When someone trains a dog to understand a command, they are actively making the dog respond to signals such as a pointing gesture and spoken words; they are establishing a signal–response relationship.
This unique communicative behaviour is perhaps taken for granted. It is not considered relevant for a definition of language or human communication by most theorists. Instead, theorists consider that these types of persuasive acts, which may involve giving other animals treats or coercing them, must have evolved within a cooperative society of apes, along with ‘language’ and other human-specific traits.
This question ties in with my previous post, in which I discuss another cognitive feature of humans: namely, ostension, the recognition of communicative intention. Indeed, if dogs could recognise, as we do, that someone who produces signals is trying to tell them (to do) something, then establishing these rewarding relationships would be a lot easier for both species. Like a human child, the dog would infer whatever it is that the human wants with his words and gestures.
Thus, when training a dog to fetch, the human is seen as trying to share an experience or mental state with the dog. He may point his finger toward the ball that he has just thrown and make other attempts to send his playful intention over to the dog’s mind. This foundational ability for ‘shared intentionality’, theorists posit, was somehow developed by our cooperative ancestors well before language and dog training.
However, dogs exist in their current form and are able to share intentional states with us because humans have bred them for desirable traits. Hence, it would seem that the behaviour of training others (particularly children) to respond to signals explains the evolution of shared intentionality and ostension, rather than the other way around.
Still, as I have explained, theorists prefer to adopt a sociocentric perspective, from which these positive socio-cognitive traits evolve irrespective of what individuals are doing to make their relationships (on which their society is based) work. The case of the dog–human society is an obvious illustration of this error, insofar as these cross-species relationships have been established by our ancestors, not by the ancestors of dogs or by the ‘force’ of natural selection.
However, consider a human-to-human example: “Suppose that we are in a bar. It is your turn to buy the drinks, and I would like another. I intend for you to believe this … and I therefore make sure that my empty glass is visible to you” (emphasis mine). There is not enough in this action to suggest that the signaller meant that he would like another drink; meaning requires, for example, “making eye contact with my friend and simultaneously tilting my wine glass”.
From a psychological perspective of humans who have inside knowledge about meaning and other socio-cognitive traits, this notion is clearly important. However, the question remains: why would an animal respond to the ostensive behaviour of making eye contact and tilting a wine glass? From an objective, biological perspective, this signalling behaviour is no different from the more primitive behaviour of making sure that the empty glass is visible to the receiver; both are a matter of one organism attempting to change the behaviour of another through signals. In nature, nothing evolves unless it gives organisms a benefit in terms of survival and reproduction. So, human receivers must be deriving an added benefit from their increased attention to signallers, with respect to an ancestral situation in which no such attention was necessary to get meanings across.
Given what I call prescription, the behaviour of language, the answer is clear: Human animals are receptive to ostension because ostension signals that there are consequences for not responding to signals. In this bar example, we can imagine the signaller getting upset with the receiver and acting in various ways detrimental to them if they do not pay attention because he really wants his glass filled, not because he is interested in their attention for the sake of it. Moreover, we can easily picture other signallers that compete for the same receiver’s attention and offer incentives, especially millions of years ago.
Ostensive-inferential behaviour, then, is an adaptation to the basic behaviour of making receivers respond. This is hard to perceive if we adopt a historicist, sociocentric perspective of communication in which receivers remain connected to signallers across evolutionary time.