Bird syntax? Interesting new study in Nature but I’m not convinced. They report that birds have a particular behaviour associated with one call ABC (scan for danger), and another behaviour associated with D (approach nest). They say that when the birds hear ABC-D, you get scanning followed by approach, but with D-ABC, no reaction, and conclude that there’s a compositionally interpreted rule for the former, but no rule licensing that latter. But surely the lack of reaction to the reverse call (approach then scan for danger) could simply be because this is behaviourally odd, as opposed to not being generable by rule? Signalling to approach a nest and then to scan for danger (as opposed to the other way around) might just be unhelpful behaviour. If that’s the right interpretation, then there’s no compositional syntax, just serially ordered behaviour associated with the order of calls, with one kind of behaviour outside of the birds’ usual repertoire.

http://www.nature.com/ncomms/2016/160308/ncomms10986/full/ncomms10986.html#close