Alternative futures

« previous post | next post »

In yesterday's post on "what's will?", I rashly asserted that "the commonest way to express a future-time meaning is indeed to use the auxiliary verb will". This provoked immediate questions and counter-claims. So I promised to devote a Breakfast Experiment™ to quantifying the choice among alternative verb forms used to express future time in English.

Let's start with the questions and counter-claims.  Joe asked

Is that true? "Going to/gonna" seems fairly common as well. When I was an EFL teacher I paid a lot of attention to the use of "will" and "going to" in conversations with other teachers, and based on my observations I feel like "going to" is actually more common. Is there a way to find out the statistics on these things?

My response was to modify my original statement, to say that "… a common way to express a future-time meaning is …", since in fact I had no evidence to offer about relative frequencies, only a guess. In contrast, Language Hat was in no doubt about the matter:

I think it's important to emphasize that "will" is not even the most common way to express a future meaning; the simple "present" is, I believe, more common in contemporary colloquial English. "We're going tomorrow" seems to me much more natural than "We will go tomorrow."

My response, again, is that we need to look at the facts. Though I agree with Steve's judgment about the cited example, it strikes me as not entirely typical. "We will" would be "we'll" even in less formal written English these days. And the implied context seems to me to shade the meaning in a direction that's more friendly to going to, in those situations where it's available as an alternative to forms of will.

So let's start by clarifying the nature and scope of this morning's experiment.

First, we're looking only at the forms of finite verbs used in clauses used to describe future-time events, states and so on, in contemporary American English.

Second, I'm going to ignore things like "I promise to be good" or "The future is bright", even though it's not obvious that their time reference is pragmatically much different from "I'll be good" or "Things are gonna be great". However, I'll include things like "until they arrive" or "when I yell 'go'", even though English pretty much insists on the present tense in constructions of that type. We'd get slightly different counts by looking only at finite verbs in main clauses, but that seems like an artificial restriction to me. And where I'm unsure about whether an example should be counted, I'll count it.

Third, as I said in responding to Joe, it's clear that genre, register and even topic make a difference. So I'll look at a variety of sources, consistent with the need to get on with my day by 9:00 a.m. or so. Let me stipulate that this is not an adequate sample — but perhaps it'll be enough to suggest the way the wind is blowing. In any case, it's a step up from simple assertions of divergent opinion.

OK, the first sample is a political speech, full of promises about the future: the transcript of president-elect Obama's December 6 radio address about his economic plans. In 773 words, he used forms of will 25 times. Of these, 4 were the negative form "won't" ("We won’t do it the old Washington way"), 14 were spelled "will" in the transcript ("It will put people back to work") , and 7 were spelled "'ll" ("We’ll invest your precious tax dollars in new and smarter ways"). He used present tense for future time reference in 2 when-clauses ("When Congress reconvenes in January, I look forward to working with them to pass a plan immediately"). There were no examples of going to, gonna, is to, shall, etc.

My second sample is a press release, a genre that is sometimes also future-oriented — "Stanford Law School Launches Intellectual Property Litigation Clearinghouse", 12/8/2008. In 1282 words, there were 4 examples of will, and one example of may that I judged to involve a clear future-time reference:

We are very excited about the research opportunities now available, as well as how the results of such research may aid the legal community not only in resolving disputes, but also in reducing the costs of resolution and providing insights for business and policy decision makers in the future.

My third sample is a news story — "Fighting Foreclosures, F.D.I.C. Chief Draws Fire", NYT, 12/10/2008. Along with 4 forms of will (2 of "will", one of "'ll", and one "won't"), there are 2 present-tense future references in main clauses. One is in a quote from the FDIC chair Sheila Bair:

And at the end of the day, I’m happy if Treasury just picks a plan and does it. Even if it’s not my plan, its better what we’re doing right now.

Another is in a quote from an anonymous White House official (which also features a future-time can:

We’re done in two months. The next administration can try to find a way out of that maze.

So far, will is more than holding its own — 33 forms of will, vs. a total of 6 future-time references using other verbal forms: 2 plain present tense in when-clauses, 2 plain present-tense references in main clauses, 1 can, 1 may. Even subtracting the political speech, will is winning 8 to 6.

But this is all formal written language, except for the quotations in the news story. Let's look at something more colloquial. I'll stack the deck by searching for "gonna" and taking the top four posts from the first U.S.-based blog I find. Alas, there's just 1 gonna, 1 going to, 3 will, 4 'll, and 1 shall (in a song quote "Shall we dance? Or maybe just limp around?").

