Key linguistic cues can help reveal dishonesty during business negotiations, whether it’s a flat-out lie or a deliberate omission of key information, according to research by Lyn M. Van Swol, Michael T. Braun, and Deepak Malhotra.
Want to know if someone’s lying to you? Telltale signs may include running of the mouth, an excessive use of third-person pronouns, and an increase in profanity.
These are among the findings of a recent experimental study that delves into the language of deception, detailed in the paper Evidence for the Pinocchio Effect: Linguistic Differences Between Lies, Deception by Omissions, and Truths, which was published in the journal Discourse Processes. Asked why the topic of deception is important to business research, negotiation expert Deepak Malhotra responds wryly: “As it turns out, some people will lie and cheat in business!”
Malhotra, the Eli Goldston Professor of Business Administration at Harvard Business School, coauthored the paper with Associate Professor Lyn M. Van Swol and doctoral candidate Michael T. Braun, both from the University of Wisconsin?Madison. “Most people admit to having lied in negotiations, and everyone believes they’ve been lied to in these contexts,” Malhotra says. “We may be able to improve the situation if we can equip people to detect and deter the unethical behavior of others.”
?Just like Pinocchio?s nose, the number of words grew along with the lie?
“Evidence for the Pinocchio Effect” fills a key gap in the field of deception research, says Van Swol, the study’s lead author. Previous studies have examined the linguistic differences between lies and truthful statements. But this one goes a step further to consider the differences between flat-out lying and so-called deception by omission?that is, the willful avoidance of divulging important information, either by changing the subject or by saying as little as possible.
The ultimatum game
To garner a sample of truth tellers, liars, and deceivers by omission, the researchers recruited 104 participants to play the ultimatum game, a popular tool among experimental economists. In the traditional version of the game, one player (the allocator) receives a sum of money and proposes how to divvy it up with a partner (the receiver). The receiver has the option of either accepting the proposed split or refusing the allocator’s proposal?in which case neither player gets any of the money. Because receivers will often reject offers they perceive as unfair, leaving both parties with nothing, it behooves the allocator to offer an amount that will be deemed fair by the receiver. In many instances, allocators choose to share half, Malhotra says.
For the purposes of the deception experiment, the rules of the ultimatum game differed from the traditional version in three ways. First, in this version, the allocator received an endowment of either $30 or $5 to share with the receiver. The receiver had no way of verifying how much money the allocator had been given, information which the allocator was not required to divulge. Hence, an allocator could conceivably give the receiver $2 and keep $28, and the receiver would be none the wiser, perhaps assuming only $5 was in play. The second change was that if the receiver rejected the allocator’s offer he or she would receive a default amount of $7.50 (or $1.25)?whereas the allocator would get no money at all.
Finally, each game included two minutes of videotaped conversation in which the receiver could grill the allocator with questions, prior to deciding whether to accept or reject the offer. This provided ample opportunity for the allocator to tell the truth about the money, lie, or try to avoid the subject altogether. “We wanted to create a situation where people could choose to lie or not lie, and it would happen naturally,” Van Swol says.
Ultimately, the receiver had to decide whether the proposed allocation was fair and honest, based only on a conversation with the allocator. Thus, it behooved the allocator to be either a fair person or a good liar.
As it turned out, 70 percent of the allocators were honest, telling the receivers the true amount of the endowment and/or offering them at least half of the pot. The remaining 30 percent of allocators were classified either as liars (meaning they flat-out lied about the amount of the endowment) or as deceivers by omission (meaning they evaded questions about the amount of the endowment, but ultimately offered the receiver less than half).
After a graduate student transcribed all the allocator/receiver conversations, the researchers carefully analyzed the linguistic content, comparing the truth tellers against the liars and deceivers in order to suss out cues for deception. They looked for both strategic and nonstrategic language cues.
“A strategic cue is a conscious strategy to reduce the likelihood of the deception being detected,” Van Swol explains, “whereas a nonstrategic cue is an emotional response, and people aren’t usually aware that they’re doing it.”
Key findings: word count, profanity, and pronouns
In terms of strategic cues, the researchers discovered the following:
- Bald-faced liars tended to use many more words during the ultimatum game than did truth tellers, presumably in an attempt to win over suspicious receivers. Van Swol dubbed this “the Pinocchio effect.” “Just like Pinocchio’s nose, the number of words grew along with the lie,” she says.
- Allocators who engaged in deception by omission, on the other hand, used fewer words and shorter sentences than truth tellers.
Among the findings related to nonstrategic cues:
- On average, liars used more swear words than did truth tellers?especially in cases where the recipients voiced suspicion about the true amount of the endowment. “We think this may be due to the fact that it takes a lot of cognitive energy to lie,” Van Swol says. “Using so much of your brain to lie may make it hard to monitor yourself in other areas.”
- Liars used far more third-person pronouns than truth tellers or omitters. “This is a way of distancing themselves from and avoiding ownership of the lie,” Van Swol explains.
- Liars spoke in more complex sentences than either omitters or truth tellers.
The researchers also examined when and whether the receivers trusted the allocators?noting instances when receivers voiced doubts about the allocators’ statements, and correlating the various linguistic cues with the accuracy of the receivers’ suspicions. They also noted instances in which receivers showed no suspicion toward deceivers.
On average, receivers tended to trust the bald-faced liars far more than they trusted the allocators who tried to deceive by omission. In short, relative silence garnered more suspicion than flat-out falsehoods. “It turns out that omission may be a terrible deception strategy,” Van Swol says. “In terms of succeeding at the deception, it was more effective to outright lie. It’s a more Machiavellian strategy, but it’s more successful.”
Possible applications
In the latest phase of their research, the team is investigating the linguistic differences between lying in person and lying via email. Results regarding the latter may be increasingly useful as a larger portion of business is now being conducted via email, and such communications leave a transcript that can be analyzed carefully?and at leisure?by suspicious counterparts. “People detect lies better over the computer than they do face-to-face,” Van Swol says.
That said, the researchers are quick to emphasize that linguistic cues are most definitely not a foolproof method of detecting lies, even among those who are trained to look out for them.
“This is early stage research,” Malhotra says. “As with any such work, it would be a mistake to take the findings as gospel and apply them too strictly. Rather, the factors we find to be associated with lies and deception are perhaps most useful as warning signs that should simply prompt greater vigilance and further investigation regarding the veracity of the people with whom we are dealing.”
About the author
Carmen Nobel is senior editor of Harvard Business School Working Knowledge.