How Baseball Explains The Nature Of Language : 13.7: Cosmos And Culture Language is more like baseball than most linguists would be prepared to admit. This has big implications for what a science of linguistics can hope to achieve.
NPR logo How Baseball Explains The Nature Of Language

How Baseball Explains The Nature Of Language

Whether someone is safe or out can only be understood within baseball. Stephen Dunn/Getty Images North America hide caption

toggle caption
Stephen Dunn/Getty Images North America

Mine is an unusual name. I have come to learn that it is isn’t possible for me to tell someone my name over the telephone.

I don’t mean that it isn’t easy.

I mean that it isn’t possible.

Nothing I say, no sound I emit, can suffice to communicate to the person on the other end of the line what my name is. Even spelling it out only works if I make use of a conventionalized system for naming the letters (such as “Alpha Lima Victor Alpha”).

We confront what I will call the Paradox of Speech (a close cousin to Plato’s Paradox of the Meno). To understand my words, you need already to know what I am saying. But if you already know what I am saying, there is no need to listen to my words.

How then is communication by speech even possible?

The solution to the paradox — understanding how speech is even possible — comes when we give up what we might have thought was a bedrock fact about language: the idea that speech is meaningful sound.

This won’t come as a surprise to linguists. It has long been known that there is no one-one mapping between physical sound (acoustic phenomena picked out by spectral properties) and speech sound. The k you hear now and the p you hear later may be realized in identical acoustic events; and radically distinct acoustic events can be heard as one and the same speech sound.

Does this mean that there is no difference between pong and kong? Is our conviction that these different words are somehow an illusion? Certainly not. But the difference isn’t one we can understand in the vocabulary of physics. It isn’t a difference in how these words sound.

If words aren’t sounds, then what are they? One possibility — this is the idea I want to explore here — is that words are like home runs and that linguistic facts are like facts about baseball.

Baseball is defined by a system of rules. The rules tell you what a home run is, and they provide criteria for deciding, say, when a ball is foul. The rules of baseball jointly entail that  there are unassisted triple plays but there are no touchdowns

To learn baseball is to learn much more than a system of rules, however. Players learn to field, run the bases and hit. And whether you are a player or a fan, when you learn baseball, you acquire a whole suite of new perceptual skills; the botched double-play, the right situation to put on a hit-and-run, the balk, these phenomena pop out perceptually for the knowledgeable observer. Even more remarkably, when you are initiated into the baseball world, you learn to care about such things as stolen bases and pick-off attempts.

Far more than merely learning a system of rules, to learn baseball is to come to be able to see and feel and be motivated in ways that are meaningless to strangers of the game. Baseball is more than a system of rules, it is a practice.

Have you ever spent time with baseball people? If you have, then you will know that the thing baseball folk do more than anything else, even during a game, is talk about baseball. I don’t just mean that they dispute plays. Baseball people are endlessly engaged in evaluating play and in developing theories — even mathematical and statistical theories — to justify or support one or another means of evaluation. They ask questions like, how much value does a player add to a team? Defensively? Offensively? And they ask the meta-question: By what methods can we decide such issues? Baseball people are concerned not just with how you play, but with the very question of how one ought to play baseball.

Let us say that a human social practice is baseball-like when the practice includes as a part of itself the activity of reflecting on, theorizing about, and critically evaluating the very practice itself. Baseball is baseball-like in this sense. And so are many, perhaps all, social practices. Art and the law provide clear examples. But not everything is baseball-like. Digestion, for example, is not. However some people like to opine about how one ought to eat and how one ought not to eat to improve one’s digestive performance; it remains the case that there is the first-order activity, digestion, and the second-order and distinct activity of talking and thinking about digestion. This sharp division into the first order and the second order, or into the practice and the ideologies about the practice, is not available to us when a practice is baseball-like.

What makes baseball-like practices interesting is the fact that, to put the point as a philosopher might, their ontologies are practice-relative. Are there home runs? Of course. But you can’t look outside of the practice of baseball for an understanding of what home runs are. In particular, you can’t turn to neuroscience, psychology, economics, biology, physics or chemistry for such an explanation. A home run is, yes. But what is a home run? It is a structure in the space of baseball. A home run is a practical structure.

Let us turn our attention back to language. As with baseball, so with language. Language is a system of rules and representations, but to learn a language is more than learning rules.

It is to acquire a range of practical and also as perceptual skills, and it is to come to have a range of feelings, attitudes and beliefs about language itself. Hearing speech is not just a matter of registering and categorizing sounds; this is what we noticed at the outset. It’s a way of paying attention to what people are doing — to what they say, to how they respond, to what they are paying attention to, and generally, to what’s going on.

