The copywriter's 'rule of three' rests on a simple observation: people are drawn to patterns of three.
Three examples feel more complete than two; three adjectives more persuasive than one. Stories, speeches and advertisements all exploit this tendency. The mind seems to settle comfortably into a trio.
Perhaps it is not surprising, then, that the number three should surface elsewhere—in places less straightforward than marketing copy. Among interviewers and deception researchers, three has acquired a more dubious reputation. It is sometimes called the 'liar's number'.
For years, the American statement analyst Mark McClish noticed a recurring detail in deceptive statements. When people were required to invent a number, they often settled on three, or on a number beginning with three. The observation was informal rather than scientific, a pattern glimpsed repeatedly across interviews. Yet it persisted often enough to attract attention.
In 2009, McClish conducted a series of studies into deceptive language. Their primary purpose was not to examine numbers at all, but numbers nevertheless emerged as a curious feature of the data.
In one experiment, one hundred participants were asked to write an account of a day in their lives. The morning section was to be truthful. The afternoon—from 1 p.m. until 6 p.m.—was entirely fabricated.
The evening returned to truth. McClish hoped to compare honest and deceptive language and identify the ways in which people altered their narratives when inventing events.
The results yielded no universal linguistic fingerprint. Participants changed their style when lying, but each did so differently. Vocabulary, sentence structure and emphasis varied from person to person. Yet another pattern appeared unexpectedly. When participants introduced times into the fictional section of their stories, they showed a marked preference for 3:00 or 3:30.
Almost half of the participants—48 per cent—selected one of these times as the first new time reference in their fabricated account. Other times were mentioned far less frequently. Given only five broad options between 1 p.m. and 6 p.m., the finding was hardly definitive. Even so, the clustering around three o'clock was striking enough to invite notice.
A second study produced a similar effect.
Again, one hundred participants were asked to invent a story. This time they imagined returning home to discover that their house had been burgled. They were instructed to include certain details: an open front door, a television left on, a missing gun collection, the number of guns stolen and their value.
The participants were free to choose any number of stolen guns. Their answers ranged widely, from two to fifty. Yet one number appeared more often than any other. Fifteen participants claimed that three guns had been taken. The next most common answer was five. Among larger figures, thirty proved the most popular choice.
Viewed another way, the pattern became more pronounced. The digit three appeared repeatedly across the responses—in the numbers 3, 30 and 35—occurring more frequently than any other leading digit.
McClish is careful not to overstate the significance of these findings. Neither study was designed specifically to test the appeal of the number three. The samples were limited, and alternative explanations are easy to imagine. Human beings may simply find three a satisfying number. It may feel plausible, memorable or instinctively balanced when constructing a story. The same psychological preference that gives force to the copywriter's rule of three may also shape the liar's imagination.
For McClish, the lesson is modest. The appearance of the number three is not evidence of deception. It is merely a point of interest—a small irregularity in the narrative landscape. When three appears, he suggests, it may be worth looking more closely, asking another question, lingering a little longer over that part of the story.
Like many clues in the study of deception, it is not a smoking gun. It is only a thread. But sometimes a thread is enough to make an investigator pause.