Presuppositions versus entailments

Families of sentences

Entailments and presuppositions are typically distinguished by family-of -sentence tests (Chierchia and McConnell-Ginet 1990). Given a sentence which appears to have some entailment or presupposition, one may negate the sentence, turn it into a polar question, or place it into the antecedent of a conditional. If the inference projects (or survives), then we have evidence that it is a presupposition; if it does not, then we have evidence that it is an entailment.

For example, both the clause-selecting verbs love and demonstrate appear to give rise to the inference that the clause they select is true;

1. 1. Jo demonstrated that she’s good at Go.
      
   2. Jo loves that she’s good at Go.

but the following examples illustrate that these inferences behave differently under negation.

2. 1. Jo didn’t demonstrate that she’s good at Go.
      
   2. Jo doesn’t love that she’s good at Go.

Likewise, under questioning:

3. 1. Did Jo demonstrate that she’s good at Go?
   2. Does Jo love that she’s good at Go?

Similarly, when we place the original sentences into the antecedent of a larger conditional sentence:

4. 1. If Jo demonstrated that she’s good at Go, she’ll have to join the Go league.
      
   2. If Jo loves that she’s good at Go, she’ll have to join the Go league.

In each case, the inference that the clause selected by love is true survives, while the inference that the clause selected by demonstrate is true does not.

We observe the same pattern when exchanging an indefinite determiner in some sentence with a definite determiner, suggesting that when an indefinite determiner trigger an entailment, an indefinite determiner triggers a corresponding presupposition.

5. 1. Bo brought a dog to his place of work.
      
   2. Bo brought the dog to his place of work.

In both of these sentences, we observe the inference that there is a dog. If we negate both sentences,

6. 1. Bo didn’t bring a dog to his place of work.
      
   2. Bo didn’t bring the dog to his place of work.

we observe that the inference persists only when there is a definite determiner. Analogous results obtain when we question the original sentences or place them into the antecedents of conditional sentences.

7. 1. Did Bo bring a dog to his place of work?
      
   2. Did Bo bring the dog to his place of work?
8. 1. If Bo brought a dog to his place of work, the boss will be mad.
      
   2. If Bo brought the dog to his place of work, the boss will be mad.

Hey, wait a minute!

Another test one can deploy (carefully!) is the “hey wait a minute” test (Fintel 2004). To apply this test, one constructs a dialogue in which person A utters a a sentence containing the expression of interest, and then person B utters the declarative sentence which is hypothesized to be a presupposition, preceded by “Hey, wait a minute!”. For example,

9. Bo: You love that you’re good at Go.
   Jo: Hey, wait a minute! I’m not good at Go!

this dialogue sounds okay, presumably because the inference that Jo is good at Go is a presupposition. Meanwhile,

10. Bo: You demonstrated that you’re good at Go!
    Jo: #Hey, wait a minute! I’m not good at Go!
    Jo: No, I didn’t!

when we have demonstrated instead of love, it becomes noticeably odder (IMO). This difference is chalked up to love triggering the inference as a presupposition, while demonstrate triggers it only as an entailment.

Sometimes the test can give a somewhat weird result, even when the relevant inference is not an entailment, as (11) shows.

11. Jo: You brought the dog to your place of work!
    Bo: Hey, wait a minute! There isn’t a dog!

Thus this test doesn’t work as reliably as the family-of-sentence tests do—it seems to be a bit more sensitive to factors apparently unrelated to whether or not the relevant inference is a presupposition;¹ it therefore must be deployed carefully.

References

Chierchia, Gennaro, and Sally McConnell-Ginet. 1990. Meaning and Grammar: An Introduction to Semantics. Cambridge: MIT Press.

Fintel, Kai von. 2004. “Would You Believe It? The King of France Is Back! (Presuppositions and Truth-Value Intuitions).” In Descriptions and Beyond, edited by Marga Reimer and Anne Bezuidenhout. New York: Oxford University Press.

Footnotes

1. What are these factors? If I had a nickel for every time I’ve wondered this, I would be a rich man.