Mixing Memory

Why are so many people convinced that we only use 10% of our brains, or that Eskimos have n words for snow, where n is as high as you need it to be for the desired rhetorical effect? Or more seriously, why have some people, particularly Fox News viewers (no, really), persistently believed in Saddam Hussein’s involvement in 9/11? Why does that used car salesman who waves at you as you drive by the dealership on the way to work every morning look so trustworthy, even though you know used car salespeople are never, ever, under any circumstances to be trusted? And why do you dig Henri Matisse’s “The Dance” so much? The answer to each of these questions, and to every other question you will ever ask, is mirror neurons processing fluency. OK, processing fluency’s not the whole answer to these questions and many more, but it’s at least part of it.

You’re probably thinking, Chris, what the hell is processing fluency, and why does it have so much power over my life? Well, I’m going to tell you, but you’re going to have listen to the story of the truth effect first. The truth effect, or more accurately, the illusory truth effect (though that doesn’t have the same ring to it), was first reported in 1977 by Hasher, Goldstein, and Toppino(1). They gave participants lists of plausible statements on three occasions, each separated by two weeks, and were asked to rate how confident they were that they were true or false. Some of the statements were actually true, and some were false, but none of them were likely to be known by their college student participants. So for example, a participant might have read, “Kentucky was the first state west of the Alleghenies to be settled by pioneers,” and “French Horn players get cash bonuses to stay in the U.S. army” which are (or at least were, for the second one) true statements, and “Zachary Taylor was the first President to die in office,” which is false.

The key manipulation was repetition. Each participant saw 60 statements in each session, and each time, 40 of the statements were new (that is, they hadn’t been seen in the previous sessions). