Degrees of Truth

This week’s readings by David Barner, Alex Stiller, and their colleagues focused on scalar implicature and pragmatic inferences. Their papers were concerned with the information conveyed by what is not said in a sentence such as “I ate only some of the cake”. An experienced listener would naturally infer that not all of the cake has been eaten. Barner and Stiller utilized experiments to test the hypothesis that children are simply less effective at performing scalar implicature than adults. Their findings demonstrate that children generally struggle not with scalar implicature, but rather the mechanics of imagining “strengthened” alternatives to a sentence construction. For example, utilizing “all” when it can be appropriately be applied as opposed to “some”. When presented with “strengthened” or more descriptive sentence constructions, children usually deem these versions more appropriate. At the end of his paper, Stiller raises an interesting point about how our “pragmatic computations operate over our knowledge about the world”. As one might expect, since adults have lived longer, they have a greater wealth of worldly experience from which to develop pragmatic inferences.

            I argue that children also consider a certain degree of truthfulness, which is lower than that of adults, when thinking of appropriate sentence constructions to describe a scenario. For example, if a child is a shown a picture of three sleeping dogs and is told “some of the dogs are sleeping”, they would agree with this statement. Barner states, in a similar case involving horses, that adults deny this statement as a “good description”. However, for all intents and purposes, “some of the dogs are sleeping” could be considered a good description because it is not untrue. Children might only consider statements which are blatantly false as problematic rather than considering “degrees of truth”. An adult considers a usage of “all” to be more appropriate because it is a more truthful and therefore a “stronger” description of the scenario at hand. What appears to separate the degree of truth to which an adult or child holds a description is a greater inferential understanding of what the speaker is more likely to say based on years of experience. Where children may stop at using “some” because they deem it sufficient, the experience of adults allows them to generate and utilize a construction involving “all.” This aligns with Stiller’s conclusions about frequency and rarity.

            Since children presumably are less familiar with hearing “all” as opposed to “some” in certain scenarios, they likely equate the two words and fail to scale them in the same manner as with numbers. I am opposed to referring to those descriptions which appropriately utilize “all” versus “some” as “good” versus “bad” descriptions, but rather as more truthful and less truthful descriptions. Some descriptions are simply more common when applied to specific scenarios, but I do not think this necessarily means those uncommon descriptions are “bad”, as might be suggested by the usage of “good” when referring to high-frequency descriptions.