'But' how do children judge it on a scale?

‘But’ how do Children judge it on a Scale? Leen Janssens (Leen.Janssens@ppw.kuleuven.be) Laboratory of Experimental Psychology, Tiensestraat 102 3000 Leuven, BELGIUM Annelies Van den Broeck (Annelies.Vandenbroeck@student.kuleuven.be) Walter Schaeken (Walter.Schaeken@ppw.kuleuven.be) Laboratory of Experimental Psychology, Tiensestraat 102 3000 Leuven, BELGIUM Abstract This study examined children’s comprehension of the conventional implicature induced by but, combined with so and nevertheless, in ‘p but q’ sentences constructed as distancing-contrastive connections. Based on the Pragmatic Tolerance Hypothesis of Katsos and Bishop (2011), a three- point scale was used as response format. Using a scale instead of a binary judgment task can reveal more insight in which factors are considered most important when processing ‘p but q’ sentences. The results indicated that the content of the p- and q-arguments plays a very important role when children process ‘p but q’ sentences. However, their use of the three- point scale also indicated that they are sensitive to the pragmatic meaning of but, so and nevertheless. These results must be interpreted cautiously since the children seemed to use the middle value on the scale around 30% of the time in each sentence category, which was not in line with our predictions. This might indicate that children experience a general incomprehension with this type of sentences and answer with the middle value on the scale because they simply don’t know the answer. Keywords: conventional implicature; but; scale; content Introduction Over the past few decades, considerable experimental research has been devoted to scalar implicatures. Grice introduced the term implicature in the 1967 William James lectures to offer an explanation for how it is possible that an utterance can mean more than what is literally said. Scalar implicatures are a subcategory of conversational implicatures and are based on a scale of informativity. For example, on the scale the use of the more informative all logically entails that some is also true. However, in an utterance such as ‘Some Belgians like to drink beer’, the pragmatic meaning of some causes the hearer to interpret this utterance as ‘Some but not all Belgians like to drink beer’ even though the logical meaning of some is ‘some and perhaps all’. According to Grice (1989), people follow a set of maxims in communication in order to understand each other correctly. That’s why the consensus applies that whenever a speaker uses a weak term such as some, a stronger term such as all does not hold. The speaker would not have been optimally informative if a stronger term applied. Developmental conversational implicature research has shown that children are less pragmatic than adults. For example, Noveck (2001) found that 89% of the 7-to-8-year- olds in his study agreed with statements such as ‘Some giraffes have long necks’, compared to only 41% of the adults. Similarly, with respect to propositional connectives, Braine and Rumain (1981) presented evidence showing that deductively competent 7- and 9-year-old children favor a logical interpretation of or (‘p or q and perhaps both’) over an implicit one (‘p or q but not both’). Adults on the same task were equivocal, though they tended to favor exclusive interpretations (Braine & Rumain, 1981). However, these (and other) studies claiming that children lack pragmatic competence have been criticized by Katsos and Bishop (2011). In their implicature studies, Katsos and Bishop (2011) argued that earlier studies mostly employed tasks that cannot differentiate between actual implicature derivation and mere sensitivity to violations of informativeness. The majority of studies concluding that children are more logical than adults used binary judgment tasks in which participants were instructed to judge an utterance as ‘true’ or ‘false’. Katsos and Bishop (2011) argued that children might not reject underinformative sentences because they are tolerant to violations of informativeness. However, this doesn’t mean that they are not sensitive to these violations. In order to test this Pragmatic Tolerance Hypothesis, Katsos and Bishop (2011, Experiment 2) instructed their participants to judge on a ternary scale how well a fictional character described certain situations. They found that children’s performance did not differ from adults’. Underinformative utterances were judged by both groups with the middle value on the scale. This shows that children understand that using for example some, when all would have been a more informative description, is not optimal. However, in a binary judgment task they would not penalize such a description as false whereas adults would. In previous research (e.g. Noveck, 2001) this falsely led to the conclusion that children lack pragmatic competence.