OK, let's try another blog found by searching for "gonna". This one is much more colloquial:

Wow. I haven't pulled an all nighter and Felt THIS fine. it took me about 4hours to do my essay because i really tried. I was about to fall asleep until i got done, and i woke up! lOL. i was like fuck it. im gonna do homework all morning so i dont have to do it tonight. But yeahh.. its gonna be a bitch going to work 2-6. Probably bust home and nap an hour or so. Then mandatory Gyming with Justin. I guess i need an energy drink today.

Three entries are available. I count 5 gonna, 3 'll, 1 won't, 1 future-time don't, and 3 present-tense future references in subordinate clauses:

It'll be something we hopefully do til we're old men smoking out of a pipe when cannibus is legal.

So even in this case, gonna only accounts for 5 of 13 futures, and barely beats out forms of will. (It's also pretty clear that in this sort of English, there are differences in meaning among the various alternative choices for verbal expression of future-time reference.)

What about genuine conversational transcription? I picked four conversations at random from the Switchboard Corpus, and found 16 forms of will, 6 instances of going to, and 5 examples of future-time reference in (other) present tense clauses.

My tentative conclusions? In formal written English, forms of will are by far the commonest way to express future time reference in tensed clauses.  In more informal writing and in speech, forms of will are less dominant, but apparently still very much in the majority.

A larger-scale experiment might count corpus instances of the strings "will", "'ll", "going to", etc., and then check by hand a random sample of each collection, in order to estimate overall corpus frequencies. This kind string-search-based method will not catch other uses of plain present-tense verbs for future-time reference, but it would give a more reliable estimate of the relative frequency of will vs. going to in various kinds of material. This would take several breakfast periods, however.

[Update: in an odd moment between other chores, I scanned this web forum discussion, which is a regular orgy of speculation about the future. I counted 293 forms of will (283 "will" and 10 "'ll") to 33 instances of going to ("gonna" didn't occur), which is almost exactly the 9-to-1 ratio cited by Alex Boulton in the comment below (293/33 = 8.88). ]


  1. Stephen Jones said,

    December 11, 2008 @ 9:49 am

    I'm not at all sure this will be a useful exercise. For a start there are a large number of cases where 'will' and 'going to' and the 'present continuous' are not alternatives.

    We rarely use 'going to' for promises, offers, requests or spontaneous decisions.
    Are you going to turn on the light?
    has very different connotations from
    Will you turn on the light?

  2. outeast said,

    December 11, 2008 @ 10:01 am

    Is that relevant, Stephen? Surely the original hypothesis was simply about relative frequencies in references to the future. Most of the verb options for future references differ in meaning, so we would not expect to find that they could be interchanged with one another anywhere.

  3. alex boulton said,

    December 11, 2008 @ 10:30 am

    I looked at will and going to for a recent paper on using corpora in language learning. Part of the background was based on Doug Biber & colleagues' Longman Grammar of Spoken and Written English, based on a large corpus of English; the resulting grammar does, I think, show more corpus influence than any other (no offence to any other grammar… including, of course, one produced by certain Language Log members).

    Because it's the result of many years' work on a large corpus, the LGSWE can perhaps provide a wider picture than intuitions or informal experiments. Interestingly, it shows will is far more frequent than going to overall. My hugely condensed summary of LGSWE on this point is given below.


    Alex Boulton. 2007. DDL is in the details… and in the big themes. In M. Davies, P. Rayson, S. Hunston & P. Danielsson (eds.), Proceedings of the Corpus Linguistics Conference: CL2007.

    Biber et al. (1999) accept the view that there are only two tenses in English, present and past, with “no formal future tense in English. Instead, future time is typically marked in the verb phrase by modal or semi-modal verbs such as will, shall, be going to” (p. 456). Detailed treatment of these items is therefore reserved for the section on modals and semi-modals. Will is the most frequent of all modals, going to the most frequent semi-modal (p. 488-489); yet will is nearly nine times more frequent overall, occurring roughly 24,600 times per million words compared to only 2,800 for going to. There are significant differences in distribution of will and going to between registers (as well as between British and American English), suggesting that there is considerable overlap in meaning and use between the two items (p. 487). Indeed, the distinction between the two is “often blurred” (p. 496). Their main meanings are found to be volition and prediction; while will is very common in both functions, going to is mainly reserved for volition and largely limited to conversation.

    Ref: Biber, D., S. Johansson, G. Leech, S. Conrad & E. Finegan. 1999. Longman Grammar of Spoken and Written English. Harlow: Pearson.

  4. William Ockham said,

    December 11, 2008 @ 10:54 am

    Don't forget that, at least in my neck of the woods, you have to include 'fixing to' (that 'g' is usually not pronounced) in this discussion. It's more or less equivalent to 'going to' or 'gonna', but not quite the same. I'm not sure I can explain exactly how it differs.

