When Negation and Epistemic Modality Combine: The Role of Information Strength in Child Language

Abstract

Negative sentences with epistemic modals (e.g., John might not come/John can not come) contain two logical operators, negation and the modal, which yields a potential semantic ambiguity depending on scope assignment. The two possible readings are in a subset/superset relation, such that the strong reading (can not) asymmetrically entails the weak reading (might not). In this situation, a potential learnability issue arises. Based on the Semantic Subset Principle, we anticipated that children’s initial interpretations would sometimes differ from those of adults because children are expected to initially prefer strong (can not) readings for sentences that convey weak (might not) readings for adults. This proposal is investigated in two experiments using Standard Italian, which is an ideal testing ground for child language, in view of its simple modal paradigm. The results of these experiments confirm the predictions of the Semantic Subset Principle. Five-year-old Italian-speaking children were found to strongly favor the scope assignment that generates strong (can not) readings, even in cases where adults strongly favor the weak (might not) scope assignment. This result is discussed in relation to some recent alternative proposals (Gualmini & Schwarz, 2009; Musolino, 2006) that do not assume any initial bias toward the strong readings.

ACKNOWLEDGMENTS

We wish to thank Rosalind Thornton, Drew Khlentzos and the audience of The Romance Turn V, Lisbon for helpful comments. A special thank also goes to Lena dal Pozzo and to the children, the parents and the teachers of the childcares where this research was conducted: S. Benedetto (Siena), Scuola Materna Mameli, and Kindergarden (Florence).

Notes

¹ In some varieties of English, the indicative forms may not and can not can both express impossibility. To avoid confusion, we have chosen to use the subjunctive form might.

² In the vast majority of linguistic contexts, the logical expressions any and some are in complementary distribution, such that any must be interpreted inside the scope of negation, whereas some must be interpreted outside the scope of negation. See Crain (2012) for examples of linguistic contexts in which some and any are assigned the same meaning.

³ Although children’s justifications of their rejections were not reported, it is worth commenting on the counter-intuitive finding that children correctly rejected false negative statements like (23) more often than they correctly accepted true negative statements like (22). This makes sense if, unlike adults, 5-year-old children generate scope assignments that make sentences true in the narrowest range of circumstances, i.e., the strong reading. If so, then both (22) and (23) mean that there cannot be a parrot/bear in the closed box. This makes (23) clearly false, and it also makes (22) false (but perhaps less clearly so). A potential problem for this scenario is that children should also have rejected false strong sentences like “There cannot be a parrot in the box.” However, comparisons like this are difficult to assess. We return to this issue in Experiment 1.

⁴ Another quasi-modal is “bisogna”, with a meaning similar to need. This form has idiosyncratic properties which distinguish it from the other two modals potere and dovere. For example, in Standard Italian, it cannot have a referential subject.

⁵ In the case of dovere, an ambiguity remains when the modal follows negation: non dovere could be either translated in English as “need not” or “must not.”

⁶ The position of the clitic ‘ci’ can vary, so an alternative to sentence (25) is “può non esserci una mucca nella scatola”, with the clitic following the copular verb. The absence of ‘clitic climbing’ here is possibly related to the absence of ‘restructuring’ effects (see Rizzi, 1982). This difference, however, has no consequences for scope assignment.

⁷ A difference is the presence of an extra object (e.g., a strawberry). The extra object was added outside the boxes to counterbalance true and negative answers. In this way, we have a false positive sentence There might be a strawberry in the box and a true negative strong sentence There cannot be a strawberry in the box. These were not among the conditions tested in Noveck (2001).

⁸ Children made comments along these lines about 34% of the time (on 22 of their 63 rejections). In the remaining cases, they gave no explanation or said “I don’t know.”

⁹ In the given scenario, Box C could contain either a horse alone or a horse AND a cow. In this context, the target sentence “There might be cow in the box” might be considered as under-informative, since the horse hasn’t been mentioned. The sentence could nevertheless be accommodated by assuming that the focus of the sentence is on the cow.

¹⁰ For some speakers, sentences (47) and (49) are false when a horse is in the box. Even under this more restrictive interpretation, (47) and (49) are not stronger than (46) and (48), since they do not asymmetrically entail those in (46) and (48).

¹¹ When uttered with a neutral intonation, negative sentences like (52) are generally unambiguous. To capture this cross-linguistic generalization, a universal constraint has been proposed to ban the covert movement of universal quantifiers from object position (Mayr & Spector, 2010).

¹² The frequency of embedded sentences in child directed speech ranges between 6% and 11% of the overall input (Morgan, 1989). Arguably, very few of these are negative sentences like (57). More generally, sentences with two scope-bearing elements embedded under DE operators are rare.