Watch Your Language

What our words reveal about our minds, but not about the world.

Steven Pinker is a man stuffed with thoughts and gifted with language, a combination that has won him an unusually wide audience. In 1995, he published The Language Instinct, which was not just a best-selling Pulitzer Prize nominee. It was a best-selling Pulitzer Prize nominee about linguistics that was the first really meaty guide to how language actually works. Three similarly successful books followed. In one more about linguistics and two about human nature, Pinker gradually emerged as a polymath pioneer in the field of evolutionary psychology, which plenty of scientists had dismissed as mere storytelling but has thrived, thanks largely to his efforts. His latest book completes two arcs, one for the language books and one for the psychology books. There’s more stuff than ever in The Stuff of Thought, and once again, Pinker isn’t shy. In devoting almost 500 pages to the way we make meaning, he confirms his place as the lead interpreter of a field long dominated by Noam Chomsky. He does it with signature clarity.

Under Chomsky’s sway for four decades, most of linguistics and related sciences focused on the structure and rules of language, at the expense of meaning. Pinker defies the old order, and he does it, fittingly, without letting rhetoric get in the way as he guides readers through the radically expanded landscape of work on language and thought in cognitive science. His book is a vast explainer, built out of his own research and the work of many careful researchers and scholars who have received little attention beyond the academic fields of semantics and experimental psychology. In one chapter, Pinker examines whether the particular language you speak influences the way you think. The idea that it does has generated a lot of attention (as well as irritation and indignation) in cognitive science over the years. Laying out the debate as he sees it, Pinker concludes that it does not, at least not in any dramatic or important way. For most of the book, however, he flips the terms, investigating a different relationship that is equally deserving of the spotlight: the way thought underpins language.

Of course, there are entirely obvious ways that thinking shapes language. The words of a language reflect the concerns of its culture, and when we use language, we usually try to say what we think. We signal in various ways what we think about what we think, and even the mistakes we make expose the way our mind works. This is all interesting stuff, but with his characteristic flair (“verbivores,” “whimperatives,” “malefactive verbs,” “momentaneous events”), Pinker explores a less day-to-day connection. He carefully picks language apart to reveal the conceptual scaffolds and preoccupations that underlie it, including pervasive beliefs about the workings of time, space, and motion, as well as the human body. Once the struts are exposed, Pinker confronts us with a surprising conclusion: Though they shape our language, these mental scaffolds have little to do with how the world really works.

Getting at the struts isn’t easy. For linguists, this kind of activity typically begins not with a grand ambition to expose the anthropic universe but instead with a small, nagging problem. Why, for example, can you say “load hay onto the wagon” and “load the wagon with hay” (which seem to mean pretty much the same thing), and yet even though you can say “toss hay onto the wagon,” you cannot say “toss the wagon with hay”? On noticing this oddity, scholars soon discovered that it was spread throughout language. Lots of verbs will happily tolerate the structural flip, but for no obvious reason, others refuse to be manipulated in this way. Certainly, language can be a messy place. Was this merely a chance variation, the result of words having individual preferences?

No one in science likes to go with randomness as a first, or a last, guess, but it took a very long time for linguists to show that there was method beneath the mess. It turns out that the goodness of the fit between the verb and the construction is not simply an individual difference in the way words are used, but is fundamentally underwritten by broad, unspoken categories of meaning—in this case, the way we think about things like movement and space.

Only one of these verbs works in both sentences because, despite appearances, they don’t really mean the same thing. In fact, they subtly frame events in very different ways. When you “load, or toss, hay onto a wagon,” all that’s being described is that some things are moved to another place. There may be lots of hay in the wagon or there may not. The sentence doesn’t really say. When you reverse the structure and “verb a wagon with hay,” the implication is that it is completely, not partially, filled with hay. The wagon is qualitatively changed by the action of the verb. A full wagon seems logical if it has been loaded with something, but the mind—and language—balk at the idea of a wagon changing state so completely because things have been tossed on it.

What’s revealing about the clashes and matches between a verb and a frame is not just that the frames are freighted with ultimately different meaning, but also that the verbs fall into natural groupings based on broad categories of meaning. Linguists recently discovered, in part because of the way a verb fits or doesn’t fit into different kinds of sentences, that many verbs hang out in invisible cliques, again based on concepts like space, or force, or motion. Of at least 85 such verb sets in English, one involves what happens when a collection of objects is distributed over a surface (blot, bombard, dapple, riddle, speckle). In another group, the verbs all describe what happens when little bits of stuff are sent in every direction (bestrew, scatter, seed, sow, strew). Yet another lot describes something that is being expelled from inside something else (emit, excrete, expectorate, secrete, spew, spit, etc.).

It is remarkable that no one is ever taught that these classes exist, let alone what they mean. Yet when a child learns the verbs of a language, they implicitly learn about the way verbs huddle together. As Pinker points out, verbs could be grouped for all sorts of meaning, for example, based on whether they describe things that look the same, feel the same, or smell the same. But they are not; what their groupings reveal is a distinctly and universally human fixation on different kinds of motion, how force is applied, how time gets parceled up, and how states change—this is the stuff of thought.

Pinker walks his readers with a firm and friendly hand through many finely detailed examples, picking out a word or exposing a metaphor, turning it over, trying it out in different contexts, and exposing its internal mechanism. Again and again, seemingly inconsequential quirks of language reveal the same cosmic preoccupations. In addition to physical objects and the laws that move them, the way we carve up the universe includes a basic taxonomy of human vs. nonhuman things, animate vs. inanimate objects, discrete objects vs. continuous stuff, and flexible vs. rigid things. Our language is also shaped by a timeline along which events are bounded or unbounded. Also fundamental is the idea of a goal, as well as an important distinction between means and ends.

Not all of these meanings appear in all languages, but some large set of them do, suggesting pretty strongly that they are fundamental to how human beings think about the world, not just how we talk about it. Pinker runs through many different experiments that show how basic some of these concepts are to human thinking, independent of language use.

Yet although they are essential to thought, the principles and distinctions revealed by language are not fundamental to the world and how it works. Time is not a line; objects are not bounded or unbounded in the ways that we construe them. Nor does the world break down in any clean way into, for example, things that are human vs. things that are nonhuman. The stuff of human thought is wrong. Well, maybe not wrong, but it’s not right, either. Rather, as Pinker shows, our default ideas about how the world is partitioned are a “cognitive lens” with which to view all the, you know, things around us. After all, he says, “though we can never directly know the world, it’s not as if one could know the world without some kind of mind.” The trick, of course, is to be aware of the lens at the same time you are looking through it.

How then should we view language? If it’s not the case that language determines how we see the world, and it’s not true that the world itself determines language, what is it? If you’re adept enough with it, then language is a paradox: revealing the universal concerns of our species, while at the same time enabling us to see, at least a little bit, beyond them.