The Yale Grammatical Diversity Project (which I plugged back in 2013) has an interesting webpage on the phrase “try and”:
Typically, try can be followed by three kinds of phrases: a noun phrase (1a), an infinitival verb phrase with to (1b), or a verb phrase with -ing (1c).
1) a. I’ll try the salad.
b. I’ll try to eat this horrible salad.
c. I’ll try adding vinegar to the salad, to improve the taste.
However, try can also combine with the conjunction and, followed by a bare verb form:
2) I’ll try and eat the salad.
This usage is very similar in meaning to try to, if not identical, but is deemed prescriptively incorrect (Routledge 1864:579 in D. Ross 2013a:120; Partridge 1947:338, Crews et al. 1989:656 in Brook & Tagliamonte 2016:320). In the next few sections, we will see that it has a number of interesting properties.
A sample of one of those properties:
Unlike with regular coordination, try and is available only when both try and the verb following and are uninflected, which means it must occur in its bare form. Carden & Pesetsky (1977) call this the bare form condition.
(Click through for examples.) Via Avva, who mentions other good YGDP pages, like “come with” (We’re leaving now, do you wanna come with?), “drama SO” (I’m SO not going to study tonight), needs washed (this car needs repaired), and repetition clefts (What he wants is, he wants a good job).
The ‘tries and finishes’/’tries and finds’ form is barely grammatical for me – I wouldn’t write it, and I wouldn’t intentionally say it, but it doesn’t strike me as impossible.
It’s quite difficult to untangle google hits for the form from rugby tries, trying in the legal sense, and separate attempts (‘tries and finds he can…’), but they do seem to be out there. (‘Tries and tells’ seems to be quite a common one.)
Just on a whim, I found that “try and” is suppressed by a factor of more than 3 after “didn’t” in American English. COCA results:
did n’t try to: 1183
did n’t try and: 64
try to: 148,651
try and: 23,050
I didn’t try any other negatives—and now I see what I should have done.
(The first several hits on “didn’t try and” looked like examples of this construction, not like “He didn’t try and the result was predictable.” I probably should have searched for “did n’t try and_vv0, though.)
The Yale site doesn’t discuss the allowable verb complements of “rather than”, which I think you could say things about.
“Come with” always reminds me of this song.
Lynne Murphy at Separated by a Common Language also has a lengthy blog post on try and, try to, discussing British vs. American attitudes, with plenty of references. Her opening:
The obligatory Tolkien quotation has, naturally, appeared here twice already:
In addition, from my comment to Lynne Murphy’s post,
Thanks for the Lynne Murphy post; I see that in the comment thread John Cowan wrote:
I myself say both; I have no idea what the frequencies are.
“Try and” seems a bit softer to me than “try to”.
Like if you were talking to a child,
“Try and not dribble when you drink your milk”
is not as formal and harsh as
“Try not to dribble …”
(Example of “try may not be separated from and by negation”)
“Try and…” constructions feel a bit like serial verbs, unlike “try to”.
Hah! Catnip to me, this kind of syntax question, having spent many a happy hour puzzling out just what the hell is going on with the many different types of subordinate clause in Kusaal …
This kind of construction is not only potentially grammatical in Kusaal, but often actually mandatory:
M tɛn’ɛs ye o lu teŋin.
I think that he fall down
“I think that he’s fallen down.”
but only
M nyɛ ka o lu teŋin.
I see and he fall down
“I see that he’s fallen down.”
*M nyɛ ye o lu teŋin is ungrammatical.
(“Try to/and” doesn’t work in Kusaal, though: mɔ “strive” takes a purpose clause introduced by ye “that”, which is a different construction from either of these, though the distinction from the first type is often only formally apparent in the tone of the verb, if at all.)
Y’s instinct is (unsurprisingly) correct, at least as far as Kusaal goes. This kind of subordinate ka-clause is in complementary distribution in Kusaal to subjectless subordinate clauses introduced by n, which almost all grammars of Western Oti-Volta languages call serial verb constructions. (Wrongly – but they are not wholly unlike SVCs, either.)
I wonder if there are some (soft) restrictions on the dependent verb. For example, “I’ll try and get some dinner” sounds perhaps more natural than “I’ll try and get up before you” or “I’ll try and get there before you.” Maybe?