A sensitivity to speech is, really, a sensitivity to human action.

And crucially, just as speech doesn’t reduce to sound, action doesn’t reduce to movement. I reach for the cup. One movement, but potentially multiple actions. In one case I am putting the cup away, let’s say. In another case I am reaching to take a sip. These differences in action despite sameness of movement are readily apparent.

From this standpoint we can appreciate that of course we can’t hear novel words over the telephone. The very fact of their novelty removes the background context that makes perceptual sensitivity to the corresponding actions possible. What makes it possible to talk on the telephone is the fact that most of the time we can rely on the presence of a shared context or background. Everything flows smoothly until we get to the point where what I am doing now — articulating my unusual name, say — can’t be read off the background we take for granted.

But most striking of all — in part because this runs against what linguists have been teaching for nearly a century — language is baseball-like in the sense defined above. To learn a language is to learn not just to do something, it is to become all at once -- a critic, a teacher and an innovator. Children don’t merely learn what words mean. They learn to ask what words mean, and, right away, to say what they mean, to tell or teach another, and they learn to correct and upbraid another’s usage.

The idea of a first-order linguistic practice that is unaffected by ideas about how one ought to speak is as strange and remote as the idea of a species of people who play baseball without ever talking or thinking about how one ought to play baseball. It would be a form of robot-like existence, certainly not a recognizably human one. Where there is understanding, there is the possibility of misunderstanding. And clearing up misunderstandings belong to the practice of communicating too.

Linguistic science wants to be descriptive. Because linguistic performance itself is obviously the product of social forces and top-down prejudices, not to mention all manner of errors, slips of the tongue, etc., what linguists seek to describe is not what people actually do and say but, as linguists put it, our underlying linguistic competence.

It is this competence that enables us to perceive, for example, that “every boy loves a person” is grammatical whereas “every boy love a person” is not. And it is this competence that allows us to understand, as it were unbiased by theory or ideology, that “a” and “person” go together in this previous sentence, in a way that “loves” and “a” do not; they form a significant structural unit, a noun phrase.

Now it is surely the case that we are sensitive to these kinds of differences and these kinds of facts. But this is “inside baseball!” Asking about grammaticality, or how words group together, is like calling balls and strikes. We can do this. We can do this well. We can do it spontaneously. But we do this within the practice. “Grammatical” and “ungrammatical” are terms of appraisal; we deploy them within language and they can’t be used to demarcate the limits of language (any more than “witty” and “boring” could be so used). We are so deeply embedded within and at home in the language world (compare: the baseball world) that we find it difficult to believe in the practice relativity of our convictions and commitments.

Linguists insist that language and writing are very different. The latter is a 10,000 year old bit of social technology. It is obviously baseball-like; the use of written language can’t be separated from our attitudes about how one ought to write. But language, speech, real language, this is thought to be part of our species-specific genetic endowment; you can no more teach or correct language than you can teach or correct the growth of limbs or digestion.

I want to take seriously the possibility that the linguists have it wrong.

Language and writing are alike not merely in that both are baseball-like; their connection runs deeper than this. First, speech, like writing, is also a technology; or better, like writing, and baseball, it is a technique. Speech happens to be older than writing by maybe 70,000 years. Second, writing has powerfully and utterly transformed our conception of language, so much so that we find it almost impossible to think of language other than in the terms afforded by writing, e.g., we think of sentences as formal strings composed of a special kind of formal string, the word. But the idea that language is made up of formal strings and that it can be studied as a system of rules governing the generation of formal strings is about as plausible as the idea that dancing is a formal string that can be spelled and written and evaluated for its hidden structures.

The deeper you go into physics and chemistry, the stranger the world gets. The objects of every day life disappear. This is precisely not the case with linguistics. Linguistics begins with a pre-theoretic conception of language as consisting in words and sentences (a conception read off of our use of writing to represent language) and this is also where it ends.

What this reveals is not that there is anything suspect about the category of the word, or the sentence. What it reveals, rather, is that linguistics is not really in the business of uncovering hidden realities in the way that chemistry and physics are. The linguist uncovers, or helps us uncover, how we think about our practice. The linguist works is in the domain of style.

And this is at it should be.

I once met a woman who was trilingual. She spoke Fulani at home, Hausa in the market place, and French at school and at the post office (and with me). She gave me the strangest look when I asked her how to translate a sentence from French to Fulani. It took me a long time to finally understand why should found the request strange. For her, the task of translating from French to Fulani, was a bit like trying to understand how to do the twist in classical ballet, or how to score a touchdown in baseball. The request barely made sense and it had even less point. She implicitly understood, as I did not, that language is not an all-purpose symbolic system governed by rules and representations.

Language is like baseball.