How Did Language Evolve? -- II

In my last blog, I noted that efforts to teach primates human language demonstrate that we are unable to teach them human or comparable languages or they are unable to learn them beyond what a 2 or, perhaps, a 3 year old human can learn. This research clearly adds no support to the view that human communication is somehow directly linked to advances in primate communication before we split off from the other primates, which was the contention that got me thinking about this issue..

In order to understand the evolution of language it will be useful to step back and take a look at what it is we ultimately want to account for. In the simplest of cases, I, as a speaker will have some goal I wish to achieve -- inform someone of something I believe they need or may want to know, or request information from them, or request them to do something, or invite them to join with me in doing something, or make a bet with them, among many other possible things. What I say will reflect my goals, the relationship I have with my addressee, background knowledge I believe we share that is salient and relevant to my achieving my goal, and aspects of the immediate context that are salient and relevant to my achieving my goal. There are surely other considerations. At the very least, human communication is a very complex business.

If I want you to give me a ride home from work and I am your supervisor, then I will go about this in much different way than if you were my supervisor. If we are personally close that fact will affect how I go about this. My knowledge of when you usually leave work would be relevant. My knowledge of how much you would have to drive out of your way to give me this ride would be relevant. If I have never made a request like this of you and don't know where you live exactly, I will likely approach this quite indirectly, perhaps saying something like "When are you leaving work today?," or "Do you live anywhere near Green Street and Elm Street?," expecting or hoping you can infer my goal from what I say. The former is so indirect the point might be missed. The latter will usually be inferred to be the start of a ride request. Utterances like these get at your ability to give me a ride home (our wanting to leave at much the same time) and your willingness to do so (you will likely not want to drive me home if this will add a great deal to your driving time.) And, in general for any goal I may have in addressing you, there will be conditions that must be satisfied before I can achieve my goal and what I say will reflect the nature of these conditions and how I want to go about addressing them. The linguistic choices I make will reflect social concerns as well, for I would want to avoid endangering our relationship by burdening you or disrespecting you.

The failure to understand the complexity of human communication and how language fits in in our efforts to communicate with each other lies at the heart of the failure of syntacticians to understand what syntax is about. The machine that drives communication is pragmatics and syntax must be as much about how how social choices influence the syntax and morphology of utterances ("Would you mind giving me a ride home tonight?" vs. Please give me a ride home tonight," both of which have different conventional meanings but can be used to achieve the same goal) as well as conventional meaning, but most remarkably the utterances we find in actual communication rather than the made up sentences syntacticians focus on can be quite ungrammatical by any normal linguistic standard.

In my blog on The Use of Simplification in Scientific Research, I describe a ride request in which the notion of a ride is never referred to directly, a request for a ride is never made directly, and the rejection of this request is never made directly. If you were to listen to this conversation (which I did in Italy of all places at a conference) you would find it quite unremarkable. On the other hand, a transcript of this conversation would drive any conventional syntactician crazy for the critical utterances are highly elliptical and hence ungrammatical by most standards.

In my book on Speech Acts and Conversational Interaction, now out in paper back but sill very expensive (go to your library if you want to read it), I cite a telephone conversation that, after a greetings sequence, begins with a "Guess what?" followed by the reply, "What?," as is normal, followed by

My car is stalled...and I'm up in the Glen...

followed by the addressee offering up an unhelpful utterance of "Oh", followed by the caller saying

And I don't know if its possible ... but see ... I haveta open up the bank in Brentwood.

The sequences of periods indicate pauses. I have omitted audible breathing noises.

The first of these examples is grammatical if you ignore the pause. The second is wildly ungrammatical. But, the addressee immediately grasps the meaning (significance) of what is said and replies and does so in an equally elliptical and ungrammatical way that is also perfectly clear in meaning (significance). As I said earlier, this conversation sounds perfectly normal despite bing highly elliptical and officially "ungrammatical."

Collapsing what the caller says that bears directly on the ride request into just one long sequence, we have (with the elided material put in in bold face):

My car is stalled...and I'm up in the Glen...And I don't know if its possible for you to pick me up and give me a ride to my bank but see I need for you to pick me up and give me a ride to my bank because I haveta open up the bank in Brentwood.

In my book, I show how seeing conversation as driven by pragmatics, not semantics, facilitates an account of how we understand such elliptical utterances. The basic idea is that the elided material addresses the conditions on a ride request being successful and is recoverable by the listener's understanding of the structure of ride requests.

Clearly, some inferencing was required by both parties for them to understand each other's highly elliptical utterances, which is to say, that speaking and thinking and hearing and thinking are so intimately intertwined that we cannot seriously discuss the evolution of language without also considering the evolution of cognition. Put somewhat differently, any account of the evolution of language must be embedded in an account of the evolution of communication and both must be embedded in an account of the evolution of cognition. To study the evolution of language as if it were some free-standing human acquisition would be a serious mistake in my view. When simplifying scientific problems to make them easier to understand one may oversimplify them to the point that one cannot make any real progress at all.

So, to account for the evolution of language we have three problems at the very least. They are

1. We must account for the development of the physiological abilities required for speaking and for what I will call "linguistic hearing," i.e, our ability to discriminate linguistic sounds from other sounds, something that happens at a very early age.
2. We must account for the cognitive abilities required to learn and use languages.
3. We must account for the development of any language specific abilities that are independent of our other cognitive abilities.

By 3 I mean to refer to the development of language-specific 'language centers" in the brain among other things.

At this point, I believe we can conclude that accounting for the evolution of language will likely prove to be impossible since we can infer little from the communicative abilities of our primate cousins and will likely be unable to infer much about the structure of our brains from the fossil record. That's why I have always backed away from the issue. But even if we are going to be unable to answer the question constituting the title of this blog, it may be useful to understand why.

One thing is pretty clear, I believe, and that is that in the course of the evolution of language, it only makes sense to assume that mutations that favored linguistic hearing or making linguistic sounds or learning languages or using languages will tend to have survived and that over time some highly specialized perceptual and sound production and linguistic-cum-cognitive abilities would have evolved. A demonstration of this would help us to understand why studying other primates is a waste of time if it is done by way of trying to understand the development of human communication.

---

Tags: primate communication, evolution of language, language and cognition, pragmatics