- In the sci-fi novel "The Three Body Problem", author Cixin Liu invents a race of aliens that is incapable of lying.

The Trisolarans as they're called, broadcast every thought they have out into the world, like always on radio transmitters.

When the Trisolarans come into contact with humans, they're horrified by our ability to intentionally deceive one another.

In linguistic terms, Liu collapses the distinctions between locution, the actual words we use, illocution, the intent behind those words, and perlocution, the effect of those words on others.

For the Trisolarans, there's no boundary separating intent and speech.

They can read each other like books.

But for us humans, communication takes place outside of the mind.

We can think one thing and say another.

Most people would call that lying and say that it's wrong, but when you really dig into the linguistics of it, "what is a lie?"

becomes a difficult question to answer.

And though we have a long history of moral judgment around it, the propensity to lie seems to be built into our DNA.

Little kids start doing it basically as soon as they realize they can.

But what really counts as a lie, and what's going on in our heads when we tell one?

Is it a sin or a superpower?

I'm Dr.

Erica Brozovsky, and this is "Other Words".

(bright music) - [Announcer] "Other Words".

- Some would say that lying is common throughout the animal kingdom.

I mean, look at this shifty bug disguised as a stick.

Or this two-faced cuttlefish pretending to be a female to hoodwink competitors.

Or this duplicitous bird faking a broken wing to lure predators away from her nest.

While the purpose of all these behaviors is indeed to deceive, few people would call them lies because they lack conscious intentionality.

First-order animal deception like passive camouflage requires no intentionality at all.

The stick insect doesn't even know that he looks like a stick.

Second-order animal deception is still pre-programmed, but only switches on in the presence of another, like an adversary.

The cuttlefish doesn't know why he exhibits female patterns when he senses another male.

He just does.

And third-order deception may be learned rather than instinctual, but still lacks a full understanding.

The mother bird may have figured out through trial and error that predators are more likely to follow her when she crooks her wing, but she doesn't really know why it works.

Fourth-order deception is pretty much the sole province of humans and maybe some apes, because it requires a robust theory of mind.

That is the cognitive awareness that other individuals have thoughts and beliefs that are separate from yours, and that those beliefs can be false.

Lying is a fourth order deception for the most part, and a simplified description goes something like this.

X believes that P is false.

X states P in order to make Y believe that P is true.

Note that in this formulation, whether or not P is actually false is irrelevant.

Most people would agree that even if P turns out to be true, as long as the X believes it as false, it still counts as lying.

So if you tell your friend that you're all out of gum because you wanted to keep the last stick for yourself, but then later find out that the package is indeed empty, well, you're still a liar.

But not every statement that fits this formula is considered morally wrong.

You might tell a coworker that her new haircut looks, "great, really great, like 'Amelie'", when what you really thought was, "No Country for Old Men".

These are pro-social lies told to save face, protect feelings, and generally grease the wheels of social interaction.

Not only are they considered acceptable, but many would say it would be rude not to tell them.

The moral distinction seems to be whether telling a lie benefits yourself or someone else.

But pro-social lies aren't always so easy to categorize.

When you tell a friend that you love her singing, yeah, you're trying not to hurt her feelings, but aren't you also avoiding discomfort for yourself?

After all, it might benefit her to hear what you really think, but that would be an awkward conversation, so it's easier for you to just lie.

This formula also doesn't account for the fact that lies don't have to only be declarative sentences.

In other words, you don't have to make an assertion to lie.

Let's say your roommate bursts into the house upset and says, "someone sideswiped my car while it was parked out front!"

That someone was in fact you, but you respond to her declarative sentence with an exclamatory, "oh no!"

and interrogative.

"When did that happen?"

Technically, no false assertions were made, and yet most people consider these to be lies.

It may feel obvious to you why, but linguists like clear cut definitions, and some have sought them in the work of Paul Grice, a philosopher of language who came up with four Maxims of Conversation: quality, quantity, relation, and manner.

Basically, these are assumed expectations that we have of whoever we're engaged in conversation with.