That and the other restrictions make this feel somewhere between a fixed expression and a fully productive contruction. I wouldn’t be surprised if some Construction Grammarians have chewed on this already.
Even so, I wasn’t taught that it even exists (unlike if I was for example) and stumbled over it later.
Scandi-Congo high-five (told you Longobardic was a North Germanic language)! My dialect lacks “to” with verbs altogether*, so “I’ll try to repair that”, Standard ich werde versuchen, das zu reparieren, comes out as the cognates of ich werde probieren, dass ich das repariere. Looks like Balkan-style infinitive avoidance, but isn’t (and doesn’t touch it geographically).
* …but not “to the”; another story.
I actually was, or rather, our school grammar for advanced learners (early-mid 70s) had a note explaining that try and is a common equivalent of try to. Of course no mention of American usage, we were learning English, after all.
“I’ll try and get some dinner” sounds perhaps more natural than “I’ll try and get up before you” or “I’ll try and get there before you.” Maybe?
FWIW, those are all completely fine to me.
These sound OK to me too, but for the latter ones I might prefer “try to”.
@dm
Probieren, dass or probieren, ob? Or does it depend on how sure you are your efforts will succeed?
No, ausprobieren, ob das so ist is where ob goes. I’m not going to try in order to find out whether I’m repairing it.
There is no try and in Shakespeare. Here are the four occurrences in Ulysses:
In each case it seems there is a particular reason for the selection of and rather than to. Even apart from registral and stylistic colouring, three of the four occurrences have to try, and arguably a distracting three-component alliteration on “t” weighs against the use of to try to (compare instances in the thread above, and note the excessive clumsiness of truly try to true if it had been used in my example 1 from Ulysses). Of the twelve occurrences of try to in all of Ulysses, only one is preceded by to. In intentionally awkward “Eumaeus”:
Looks like Balkan-style infinitive avoidance
Kusaal hasn’t got an infinitive to avoid, but gerunds do turn up as verb objects to a limited extent. However, the only case that even looks as if it ought to have a “purpose” sense is after bɔɔd “want”, and that actually means, not “want to”, but “be about to”:
Bɛog la bɔɔd nier.
morning the want appearing
“Morning is about to come.”
Bɔɔd as “want to” takes a purpose clause with ye:
M bɔɔd ye m kul.
I want that I go.home
“I want to go home.”
Though if there is no subject change, you can use the not-serial-verb construction, though it’s not very common after bɔɔd:
M bɔɔdi kul. (historically, M bɔɔd n kul.)
I want.LINKER go.home
Historically, n VERB probably was a sort of infinitive. Even now, if you ask a Mooré speaker how to say “(to) go home”, they’ll say “n kuili” rather than just “kuili.”
As a synchronic analysis in Kusaal, that runs into a lot of problems, though (like the fact that n VERB can perfectly well follow an intransitive main verb.)
But there’s a lot more similarity between some common English constructions and canonical serial verb constructions than has generally been recognised. Myself, I think what CGEL calls “catenative clauses” in English are very much the same kind of animal as SVCs. SVCs are less exotic than people tend to think, once you start looking into the syntax of SAE languages more deeply.
@David Marjanovic: ich werde probieren, dass ich das repariere
How is that not Balkan-style infinitive avoidance? Is the repariere subjunctive or indicative? It’s impossible to tell.
There is no try and in Shakespeare.
There is only do and, my padawan.
There is no try and in Shakespeare.
or in baseball, i hear.
For whatever it’s worth, intuitively try and feels to me a bit more like try + gerund, with the sense not so much “attempt to do something” as “do something as an attempt (at something else)”. For example in I don’t know if he’s home; try ____ the doorbell, I think my preferences would be ringing > and ring > to ring. Likewise in cases where success is doubtful try to sounds better to me: I’m going to try to/?and jump that six-foot fence, or (to a crying toddler) Try to/?and tell daddy what the matter is, or You are linguist, yes? So listen, and try to/?and understand. But it’s very possible I’m imagining nonexistent distinctions.
TR, I’d agree. My sense is that “try ringing / try and ring the doorbell” means to try to see if someone is home, while “try to ring the doorbell” means the doorbell might not work, in which case you should knock on the door or something.
