Quick Summary:

What's in our minds when we throw an if/then sentence out there? How do we work out what worlds we may be talking about? In this episode, we talk about the semantics of conditionals: what an "if" looks like logically, why a simple logical arrow isn't enough to capture the complexities of conditionals, and how we change what possibilities we allow ourselves to think of based on what our "if" clause holds. 

Extra Materials:

While our episode about the meanings of conditional sentences had many different kinds of examples, we focused our explanation mostly on what are traditionally called indicative conditionals, like this one.

(1)    If you study enough history, you will pass the test.

An indicative conditional is called that because its first clause — the if-clause — is said to be in the indicative mood, which here just means that the verb is in its simple present form. In terms of how they’re typically used, indicative conditionals seem to go hand-in-hand with the idea that the possibility being expressed by the if-clause is a fairly likely one. Or, at least, one that isn’t totally out-of-the-question.

But indicative conditionals stand in contrast to what have been variously called subjunctive or counterfactual conditionals. And as we’ll see, these two terms aren’t entirely interchangeable, with neither one actually being a very good description of what’s going on underneath the surface. So, let’s dig in a bit and look at some examples.

(2)    If you hadn’t become a violinist, you would have become an assassin.

In this case, the if-clause is in a form that’s often labelled the subjunctive mood; at first glance, these kinds of sentences are usually used in conversations to suggest that the possibility of the first half being true is either very low, or else completely contrary to established fact (i.e., counterfactual).

The idea that the if-clause is in the subjunctive mood, though, doesn’t make a whole lot of linguistic sense. The forms that the verbs “have” and “become” take on in that sentence appear to involve the same tense and aspect morphology found in other non-subjunctive contexts, like the one below.

(3)    By the time she was a young woman, she had become a violinist.

And while the verb “be” does occasionally take on the special form “were” inside these sorts of conditional sentences, at least from the standpoint of prescriptive grammar, many find the more regular past tense form “was” equally acceptable.

(4a)    If she weren’t a violinist, she would have been an assassin.

(4b)    If she wasn’t a violinist, she would have been an assassin.

Since most verbs don’t have any special subjunctive form in these sentences, and since the verb “be” only has one in the first and third persons, and only for some speakers, classifying the if-clause as being ‘subjunctive' seems like a misnomer.

(English does have a subjunctive mood, which is easiest to see in mandative sentences, like those expressing a demand. For a number of speakers, the embedded verb in “It’s crucial that he always arrive on time” must appear in its bare uninflected form, in contrast with the very same verb found in “It’s clear that he always arrives on time,” which shows up with its usual third person singular suffix.)

Moreover, calling conditional sentences like the one in (2) above counterfactual doesn’t quite work either. Consider the sentence below, and imagine it being uttered by a doctor after they’ve been observing a patient for some time.

(5)    If he had been infected by the flu, he would be showing exactly these symptoms.

It’s easy to see how this sentence could be used by someone who already believes that the first half is true, in an attempt to convince someone else that a sick patient really does have the flu. So describing the sentence in (5) as ‘counterfactual’ in this scenario seems inaccurate.

Still, there does seem to be some kind of difference between a prototypically indicative conditional like the one in (6a), and its canonically subjunctive/counterfactual counterpart in (6b). Roughly speaking, the second one still feels like it’s somehow more remote. So, what could this difference be?

(6a)    If Howard gets the intel in time, he’ll be able to stop the terrorist plot.

(6b)    If Howard had gotten the intel in time, he would have been able to stop them.

Nailing down the exact answer to that question is challenging, because this is an active area of research and debate. Some of the proposals that have come out of the field of philosophy handle conditionals like the one in (6b) by applying an analysis that’s similar to the idea of strict implication, which is something we considered and ultimately rejected in our episode. On this view, the word “if” basically goes back to being a binary connective, linking two clauses together and asserting that the possibilities described by the first form a subset of the possibilities described by the second (alongside some additional details about which possibilities are up for consideration). But the existence of sentences like (7), where each if-clause connects up to a noun phrase instead of a completed thought, strongly points towards the restrictor analysis we ended up with.

(7)    The likely outcome if their mission fails and the likely outcome if it succeeds are very different.

One promising and more recent alternative that fits together nicely with our restrictor theory assigns the same overall structure to both indicative and non-indicative conditionals, and locates the difference between them in two other places: the choice of modal verb, and the moment in time being referred to. In ‘subjunctive’ conditionals, the modal verb is usually “would,” while the sentence as a whole is generally in the past or past perfect. Unlike the modal “might,” which conveys epistemic possibility — meaning we use it to talk about what may or may not be true, according to what we know — it’s supposed that “would” communicates historical necessity, which means we use it to talk about what’s going to happen in the future, given what’s happened in the past.

If this is the right way to think about the meaning of “would,” then in certain conditionals, the modal conspires with the tense/aspect to essentially rewind history back to the time of some salient event, and to then say something about what would have surely gone on to happen, given the proposition represented by the if-clause.

Put another way, in both indicative and subjunctive/counterfactual conditionals, the if-clause serves to restrict the possibilities being quantified over by whichever modal verb happens to be there. In sentences with a “must” or a “might” in them, this often means temporarily adjusting our state of knowledge and then making a claim about what follows from that update. In a sentence with “would,” this means running the clock back to some earlier period, and then incrementing the passage of time by one sentence, to observe the results.

So the supposedly deep divide between these flavours of conditional isn’t necessarily as dramatic as it’s always made out to be! 


So how about it? What do you all think? Let us know below, and we’ll be happy to talk with you about how conditionals work and what they mean. There’s a lot of interesting stuff to say, and we want to hear what interests you!

blog comments powered by Disqus

Previous Topic:                                                                                                                                                                                                            Next Topic:

The Syntax of Silence                                                                                                                                                                              Coming soon!