Agent tardigrade has proposed a most cunning idea to rescue the doctrine that English ‘will‘ encodes for future time. We are grateful to have our attention thus drawn to the matter: the idea is well worth thinking through. Here it is, as I understand it.
Upon hearing a knock at the door, we might opine
That will be the milkman.
In our design, this is a judgement about present fact, not any kind of prediction about the future. For according to us, ‘will’, being an aboriginal form, encodes presentness. Not being in epistemic contact with the facts (it isn’t as if we have looked out of the window and seen the man) we are not in a position to assert the proposition encoded by
That is the milkman.
So we do only thing available to us, offering a verdict based upon, and hopefully justified by, reasoning which we are prepared to make available on demand. (He always knocks at this time on a Saturday, and no-one else would trudge out to this benighted spot on the off chance of finding us in). As always, the other modals, and hence other verdicts, are available. Our verdict might have been any variant of the natural interpretations of
That will/could/may/might/should/must/needn’t be the milkman.
No-one is tempted to accord futurity to the verdicts expressed using the other modals, and nor, say we, should they be tempted to accord futurity to the verdict expressed by ‘will’.
But here is where the cunning idea comes in. ‘Will’ is unique amongst the modals, says tardigrade, in always encoding for future time. So the message envisages a future event. The future event in question of course cannot concern the presence of the milkman, for that is irreducably present. Instead what is envisaged is a potential future discovery of the facts of the matter, a later discovery that it indeed was the milkman who was knocking.
Cunning. The idea is that such uses of ‘will’ are elliptical – that when we say
That will be the milkman
our assertion should be understood as shorthand for something more complex, namely
That will-if-we-ever-come-to-know-the-facts-turn-out-to-have-been the milkman.
And that use of ‘will’ is, just as tardigrade theory requires, a straightforward prediction of a potential future event. Thus preserving the doctrine that ‘will‘ always encodes futurity.
What are we to make of this? For me, there are always two questions in such cases. (1) Can this proposal cope with all the teeming evidence? Does this theory work across the board, with all uses of ‘will’? And (2) if not, what made the proposal plausible in the first place?
Let’s take the first question first. I begin with the observation that tardigrade theory just does not apply to one large grammatical class of ‘will’-sentences, exemplified by
Nowadays, Grannie will often be found testing the famous echo in the British Museum Reading Room.
No thought here at all about the future. The natural interpretation of the sentence is a claim about Grannie’s present habits, and the presence of ‘nowadays’ nixes the possibility of a tardigrade translation. (Which would presumably have to be the nonsensical:
Nowadays, Grannie will-if-we-ever-come-to-learn-the-facts-turn-out-to-have-been often found testing the famous echo in the British Museum reading Room).
And as always (say we) a wider panoply of the modal paradigm is available. For the above is just one variant of
Nowadays, Grannie will often/can sometimes/may occasionally be found testing the famous echo in the British Museum Reading Room
And there are corresponding transported versions, now interpreted as claims about Grannie’s past habits:
In those days Grannie would often/could sometimes/might occasionally be found testing the famous echo in the British Museum Reading Room.
So tardigrade theory cannot hope to provide a completely univocal account of English ‘will’. Could it at least hope to unify the two distinct interpretations (present and future) of
Her Majesty will be drunk?
Stay tuned for Part II.