The Maxim of Quality assumes that the speaker will try to be truthful.

Do not say something you know to be false, or that you don't have enough evidence to assert.

The Maxim of Quantity assumes that the speaker will give the appropriate amount of information, no more and no less.

If I ask you what you had for breakfast and you say bacon and eggs, I'll assume you didn't also have pancakes, but I don't need to know whether you put pepper on your eggs or what brand of bacon it was.

The Maxim of Relation assumes the speaker will only include relevant information.

If I ask you how to cook a turkey and you mention that eggs take nine minutes to boil, you violated the maximum violation.

And the Maxim of Manner assumes the speaker will try to be as clear as possible, which includes not using obscure words, repeating oneself, or recounting events out of order.

Together, these maxims portray speech as not a solitary act, but a cooperative effort between individuals.

The literal words themselves are not enough to convey the speaker's true meaning.

Only when they interact with a complex set of assumptions and expectations do they produce implied meanings, which Grice calls implicatures.

For example, if Bert says to his wife, "I ate some of the lasagna", she would assume the implicature that he didn't eat all of the lasagna, even though he didn't explicitly say so.

By asking your roommate, "when did that happen?"

You're creating the implicature that you don't know darn well when it happened.

Lying by implicature.

A famous case of lying by implicature occurred in 1966, when a film producer named Samuel Bronston was questioned under oath during a bankruptcy examination.

"Do you have any bank accounts in Swiss banks, Mr.

Bronston?"

"No, sir."

"Have you ever?"

"The company had an account there for about six months in Zurich."

As it turns out, Bronson had indeed had a Swiss bank account, an admission he carefully dodged with answers that were technically true, but violated the Grice Maxims.

"Do you have any Swiss bank accounts?"

Well, no.

At that moment he didn't, but by leaving out crucial information, he violated the Maxim of Quantity.

When the lawyer followed up with, "have you ever?"

He replied that the company did, which violated the Maxim of Relation by introducing irrelevant information.

When the truth came out, Bronson was accused and found guilty of perjury, but he appealed the decision all the way to the Supreme Court, which in 1973 reversed the conviction, ruling that though he was clearly evasive, he didn't say anything factually untrue.

It was the lawyer's responsibility to extract a declarative sentence that could be proved false.

In other words, implicatures cannot be considered perjury.

This may all sound over complicated, but when you think about it, lying, or at least lying well, is really complicated.

Not only do you have to keep track of what's true and what's not and what lies you already told, but also what your victim knows or doesn't know, and what you are supposed to know or not know.

All expressed in a dizzying array of subconscious implicatures.

no wonder they call it a web of lies.

That can explain why some studies have found liars to exhibit tells like hesitations, grammatical errors in abnormal speech rates, almost like a computer struggling under a high processing load.

Other studies suggest that liars tend to make shorter statements with fewer unique words, which makes sense, since the more you say, the more opportunities to get found out.

They also tend to avoid first person pronouns like I and me, as well as emotional words, perhaps to distance themselves and their feelings from the act of lying.

Despite these findings, in reality, it's notoriously difficult to determine with consistent accuracy whether someone is lying, because human behavior is unique and complicated.

Even technology like polygraph machines can only measure certain changes in a person's body that might indicate nervousness or guilt.

But there could be a lot of perfectly honest reasons for that.

You'd think that the absence of these face-to-face cues online would increase the frequency of our lies in emails, texts, and social media posts.

But according to some studies, this doesn't seem to be the case.

People tend to lie online about as much as they do IRL, as long as it's not anonymous.

While the ability to meticulously craft your fibs in private probably makes it easier, that temptation may be offset by knowing that your lives might be recorded permanently, and in many cases available for anyone to see.

It kind of reminds me of how the Trisolarans communicate.

Instead of the delicate face-to-face dance that Paul Grice calls a cooperative effort, we plaster our feeds with every little thought that comes to mind.

Maybe that's why going on social media can sometimes feel a bit like a Trisolaran Thanksgiving.