Some Simple Ways to Spot the Language of a Liar
A new study on the language of lying shows how to figure out who to believe.
Posted Mar 24, 2020
In turbulent times, when the news media bombards you with information from every direction, it’s hard to know who is telling the truth. In the case of the coronavirus pandemic, knowing who to believe can literally mean the difference between life and death.
You may be told by one government official that successful treatment is just around the corner, while others say it will take months to find a cure. A local agency tells you to stay at home, but another leader says it’s alright to go out and visit with friends and family.
In your own personal life, you may be faced with potential liars all the time. A close friend sends you a sweater as a birthday gift and, after you excitedly open the package, realize that it’s something you could never actually bring yourself to wear. The color is completely wrong and you find the fabric itchy. You obviously have to acknowledge the gift, and so decide that the easiest way is through a text. What will you say?
According to a new study on the language of everyday deception in text messaging by Ball State University’s Thomas Holtgraves and Elizabeth Jenkins (2020), you’re in a fairly common predicament. People send deceptive messages all the time, the authors argue, using “small and easy lies” (p. 1), or what you may think of as “white lies.” Much as the case with the ill-fated sweater, people don’t tell lies, according to the authors, due to malicious intent. Instead, people lie to protect the impression they create on others or to preserve someone else’s feelings.
You might know the truth vs. reality in the case of your own lies, white or otherwise, but how well can you tell who’s trying to manipulate your impression of them? Unlike face-to-face interactions, online or text messaging liars don’t have to worry about their non-verbal cues such as eye gaze, pauses, or body language. Instead, lies will come through in their words. The Ball State researchers began to address this question by first examining the naturally occurring lies that live in people’s text message histories.
Using a sample of 65 college students (79% female), Holtgraves and Jenkins invited participants to come to the lab with their cellphones in hand. From their most recent texts, participants provided the last 10 non-deceptive text messages they had sent, the last 10 deceptive messages, and the last 10 deceptive messages sent to them by others.
Since there was no way of knowing the “truth” of those received messages, participants had to decide on their own whether the texts seemed authentic, based on their knowledge of that person. In other words, if you wanted to tell your gift-giving friend you loved the sweater that you hated, you would have to take into account the history the two of you share, which could make your deception detectable.
These honest and deceptive messages then became the basis for the next phase of the study. A new group of participants rated each text message generated by the original participants along with scales measuring truthfulness, believability, and sense of confidence in the truthfulness judgment. The findings showed that these perceivers in fact were reasonably good at the task, providing higher truthfulness and believability ratings for the non-deceptive texts compared to the lies.
Digging into what gave the lies away, the authors found that truthful texts used fewer negations (e.g. “can’t,” “don’t”), and negative emotions (e.g. “afraid,” “warn”) and were less likely to include first-person pronouns. Truthful texts also were more logical and assertive (higher in “clout”). In other words, when you’re sounding tentative or if you’re rambling in your text, it may seem as if you’re trying to cover something up, and the recipient will be able to sniff this out.
Although there were no actual differences in word length between deceptive and truthful texts, perceivers were more likely to regard a text as genuine if it involved more words. This finding makes sense in that it’s generally easier to lie in a short message than a long one. All you have to do is say “thanks, I love it” if you’re trying to disguise your feelings about the sweater but if you really like it, you’ll be more likely to say what it is you like about the present. This feature of short texts seemed to influence how truthful they seemed, even if they weren't lies at all.
The next step in the Ball State University project involved creating an experimental simulation with pre-determined truthful and deceptive messages. In addition to examining which linguistic features of texts would predict the ability to detect a lie, the authors also tested the impact of the sender’s apparent status or power. This manipulation corresponds to what happens in the real world when you receive messages from people in a position of authority, whether it’s a boss sending you a text or a political leader sending out a general announcement in a tweet or as a quote in a press release. You might be more likely to believe the high-power messager, unless you tend to take the words of leaders with a strong dose of cynicism.
In part one of the experiment, an undergraduate sample of 46 students (two-thirds female) composed texts under instructions either to commit a white lie or tell the truth. To standardize the texts, the researchers devised eight common scenarios varying in the power relationship to the supposed recipient. See what you would come up with as truthful texts to this example (the low power alternative is in parentheses):
Your boss (coworker) met with you for the fifth time this week. You are somewhat annoyed by how needy he or she has been lately and how much of your time he or she is taking up. One hour after meeting, your boss (coworker) texts you the following message: “Does it bother you that I take so much of your time? I promise to be less needy in the future!”
Unlike the naturalistic part of the study, deceptive messages were actually shorter than truthful messages, supporting the idea that it’s easier to lie in small doses. Like the first study, though, deceptive messages actually did contain more negations (e.g. in response to the above you might cover up your feelings of annoyance by saying “Don’t worry about it”). There were no differences in whether the texts used first- or third-person pronouns or conveyed negative emotions, nor did they differ in their use of verbs. There were no differences based on the power differential represented in the prompt.
In the second phase of the experiment, participants rated the messages generated in part one in such qualities as truthfulness and believability as well as how helpful or hurtful the messages were. Once again, raters were able to separate truth from deception. Suggesting that there’s a reason people commit white lies, raters regarded deceptive texts as less hurtful than the truthful ones but also less helpful. In other words, the truth may hurt but, in the long run, will contribute to your ability to grow.
It’s important to keep in mind that the Ball State study was focused on very specific types of communication, i.e., text messages. However, its great strength was that it built on actual real-life interactions with even the scenarios involving realistic situations. Indeed, if you think about the majority of your texts between the people in your life, they probably also have a strong emotional quality. You’d be more likely to compose a deceptive text about how you’re feeling about the other person than about whether the weather is nice or not.
Moving now from the interpersonal to the larger realm of deceptive messages that you encounter from speakers who you don’t actually know, what can you take from the Holtgraves and Jenkins study? How can you know who to believe when you’re getting those multiple messages?
The findings suggest that people who lie tend to ramble more, seem less certain of the facts, and provide few discernible (or verifiable) facts. Secondly, it appears that if you’re able to read a text translation of a speech or message, you might be better at detecting deception than you might give yourself credit for. Although it may be true that people lie to a distressing degree, it may also be true that recipients of these messages can sniff out fact from fiction.
To sum up, deception may be an inherent part of all human communication. By figuring out how to interpret the content of what you’re reading or hearing, you can find the path to the information you need to guide your important life decisions.
Holtgraves, T. & Jenkins, E. (2020). Texting and the language of everyday deception, Discourse Processes, DOI: 10.1080/0163853X.2019.1711347