Re absence from Shakespeare, if try and > try an + elision (of e.g., “I can”), then there is a similar construction here:
Midsummer Night’s Dream, V.1:
I see a voice: now will I to the chink,
To spy an I can hear my Thisby’s face. Thisby!
@PlasticPaddy : I think here an simply means if. Look up the entry for and, an in the Crystals’ Shakespeare’s Words.
Which makes me wonder if the to try and construction is simply a survival of this Early Modern English usage of an(d) meaning if.
@Noetica : According to the Ulysses critical edition in the James Joyce Digital Archive, the example from Eumeus (that’s apparently Joyce’s spelling) was “try to” in the notesheet where he first wrote down the phrase.
I think modern spelling editions of Shakespeare as a rule use the spelling an for the meaning if, thus introducing a division that didn’t really exist in Early Modern English.
Ulr:
Many thanks for that intelligence regarding the critical edition.
And yes, it makes sense that he’d prefer “Eumeus”, matching his preference for “Dedalus” over “Daedalus”. How very Italianesque and Jesuitically medi[a]eval of him.
“try to ring the doorbell” means the doorbell might not work
or you might not be able to reach the button/rope
For example, an infinitive – probieren – is right there. Rather, what’s being avoided is zu. Suppose you’ve tried to repair it and concluded that’s impossible: Standard German suggests das ist nicht zu reparieren. Dialect? Das ist nicht zum Reparieren – which can also mean, and in Standard German can only mean, “this wasn’t intended to be repairable”.
It’s indicative, which is trivial to tell because the “present subjunctive” (Konjunktiv I) is completely extinct in the dialect. Even in the standard it’s restricted to certain registers and specialized uses.
(By far the most common of these is the use by journalists for “that’s what they said, I’m just reporting it, I’m not taking a stance on whether it’s true”. Check out my presentations from 19 years ago and 9 years ago, with further discussion including this. On rare occasions this can be extended; the entire novel Die Vermessung der Welt contains no direct speech at all, but it’s mostly dialog – most sentences are in the Konjunktiv I.)
In Kusaal
M bɔɔd ye m kul.
“I want to go home.”
kul “go home” is imperative* rather than indicative: the forms are identical in positive subordinate clauses other than content clauses, but if you make the subordinate clause negative it takes the negative imperative marker da, not the negative indicative pʋ:
M bɔɔd ye m da kulɛ.
“I want not to go home.”
Contrast
M pʋ bɔɔd ye m kulɛ.
“I don’t want to go home.”
* “Subjunctive” is not a useful category in Kusaal, because the formal differences between main-clause VPs and subordinate-clause VPs are orthogonal to mood.
For whatever it’s worth, intuitively try and feels to me a bit more like try + gerund, with the sense not so much “attempt to do something” as “do something as an attempt (at something else)”.
I think you’re right; your doorbell example is convincing.
By far the most common of these is the use by journalists for “that’s what they said, I’m just reporting it, I’m not taking a stance on whether it’s true”
And that use is less and less consequently maintained both in written and spoken media, compared to 30-40 years ago.
Sometimes “try” in these constructions has a somewhat euphemistic sense. If I say “I’ll try to call her back this afternoon,” the real uncertainty is not (to analogize to the doorbell) whether the telephone will actually work if I press the right buttons, the uncertainty is whether I will in fact get around to pushing the right buttons versus being distracted by other things to the point of never getting to it. I don’t however have any useful sense of whether “try and” conveys that sense more ore less strongly than “try to.”
@DM My dialect lacks “to” with verbs altogether*
What about a construction like Es ist schön im Winter schizufahren?
Modern Bulgarian lacks an infinitive whatsoever, and English “to” is extremely similar to way the particle “да” is used.
Polysynthetic languages (in the Mark Baker “Polysynthesis Parameter” sense) are supposed always to lack infinitives, for some arcane Chomskyite reason that I can no longer remember.
As with all such pieces of abstract filigree, I think there are in fact counterexamples. Though part of the problem is that there is no truly robust definition of “polysynthetic”: Baker’s “parameter” doesn’t really pan out, and neither does anything else.
Even if you can concoct a suitably generalised cross-linguistic definition of “word”, there are lots of quite different ways that a language can end up with often having lots of morphemes in its words.
