I recently noticed that German, English and Spanish seem to have a parallel colloquial use of their future tense, in which it's used to express a hypothesis about the present:

Literal meaning: I think he's driving right now.

English: He'll be driving right now.

German: Er wird gerade Auto fahren.

Spanish: (Supongo que) estará conduciendo ahora mismo.

There's a parallel use of future perfect to express a hypothesis about the past ("he'll have driven home yesterday"), which also seems to work in all three languages.

After a bit of research, it seems that at least French and Italian do the same thing, but without knowing the correct terminology for this it's hard to find out for others.

My question: where did this parallel usage come from? It can't have been inherited from PIE, considering that the future tenses in question were introduced later and don't share a common origin. Was this some form of cross-European borrowing (and if so, did the Germanic languages borrow from the Romance ones or vice versa), or was it invented independently - is there some sort of natural outgrowth in meaning of the future tense happening here?

Also, does anyone know whether there's a name for this?

  • 1
    Is it just that the natural tense / construction to use for a prediction easily generalizes to other kinds of conjecture?
    – rchivers
    Jul 11 at 17:30
  • 2
    The English and German cases are not future (Germanic languages only have past and present tenses), but rather auxiliaries with alternative uses. Eng will is a modal auxiliary meaning 'want, intend', like German wollen, and German werden is an inchoative meaning 'become', like English get. They are picked up in many other idiomatic constructions. Afaik, this has zero to do with the novel Romance futures formed by suffixing local reflexes of Lat habere auxiliaries.
    – jlawler
    Jul 11 at 19:16
  • 4
    @jlawler They are not future tenses, but along with the simple present, they are the primary constructions used to express the future, so yes, they are future. There’s obviously no morphological or etymological relation to the Romance futures, but their semantic usage is roughly the same, and they share the feature of also being used for suppositions about the present. Jul 11 at 20:09
  • 2
    This happens in Semitic too. E.g. the "imperfect" in Biblical Hebrew has a default non-past interpretation, but can be used in all tenses with a meaning of epistemic modality, "might" etc The link is that both future and this modal meaning are uncertain. I would expect some results if you search Google Scholar for "future tense epistemic modality".
    – Keelan
    Jul 12 at 6:10
  • 2
    There's a difference though between cases where you have a irrealis marker to start with - sounds like that includes Korean, and arguably English / German / Biblical Hebrew - and cases where a future marker collapses into a general irrealis marker, because "future" is not a tense concept or is not a very sticky one.
    – rchivers
    Jul 12 at 17:52

Epistemic modality ("He will be in London now") and futurity ("He will see her tomorrow") are generally associated with each other. This seems to be fairly common cross-linguistically, but there are enough subtle differences between them to make it quite individual to the speech variety in question.


In Old English, no synthetic future tense is attested, and future meanings are usually expressed with present tense verbs.

However we do see hints of some kind of future meaning derived from both "sculan", a modal verb meaning "must/should" and a transitive main verb meaning "owe"; and "willan", the modal and transitive main verb meaning "want".

E.g. from the years 996-7, Ælfric's Lives of Saints, "Of Saint Maurus", line 327 (or 330):

and þæt þæt he þe sæde is soð be dæle, swa þæt se mæste dæl ðinre muneca sceal of life gewitan binnan lytlan fyrste.

and that that he to-thee said is true in part, viz that the greatest part of-your monks shall from life go within little time.

and what he said is partly true, that is, that most of your monks will die within a short time.

There is a link drawn in Traugott (1989), to the Latin future active participle, in -turus. But Traugott goes on to say:

All these and other 'future'-oriented examples of OE willan and *sculan involve either a generalization, the 'prophetic/obligated' future, or simply 'later time' (relative tense). They are not deictic futures in which the time of utterance is the reference point. In other words, they are not fully subjective tenses dependent for the interpretation on knowledge of speaker time.

It was in Middle English that the "intentional subjective future" became strong grammaticalised, e.g. from 1386 Chaucer's Canterbury Tales:

I wol gladly yelden hire my place.

I will gladly yield my position to her.

Epistemic modality seems to have emerged a bit earlier, to my eyes at least. Traugott cites the following from version E of the Anglo-Saxon Chronicle, dated to 755:

swa swa mænige sæden þe hit geseon sceoldan.

as (=) many said that it see should.

as many who supposedly saw it (must have seen it) said.

However, the subjective version, where the use of the construction "shall/will + infinitive" implies the speaker's supposition, is much later. Visser (1966) dates it to the 19th century in Standard English, with earlier occurrences limited to Scottish and Northern English, and states that even as late as 1903, it was not wholly accepted.

Cross-linguistic differences

In Italian (as well as Spanish and Modern Greek), there is a use of the synthetic "future tense" in an epistemic role where English would take the must modal verb + infinitive construction, or the plain "present simple":

Non è a scuola. Sarà malato.

not is at school. be-FUT.3S ill

He is not at school. He must be ill.

Interestingly, French behaves like English, using the present of devoir + infinitive in this case. One of the proposed synchronic explanations of the difference is the future ratification hypothesis. Hence, we can see how there are differences in the precise nature of the epistemic modality covered by each construction in these languages, providing evidence that these developments are at root independent.

Senti's 2015 study on medieval, 15th- and 16th-century Catalan also shows how the division of labour between epistemic modality ("inference"), future tense, and deontic modality ("necessity") for the verbs deure and haver changed quite dramatically between the centuries. In Modern Catalan, deure has pretty much completely lost its temporal futurity. Again, each individual language's tense system has developed differently.

The case of Modern Korean 겠 (Yale: geyss) is very interesting, as it is a "future" that still retains a strong flavour of both deontic and epistemic modality. Its emergence dates from the transition of Middle Korean to Early Modern Korean in the 16th century, emerging from the merger of -게 ᄒᆞ- (-gey ho-), which was a causative [now taken over by -게 되다] then a passive, and anteriority/perfective marker -였- (-yess-); the product of this gained a meaning of "definite imposition of obligation". From that field of deontic modality, the epistemic modality and futurity meanings appeared. Of course, there are many other future-time constructions in Korean, and the specific shades of modality, futurity, voltion and assertion (as well as limits of politeness/formality levels) are all entangled.

Hence, these concepts are intimately related, whether it's in English, in Tundra Nenets, or in Mandarin Chinese. It's interesting to see what constructions will fill the role of futurity and epistemic and deontic modalities, and in what ways they'll continue to develop.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.