Reference games are fun

After taking Chris Potts' class on semantics and pragmatics, I think that conversational implicatures are often some of the most fascinating parts of linguistics.

Barner et al's research into the differences between 'some' and 'all' with children is exceptionally interesting, as it points towards ways in which children initially have learned their reasoning. Children respond better to more specific statements as opposed to more general ones. This to me suggests that they have not yet learning the general rule of how 'some' and 'all' work, and are working on a case by case basis. I found this ironic, as I remember us learning about how we hypothesized that it was impossible to learn certain grammar rules through a case-by-case reading earlier in the quarter.

Furthermore, the paper suggests that children cannot understand stronger reasoning than the language suggests, they very much take things at face value. This inherently makes sense to me, as a lot of what we understand from statements in conversation, the scalar implicatures etc, are not natural or inherent. The fact that 'some' does not mean 'all' is not hidden in the definition, it is learned through usage, which children will not have been exposed to.

Stiller, Goodman, and Frank's paper discusses much of the same ideas, focusing on actual experimental data. I worked in Professor Goodman's lab for a time, and also worked a lot with similar reference games to the ones discussed in this paper last year in Potts' class. Interestingly, I started work on a project that was closely inspired by and related to this paper, as my friend and I were interested in how we could force certain types of implicatures based on world knowledge. We were trying to think of ways to put people in the position where their world prior was slightly changed (describing the temperature of a scene in F as opposed to C, and then asking questions about it). This was different from ad-hoc implicature, as we were working along similar ideas as another paper by Goodman et all, "Wacky worlds", but essentially wanted to theorize what would happen if people were forced to choose between a 'purely pragmatic' choice, and one that was colored by their changed world prior (not the original one), which one would be favored.

Overall, I think that both these papers are very cool, and a great introduction into one of my favorite subjects in linguistics.