(It’s an interesting book though. And a good example of a Chomskyite not just assuming that all languages are essentially English in disguise.)
Just reading this somewhat negative, though fair-minded and respectful, review of Baker’s book,
https://www.researchgate.net/publication/269861258_The_Polysynthesis_Parameter
(which actually does maintain that Baker analyses Mohawk as too like English syntactically) I was struck by the Mohawk example
Ak-itshénʌ érhar wa-ha-níye-‘
“My pet dog barked”
confirming yet again that All Threads Are One.
No try to, either. Both constructions were relatively new and uncommon in Shakespeare’s time — try and may even be older, or they may be about equally old. (YGDP sloppily cites “Webster’s Dictionary 1989” on this point; what they mean is Merriam-Webster’s Dictionary of English Usage, which has a thorough entry.)
Shakespeare only used try in older senses, usually meaning test, put to the test, as in try someone’s patience, strength, etc.. Even Shakespeare can’t be ahead of the curve on every single thing.
Well adverted, KTS. There is no “try to” in Shakespeare: no more than in Yoda.
Shakespeare only used try in older senses, usually meaning test, put to the test, as in try someone’s patience, strength, etc.
I wonder. Perhaps in Helena’s “What I can do can do no hurt to try” (All’s Well That Ends Well), intransitive try is reasonably construed as meaning attempt (in the most common modern sense of that verb).
Or less innocently, this from Cymbeline:
Or even this from Henry VI, Part I:
Or this from Much Ado About Nothing:
David Eddyshaw: are you implying English is polysynthetic EDIT: but it seemed like it.
Obligatorily turned around and the -zu- dropped, but the infinitive stays intact: Im Winter schifahren ist schön.
(Vaguely reminds me of Kusaal clause nominalization, actually.)
are you implying English is polysynthetic
No, French is the only polysynthetic European language.
Lithuanian, as well.
Noetica, thanks, those examples are interesting; I think all of them could also be construed as “test”, but they illustrate how the meaning broadened. And I’d overlooked a relevant sense in the OED that does predate Shakespeare, 15a. “To test one’s ability to deal with (something); to attempt to do, perform, or accomplish (an action); to venture upon, to essay.” Your examples could fit there. So it’s not really the sense of try for “attempt” that was too new for Shakespeare, it was the constructions with to/and + verb.
(Try, attempt, endeavour: all loanwords. I couldn’t think of any native equivalent; the Historical Thesaurus gives fand, which apparently didn’t survive past Middle English. Seems like an unstable semantic area.)
Our European polysynthetic languages are French, Lithuanian …
Nobody expects the European Polysynthesis!
I couldn’t think of any native equivalent
“Strive” occurred to me: but, lo and behold, that seems to be a Frankish loan via French!
On the instability: Kusaal mɔ “try, strive” has an impeccable Oti-Volta pedigree, but the original meaning of the etymon seems to have been “wrestle” in the literal physical sense.
Polysynthetic French was discussed a few years ago in Guillaume Jacques’ blog, here, with much elaboration in the comments by David M., including on the question of whether his variety of German is also one of those dirt-common polysynthetic European languages.
Looking at the Lithuanian paper (to which Betteridge’s Law applies), you could make much the same arguments for Oti-Volta languages like Nawdm or Waama: verb suffixes expressing reversal, location (both have productive verb derivational suffixes expressing “do and go away”, “come and do”), and how often the action of the verb applies, and suffixes capable of making a verb out of a noun made out of a verb by suffixation etc.
Waama puts direct and indirect object pronouns as proclitics immediately before the verb, too, so you could do the French-as-polysynthetic dodge of pretending that they are agreement flexions. And all the Oti-Volta languages compound adjectives and even demonstratives with their head nouns, of course …
But I agree with Arkadiev that the real moral is that “polysynthesis” is really a very vague and poorly defined concept.
Just the other day I was commenting on Mary Linn’s Yuchi grammar, which shows by proper analysis of the stress system that many morphemes previously regarded as affixes are actually clitic words, thereby dropping the morphemes-per-word count very significantly, which (in some sense) makes Yuchi a whole lot less “polysynthetic.” Without anything about the language having actually changed, in reality.
