs aNd dAlE(hE wAs bEiN aN AsS)aFtEr tHaT wE wEnT eAt PiZzA aNd tAlKeD tiL liKe 2:15
[...]
wE wOkE uP aRoUnD 9:30 aNd aTe bReAkfAsT tHeN wEnT sWiM.
[...]
I saw Dale had called so i called him abck and talked to him for a while.then we ate n watched t.v. then went swim until 5:00.
[...]
I woke up at 7:15.Left my house at 8:00.Me and my sister Aimee left and went get Devon V.
[...]
N i called Melanie and went meet her in Moreauville.
[...]
I woke up at 9 n waited for his call.At aroung 11 o'clock I went get my cell phone n saw that i had a missed call from a # so i called it back....
It looks to me like there's a certain amount of microvariation in the grammar of go V -- this is the kind of stuff that emerges when we can search the informal writing of millions of people.
The basic go V construction has been around in English for a long time. The OED explains in its entry for go that
32. Instead of, or in addition to, the place of destination, the purpose or motive of going is often indicated. This may be expressed in various ways: a. by the simple inf. to go get : to go and get; to reach; to work hard or ambitiously (cf. GO-GETTER). Now colloq. and U.S.
go look! used to convey a contemptuous refusal to answer a question (obs. exc. dial.; common in Derbyshire).
and gives citations all the way back to Beowulf:
Beowulf (Z) 1232 Eode þa to setle. 1375 BARBOUR Bruce I. 433 Ga purches land quhar euir he may. c1386 CHAUCER Shipman's T. 223 Lat vs heere a messe and go we dyne. c1475 Rauf Coilȝear 157 Ga tak him be the hand. 1542-5 BRINKLOW Lament. (1874) 111 That I shulde go pour out my vyces in the eare of an vnlearned buzarde. 1591 SPENSER Teares Muses 398 Now thou maist go pack. 1602 Narcissus (1893) 87 Come, daunce vs a morrice, or els goe sell fishe. a1625 FLETCHER Mad Lover II. i, There's the old signe of Memnon: where the soule is You may go look. 1668 HOWE Bless. Righteous (1825) 199 We mighte as well go preach to devils. 1724 DE FOE Mem. Cavalier (1840) 71, I bid him go take care of his..things. 1795 Ann. Agric. XXIII. 315 Nor does the drilled corn..go lie (as the farmer calls it) so readily as the broad-cast. 1813 JANE AUSTEN Lett. (1884) II. 216 Your Streatham and my Bookham may go hang
Thomas (the guy who wrote in with the question) suggested a theory about the origins of the construction:
"Go X" seems to be a "short form" for "Go and X", which is what I would say for other forms, such as "He's going and checking the laundry" and "I went and checked laundry."
The "full" form also strikes me as correct where the shortened form would work, as in "I would go and check the laundry, but this blog is really interesting."
So, my question is, what is the deal with these "go" forms? What's the role of "go"? Are there other verbs that get clumped together with other modals/main verbs in bare forms (in infinitives, with modals) but then get re-parsed as coordinate phrases (X and Y) in conjugated forms? (or is that what's going on?)
Since go V has been used in English for such a long time, if it derived historically from go and V, it happened a long time ago. The OED gives go and V a separate sub-entry, which also draws attention to its connotation, which CGEL (p. 1303) calls an "emotive ... overlay of disapproval, annoyance, surprise, or the like":
c. by and with a co-ordinated verb. In the modern colloquial use of this combination the force of go is very much weakened or disappears altogether. In the positive imperative go is often nearly redundant (cf. L. i nunc, et...); otherwise, to go and (do something) = ‘to be so foolish, unreasonable, or unlucky as to ----’. So in the vulgar phrase (I have, he has, etc.) been and gone and (done so and so).
The citations go back a thousand years, but I'll select one that is s fairly accessible to modern readers, and also clearly tinged with disapproval:
1558 SIR T. GRESHAM in H. H. Gibbs Colloquy on Currency App. 6 Againste all wisdome the seyd bishoppe went and vallewid the French crowne at vjs. ivd. 1
CGEL groups go and V with with try and V, be sure and V, etc. Here's the discussion of try and V:
This idiomatic construction is syntactically restricted so that and must immediately follow the lexical base try; this means that there can be no inflectional suffix and no adjuncts. She always tries and does her best and We try hard and do our best can only be ordinary coordinations. There are two forms that consist simply of the lexical base: the plain form ... and the plain present tense ... But the verb following and is always a plain form, as is evident when we test with be: We always try and be/*are helpful. In spite of the and, therefore, this construction is subordinative, not coordinative; and introduces a non-finite complement of try. Be sure works in the same way as try, except that the lexical base of be is only the plain form, so this time there is no plain present tense ...: We are always sure and do our best is not possible as an example of this construction (and unlikely as an ordinary coordination).
Other instances of the same general type include sit and V, stand there and V, be an angel and V -- but each case has its own peculiarities, as you can discover for yourself. And the existence of a parallel form lacking the and is peculiar to go; so I don't think that the hypothesis of and-deletion gets us very far in explaining the distribution of contemporary forms.
Still, there are some significant affinities. Checking off the characteristics mentioned by CGEL in connection with the V and V pattern: For most of us, the go of go V must be the lexical base go (whether as the plain form, as in Let's go get it, or the plain present tense, as in I always go check the water level at noon), and not an inflected form, as in *We went get it, or *I'm going get it immediately. And for most of us, interpolated adjuncts are no good: *Go right now do your homework.
Meanwhile, I hope that Thomas went and checked the laundry.
[Update -- Geoff Pullum writes:
With some reluctance, I changed "can only be an ordinatary coordination" to "can only be an ordinary coordination" because it was in a quote from CGEL. Clearly the words "subordinative" and "coordinative" a couple of lines earlier had apparentatly inductated in you a kind of maniacatal predisposatition to insert "-at-" in lexatemes that etymatologicatally providated no reasonatable excuse for such insertation. But in making my correctation I have destroyed a nice piece of genuinely attestated evidence for a sort of speech error in writating.
I agree that the morphophonemics of typing errors is a unjustly neglected research area. And I've occasionally considered registering "Word's Worst Proofreader" as a service mark with the USPTO.]
Posted by Mark Liberman at October 16, 2007 07:50 AM