Weird expression

I’m a philosopher of language, and if you don’t know much about what philosophers of language do then you might think that means I know all about grammar and syntax and subordinate clauses and that sort of thing. Sadly it doesn’t: philosophers of language can get by without knowing much grammar at all. (You should see some of the things we say about the word ‘that’.) Nonetheless, today I’m writing about grammar. I do hope I’m not becoming one of those ill-informed windbags that Geoff Pullum so entertainingly excoriates on Language Log, but time will tell.

Anyway, the expression I’ve been wondering about is ‘try and’, as in ‘I’ll try and get home in time for tea’. I understand that some people don’t like this expression, and pretend to misunderstand it as (roughly) ‘both try and’. If it meant that then you couldn’t try and do something without succeeding, and this isn’t what people mean. They use it to mean what ‘try to’ uncontroversially means. Maybe people pretend to misunderstand it because they think it’s new and they don’t like new things. Or maybe they dislike it because it’s weird.

The OED gives six examples of ‘try and’ being used like this, between 1686 and 1883, and I saw another example on the BBC website yesterday, even though I wasn’t consciously looking for one. I’ve forgotten where it was though, so I deliberately found another example. It's in the text under this video. It’s even easier to find examples in reported speech, which I suppose suggests it’s a bit informal. The OED says it’s colloquial, so I guess they agree.

Anyway, the grammar they give for ‘try and’ is that it’s followed by a co-ordinate verb, whereas ‘try to’ is followed by an infinitive. I don’t know whether that’s right. Consider these two:
 

• ?I try and be the best.
• *I try and am the best.

To me, the second sounds totally ungrammatical if it’s meant to mean ‘I try to be the best’, and the first sounds a lot better. It seems to be the one people use, anyway. But if I’ve understood the OED’s rule correctly (I blush to confess that I’m a bit unfamiliar with the terminology and after looking it up I’m still not quite sure), they predict that the second is better. It isn’t.

Leaving ‘be’ out of it for the moment though, I think the only times ‘try and’ really sounds fine are when the second verb is the same in the infinitive form and in the form ‘try’ takes in the sentence. These all sound pretty bad to me, where ‘try and’ is meant to mean ‘try to’:

• *I am trying and find my keys.
• *I am trying and finding my keys.
• *He tries and find his keys.
• *He tries and finds his keys.
• *I tried and find my keys.
• *I tried and found my keys.

I don’t actually know whether I use ‘try and’ in my own speech, and a search of my blog seems to reveal that I don’t use it there, but I'm used to reading and hearing it so my grammaticality reactions shouldn't be too far off. I’m fairly sure that people who do use it only do so when ‘try’ is in the plain-looking form, so (except with ‘be’) it doesn’t matter whether the verb following ‘and’ is supposed to get modified or not. All the OED’s examples are like that. I guess when people are using a form with 'trying', 'tries' or 'tried', they go with ‘try to’.

This seems an odd state of affairs. Maybe what’s going on is that there are two rules, which in a lot of constructions can’t both be obeyed: use an infinitive verb and use a co-ordinate verb. If you can’t do both, it sounds wrong. Is that plausible? It doesn’t seem terribly plausible, especially when we’ve had at least 327 years to get used to it. Our language-processing machinery has had plenty of time to work out how it’s analysing the construction. But I can’t see what else might be happening.

Another problem is that this explanation doesn’t predict that ‘try and be’ sounds OK. On the other hand, while Google Ngrams has no results for either ‘he tries and be’ or ‘she tries and be’, the former gets plenty of genuine results from a standard Google search and the latter gets none. That’s pretty weird, and I’m not sure what kind of theory would predict it. But it’s how we talk. If you know of any proper linguists who know what the deal is with ‘try and’, do let me know, and if the OED really do have it wrong then maybe let them know too.