My own analysis of a Kusaal “compound” like bʋpielkaŋa “this white goat” is that, despite the orthography, it actually consists of three bound words, and is therefore not “one word” containing four morphemes (there’s an explict singular marker in there too) at all. You pays yer money …
There’s a reasonably famous short story by Fritz Leiber, “Try and Change the Past “.
It really feels different from “Try to Change the Past”, even though I can’t work out how. FWIW, the story is about trials that do not succeed.
Since I’ve been switching between LH and other blogs, I noticed this natural usage because I’d seen this “Try and” thread:
https://freethoughtblogs.com/singham/2025/08/16/the-summit-that-wasnt/
I think “try and” is especially suitable for examples like Leiber’s that say the attempt don’t work. See for instance Try and Stop Me, Bennett Cerf’s first book of jokes and anecdotes.
https://en.wikipedia.org/wiki/Try_and_Stop_Me_(book)
My answer was that we’re faking it quite well in verb-second word order – but verb-last order suddenly reveals that the enclitics still count as words for purposes of word order, because they suddenly move to the other side of the verb (and attach to the end of whatever word happens to be there). French doesn’t do anything like that outside the written-or-so register.
“dare and try” in the wild
Not sure Giacomo Ponzetto is a native speaker (though his English is, of course, excellent).
One more note for posterity: as an American, I would never write this, but absolutely would say it. With the caveat that I would do so if the ‘d’ wasn’t pronounced. “I am GOING TO try AND write an essay” sounds terribly mixed in register. “I’m gonna try an’ write an essay” sounds perfectly natural.
“Gonna try and love again” – Randy Meisner (recorded by the Eagles)
As a non-native speaker, maybe I’d go with “I’ll try and write an essay” if I had to use the “try and” construction.
I agree with Kevin about pronunciation: fully pronouncing the d in “try and write” feels a bit unnatural.
Speaking of the various older senses of “try”, the YGDP (citing Tottie 2012) seems to have misread its earliest example of “try and”:
That’s not the kind of “try and” we’re talking about, it’s an ordinary coordination: “try” and “discern” are synonyms there, using the oldest sense of “try”, to separate, set apart, distinguish. (This is still the meaning of trier in French today, as in trier les déchets, separate trash from recyclables.) Searching for pre-1600 examples of “try and” gets mostly coordinations of synonyms, like “try and examine” or “try and prove”.
I was wondering to myself to what extent the “come and VERB” construction is parallel to “try and VERB,” and noticed upon a moment’s introspection one significant lack of parallelism, which is that you can almost always omit the “and” in the former but not the latter with “come VERB” being equally grammatical to “come and VERB” but “try VERB” (with the infinitive) being ungrammatical at least for me.
to what extent the “come and VERB” construction is parallel to “try and VERB,”
The linked YGDP article includes discussion of what it calls “motion verb pseudocoordination”.
Thanks to mollymooly for the pointer (which revealed that I had failed to diligently scroll through the linked source all the way to its very end …). This separately seems a bit outside the usual scope of YGDP coverage, because it doesn’t seem to me that “try and” is really a regional or dialect variation versus merely a usage that occasionally attracts prescriptive ire. Are there really speakers (at least American speakers) who never utter it naturally even when their prescriptivist superego lets its guard down?
pseudocoordination
I can’t find a short snappy account of this, but Van Valin’s concept of “cosubordination” seems to have a lot of relevance:
https://rrg.caset.buffalo.edu/rrg/RVVCosubChallengesreprint.pdf
In a nutshell: with clauses, coordination versus subordination is not a neat dichotomy.
There’s been a fair bit of work along these lines, across very different languages from a typological standpoint. Canonical serial verb constructions, Papuan-style clause chaining, and English constructions like these have all been grist to the mill.
Chapter 6 of this thesis
https://www.ideals.illinois.edu/items/123248
is about English “try and.”
I got interested in this in the context of Kusaal’s not-serial-verbs, though I eventually decided to adopt purely formal criteria for coordination versus subordination while specifying just where semantic subordination does not align with those formal criteria. (The appropriate formal criteria for Kusaal are pretty different from English: ka “and” is more often subordinating than coordinating.)
