Richer-looking signals still convey states and trigger responses rather than referring

The argument

Those who hold that animal communication is not very complex accept most of the observations and dispute what follows from them. Their central distinction is between a signal that carries information and a system that can say things. A call which reliably differs with the kind of predator carries information, in the sense that a listener who responds to it does better; but on this account the mechanism is an inherited link between a stimulus, a call and an escape behaviour, with nothing in between that resembles a thought about a predator. Proponents argue that treating such links as reference describes what an observer can infer rather than what the animal is doing. The evidence they rest on is what the systems cannot do. Animal signalling, on this view, is bounded: each species has a small inventory, the inventory does not grow within a lifetime, individuals do not coin new signals for new situations, and nothing in the system allows a signaller to describe something absent, to ask, to deny, or to report what another animal did the day before. A system with a fixed inventory and no means of combining items into new meanings, proponents argue, is not a simpler version of language; it is a different kind of thing. The position is therefore about kind rather than quantity. Proponents do not claim that animals are insensitive or that signalling is crude, and they accept that it is finely tuned by selection. They claim that fine tuning is not complexity in the relevant sense, and that the interesting question is why one species crossed into an open-ended system while the rest did not.

Premises

Counter-arguments

Those who hold that animal communication is complex reply that the standard being applied guarantees the result: define complexity as the full set of features of human speech and every other system fails by definition, however much structure it has. They also dispute the factual claims. Inventories in some species are learned and do change within a lifetime, signals in others are combined into order-sensitive sequences, and the absence of questions or denials may reflect what those species need rather than what their systems could carry. Setting the bar at open-endedness, they argue, answers a different question from how complex the signalling is.

Rejecting the premises

[Rejecting P1] The thinner mechanism is assumed rather than shown: listeners respond to calls played with no predator and no caller present, and adjust when a call conflicts with what they can see, which is hard to explain as a reflex. [Rejecting P2] The factual claims are contested: in several species the inventory is learned and does change within a lifetime, and signals are combined in ways that carry meanings the parts lack. [Rejecting P3] Treating open-endedness as the criterion of complexity settles the question by definition, and a bounded system with reference, combination and learning is a simpler version of the same thing rather than a different kind of thing. [Rejecting C] The conclusion answers a question about language rather than the question actually asked, which is how complex animal communication is.