How do our words relate to our thoughts? What does language tell us about human nature? What are we doing, language-wise, when we swear, use innuendo or name our babies? Harvard professor Steven Pinker examines these questions -- and much more -- in his terrific new book, The Stuff of Thought. We had a fascinating conversation with him about his research into language and cognition:

VT: How is language a window into human nature?

Steven: We share our knowledge -- and our understanding of the world -- through language every time we teach, every time we instruct, every time we give advice. We negotiate our social relationships through language through both what we say and what we hope people will infer by reading between the lines. We also express our emotions through language, especially when we swear. We reveal something about our tastes and identities when we apply names to things, most notably to our children. These are just a few of the ways in which language can shine a headlight on human nature.

VT: Do even parts of speech reveal something about us?

Steven: I argue that verbs illuminate our notion of causality. Nouns say something about the way we construe substance and matter. Prepositions tap into our intuitions about space and force, and tense offers a window into our conception of time. Moreover, the semantics of these parts of speech aren't particular to language itself but tap into the way in which we organize and manage our affairs. So a similar concept of causation that is expressed in our verbs is the one that we intuitively apply to moral and legal responsibility.

VT: Can you give us an example?

Steven: With causative verbs, verbs like "to open," "to melt," "to bounce," "to bend" and so on, the causation has to be a direct and immediate for the verb to apply. So if I slide down a dimmer switch and the lights go dim, you could say "I dimmed the lights." But if I turn on my toaster and the lights dim, people don't say that that's dimming the lights. If I reach for a doorknob and push the door open that's called "opening the door." But if I open the window and a breeze comes in and the door blows open, I don't call that sequence opening the door.

The causation has to be immediate, without smaller intervening links in the causal chain. That is also a requirement in our allocation of legal and moral responsibility. When there are too many links in the causal chain, we tend not to find the agent responsible, or at least have our doubts. For example, when an assassin shot President James Garfield in 1881, the bullets weren't fatal. But because of the harebrained medical treatment he received at the time, Garfield lingered on his death bed for three months until he died of infection -- probably from the doctors probing his wounds with unwashed hands.

During his trial, the assassin said in court, "the doctors killed him, I just shot him." The semantics of "kill" coincide with our assessment of legal and moral responsibility. That is because of the intervening links in the causal chain, legitimate doubts could arise as to whether this man was guilty of murder or only attempted murder.

VT: In your book you talk about innuendo and how that comes into play when we communicate.

Steven: A lot of our speech is conveyed by innuendo instead of directly. Every time we make a polite request like, "if you could pass the salt that would be awesome," we are not blurting out what we mean to say, which is, "give me the salt." We count on our listener to see through that nonsensical statement to infer that we are really issuing an imperative.

The reason that we do this is that we don't like to boss people around, at least not strangers or polite acquaintances, which is what you do when you issue a bald imperative. By veiling the request in a conditional, you're not presupposing the listener's compliance, as if they were an underling or flunky. You're signaling to your listener that your relationship is not one of a dominant and a subordinate. At the same time, you really do want the damn salt.

VT: Right.

Steven: You can eat your cake and have it, too. Other examples are a bit more complicated: There is the classic sexual come-on, "Would you like to come up and see my etchings?" And the stereotypical threat, "Nice store you got there, would be a real shame if something happened to it." And the veiled bribe, like, "Gee officer is there some way to take care of the ticket here without going to court?"

What we have here is framing a message so that a willing partner can sniff out the proposition and accept it, whereas an unwilling partner can't nail you for it. The dishonest cop who is bribable can sense the bribe underneath the innuendo, while the scrupulous cop who would otherwise arrest you for bribery couldn't make the charge stick in court.

VT: These innuendos make sense within the context of our particular culture. But would other English-speaking cultures like England understand them? Does culture shape language, or the other way around?

Steven: We see culture shaping the language because languages never come out of the blue. It's not as if Martians came to Earth and handed out grammar and etiquette books. So they must arise from people trying to lubricate their interactions as they manage their affairs. For example, there are differences among cultures in the amount of politeness that is considered to be the baseline. Deviations from that are considered polite or rude being relative to the cultural baseline. And that, in turn, is systematically related to properties of the culture.

For example, hierarchal cultures tend to have more of the kind of politeness where you express deference to someone else. Egalitarian cultures have more of the kind of politeness where you include people in your circle. Think of a waitress at the coffee shop calling you "honey" or other bogus terms of endearment. That's a different kind of politeness, one that's more prevalent in a working-class setting.

All cultures have some kind of indirectness and politeness and euphemism, because all cultures are faced with a problem that language has to do two things at once. It has to convey a message like a sexual come-on or a bribe or an imperative or a threat -- whatever it is you want to get across. But at the same time, it's got to negotiate and maintain a kind of relationship that, by the mere act of issuing an imperative or a sexual proposition, you are in danger of changing. Understanding the literal meaning, which we see through, as well as the underlying content, is a means of attaining both goals at once. We use language at one level to convey the information, and at another level to maintain the relationship.

VT: At the beginning of our conversation you spoke about how we shine a headlight on human nature by the way we name our children. What do you mean by that?

Steven: In cultures where parents have the complete freedom in naming a child (unlike cultures, say, where you have to name your child after a saint or a dead relative), people try to be a bit distinctive, but not too distinctive. On the one hand, they follow Sam Goldwyn's advice: "Don't name your son "William"; every Tom, Dick and Harry is named William." At the same time, they don't want to give their child such an outrageous name that they mark the child as coming from a family of misfits or greenhorns (unless one of them is a rock star or a famous actress -- who feel they can make their own rules).

VT: They can get away with it.

Steven: They can get away with it because of their status is so high. But the rest of us want to be just a little bit distinctive, not too distinctive. We also want a name that does not sound too geriatric, so we tend to avoid the names of our grandparents' generation. The problem is that when everyone is trying to be a little bit fresh and moderately distinctive they're in danger of being moderately distinctive in the same way.

VT: That could lead to problems.

Steven: Parents can think to themselves, "Oh, Chloë is a nice sounding name, and I'm giving my daughter that name because it sounds right to me." But it sounds right to a lot of other people at the same time, too. That's why parents have the experience of taking their Chloë to the daycare center and being shocked to find that there are three others who are named Chloë.

VT: But names don't stay static. It seems that what people perceive as "somewhat adventurous" evolves over time.

Steven: That's right. As names get older, and hence less prestigious, they also become less apt for the new generation. You don't want to name your daughter "Mildred" or "Ethel" or "Hortence" because it conjures up the image of a lady in an old age home. That's not the way you wanted to think of your newborn baby.

(Photo credit: Rebecca Goldstein)