When I first came across “try and” I thought the motivation is euphony (instead of “to try to”).
drasvi: “to try to” seems analogous to “да опитам да” — “да” analogous to “to” (same functional clitic) and “try” analogous to “опитам” (as in “attempt”). It almost means the same, and has the same stress pattern.
When I first came across “try and” I thought the motivation is euphony (instead of “to try to”).
I, a native speaker, always assumed the same thing.
Doesn’t work. For one thing, the initial “to,” added to the phrase to make the idea seem plausible, is not generally used (“I’ll try to/and,” “You wanna try to/and,” etc.). For another, could you vary the “to” in, say, “tend to” with “and” for euphony? No.
The thesis that David E linked above (by Daniel Ross, 2021) covers the historical development of “try and/to” thoroughly. Some interesting points:
* For a while, from the 1300s to early 1600s, “prove” could mean “try” (in the current sense), and you could “prove to do” something. The OED also has one quotation with “prove and do”.
* Ross argues that both “try and” and “try to” developed by reanalysis of previous constructions; he includes the sentence with “trye and throughly discerne the veritie from the falsehood” as an example that’s ambiguous, and could be reinterpreted as the new “try”, expressing effort and intention. Meanwhile, purposive to-infinitives (“try to (= in order to) do something”) also came to be interpreted with the new “try”.
* In the early days “try and” and “try to” were about equally common, but since the 1700s “try to” has taken a large and increasing lead (in written records, at least).
* At first, “try and” was limited to infinitives and imperatives. It wasn’t until the 1800s that it started to appear in finite clauses, but only as long as the verb had no inflection (e.g. “whilst I try and get you some breakfast”), i.e. the bare form condition.
This might be an example of a late intrusion from outside the Balkan Schprachbund going back in. I mean — Maltese, FFS?
The Chronology fits.
Like the Latin Empie and Aromaian! I just created a new crazy new idiotic theory! (the way they go, someone has already come up with it). But it was interesting, was it not :please:. I’m sure someone has already come up with it.
https://www.youtube.com/watch?v=mlm_SIw2YnI — Seriously, this style of Portuguese singing sounds is even more phonetically Bulgarian than Romanian is — I can’t even make out the differences in the phonotacticts out at first.
Sounds extremely Brazilian …
David Eddyshaw … Yes, they’re a Brazilian band?
* For a while, from the 1300s to early 1600s, “prove” could mean “try” (in the current sense), and you could “prove to do” something
So close in meaning to its etymological doublet probe and exactly like German probieren.
David Eddyshaw: Do you mean that they don’t sound peninsular Portugese, or you don’t see the similarity with Bulgarian phonotacticts in that particular kind of song? (with either Brazillian or peninsular Protuguese?)
Genuinely very interested.
EDIT: I can tell, eventually, if it’s Bulgarian or Romanian rap — it takes me more time to discern whether it’s Bulgarian or Poruguese rap, even if Bulgarian and Romanian’s basic vocabulary is much more shared. It’s the cadence.
South African dialect users say, “come with” without any following noun or pronoun. I take it to be a borrowing from Afrikaans, “Kom saam” which means “come together (with)” but I may be wrong.
It’s also used in some US regions as a calque from (I think) German.
I happened to hear some Bulgarian here in Berlin today. Just a few words – but at first I thought it was Portuguese (which I’ve heard much more often). Something about the exact sounds of /ɛ/ and /ɔ/…
Yes and yes, but there it means only “come along”; does the South African version mean “come together” as in “gather here”?
@V:
I meant that it sounded distinctly Brazilian. as opposed to Iberian.
Portuguese-of-Portugal actually does sound distinctly quasi-Slavonic to me; however, as the sole Slavonic language I can confidently identify more accurately than “eh, Slavonic of some kind” is Russian, my expertise in this area is nugatory.
@DE: It’s not only you; back in the 90s I flipped through the stations on my satellite TV receiver in Uzbekistan and happened on a station where people spoke something that sounded like Russian, but I couldn’t understand a word. After a couple of minutes some text showed up and I realized that it was a Portuguese station and the people spoke Portuguese.
That has, of course, come up here before; cf. this quote from 2013:
More in the comments, e.g. Chris:
And Athel Cornish-Bowden said something similar in 2020:
Since I’ve been reminded by the Yagoda on Fowler thread, here’s Fowler 1926: