g. when told to point to the man with the hat in the context of two men, each with a hat). An extensive developmental literature investigates whether children are aware of the ambiguity of these instructions (Asher, 1979, Robinson and Robinson, 1976, Robinson and Robinson, 1977, Robinson and Robinson, 1982, Bearison and Sorafenib cell line Levey, 1977, Ackerman, 1981, Flavell et al., 1981, Beal and Flavell, 1982, Robinson and Whittaker, 1985 and Plumert, 1996; Beck et al., 2008; among many others). Two of the major findings suggest that they are not. First, children do select a referent in spite

of the ambiguity, and, second, they report that the instructions they were given were adequate. The latter is typically investigated by asking the child to tell the experimenter if s/he gave them enough information or not. For example, Robinson and Robinson (1982, experiment 1) report that when asked “Have I told/shown you enough about my card for you to get it right?” (ibid.: 273) 39 out of 52 children aged between 5½ and 7 agree that they have been told enough when in MLN0128 order fact the experimenter’s instructions were underinformative. Similar findings are reported in their second experiment,

and in several other studies where the question was phrased in terms of a binary choice (Robinson & Whittaker, 1985, experiments 3 and 4; Beal and Flavell, 1982 and Flavell et al., 1981, who asked children “Do you think the instructions Aspartate told you in a good way or in a not-so-good way how to [complete the task]”). Nevertheless, Beck et al. (2008), Nadig and Sedivy, 2002 and Nilsen and Graham, 2009 and others present evidence that children may be sensitive to the ambiguity in the referential

communication task, albeit in more indirect ways. Such evidence has also been available early on in this line of work, as Patterson, Cosgrove, and O’Brien (1980) report that children showed longer reaction times for ambiguous than non-ambiguous messages, and made more eye-contact with the speaker. Plumert (1996) reports that children were delayed in starting to search for an object when the instructions did not disambiguate the hiding place; and Flavell et al. (1981) report that asking children to follow ambiguous instructions to build a model elicited pauses and puzzled expressions. Moreover, Jackson and Jacobs (1982) and Brédart (1984), who used the sentence-to-picture matching paradigm, report that children are very good at selecting the referent for which the instructions would be informative, rather than the referent who was compatible with the instructions but for which the instructions would have been underinformative. These findings tentatively suggest that children can detect ambiguity, but for some reason resist correcting their experimenter.