  5. Moshe Jacobson said,

    December 11, 2008 @ 11:25 am

    Phrases such as "til we're old men", "when cannibus[sic] is legal", "When Congress reconvenes", "until they arrive", etc. shouldn't count, in my opinion, because they do not express volition or prediction. Instead, they're making assumptions of the occurrence of a future event in order to express volition or prediction of something else around the same time. In other words, these phrases can't stand alone in a sentence.

    Also, politicians' speeches use "will" a lot because it makes a weaker promise than "going to". So I don't think looking at such a speech is fair. Just my opinion.

  6. Crafty Green Poet said,

    December 11, 2008 @ 1:50 pm

    when I was learning italian I learnt that in that language the future tense actually implies uncertainty and that the present tense is usually used in cases where something is certain. I think its actually that way in English too and although outeast has a point that this kind of comment isn't relevant to the relative occurneces of certain words in reference to the future, I think it is relevant, because the actual future is different from the probable or promised future.

  7. John said,

    December 11, 2008 @ 2:01 pm

    I know it was just random that my blog happened to be in the top 4 on your blogpulse search, but I appreciate the link all the same. And this is a really interesting article. Not something I often consider when writing, or in general.

    I'll be adding this blog to my reader! (yeah, that was intentional)

  8. Dan Holden said,

    December 11, 2008 @ 2:46 pm

    You completely forgot about "finna" or "fittin' to" as my students are fond of saying.

    You could also quote my semi-okark raised father is "fixin'" to do it.

  9. Claire said,

    December 11, 2008 @ 4:37 pm

    Stephen Jones:

    As a bit of an aside, interestingly, in some Scottish dialects, there's no difference between the sense of your two example sentences.

    'Will I put the kettle on?', for my friend from Inverness, means 'Shall I put the kettle on?'

    and 'Will you put the kettle on?' has much the same sense as 'Are you going to put the kettle on?'

  10. Lukas said,

    December 11, 2008 @ 5:27 pm

    When I learned English in school (as a foreign language), we were taught some pretty clear cut rules on which form to use to express the future. Basically, the present continuous form was to be used for scheduled events, "going to" to express mere intent, and "will/shall" for everything in between. How far is that from reality?

  11. language hat said,

    December 11, 2008 @ 5:29 pm

    Language Hat was in no doubt about the matter

    Heh. Language Hat tends to express himself in a more emphatic and confident way than is really necessary. But I emphasize that 1) I did say "I believe," and 2) I was specifically talking about "contemporary colloquial English"; I am entirely unsurprised by the prevalence of "will" forms in speeches and the like. It's interesting that your sample, small as it is, produces a prevalence of "will" forms, and I'd be curious to see what a more extensive investigation would show. I'm always willing to reverse my judgments when presented with facts.

  12. Karen said,

    December 11, 2008 @ 6:27 pm

    I think (just off the cuff) that "will" is more common than we think because it's so often contracted into near unnoticeability. I have the idea that I don't say "will" very often – and mostly in questions or commands, but I do say "'ll" a lot. "will" is pretty emphatic in statements.

  13. Chad Nilep said,

    December 11, 2008 @ 7:29 pm

    I had the great pleasure to take a class with Doug Biber and Randy Reppen at the 2003 LSA Summer Institute. Most mornings Biber would start the session by asking the students in the class for an intuition about relative frequency, etc., and then show us the data as determined by Biber et al. 1999 or other corpus research. Our intuitions were almost invariably wrong.

    One of those intuitions we were wrong about was the frequency of 'going to' and related forms in casual speech. I – and I think most of the class – thought that 'going to' etc. would be more common that 'will' in casual speech. As Alex Boulton suggests, it ain't necessarily so.

  14. sleepnothavingness said,

    December 11, 2008 @ 8:39 pm

    Then there's the variant, "I'm off to make one of those" where "off" implies intent rather than motion (from the same part of the world that understands "while" to mean "until", causing mass confusion when a traffic sign once read "Wait here while light remains at red").

  15. Molly said,

    December 12, 2008 @ 3:25 pm

    sleepnothavingness: please tell me what part of the world that is. I'm assuming it's not the sleepless. thanks

  16. sleepnothavingness said,

    December 13, 2008 @ 12:11 am

    Molly: North Lincolnshire (East coast of England).

    "Well I'm not off to wait while that bloody light goz red!" – maybe some South American linguists will make a study of the soi-disant Yellowbelly some day.

  17. sleepnothavingness said,

    December 13, 2008 @ 7:32 am

    Clarification: "You'll etter wait while Christmas" would be a better example.

RSS feed for comments on this post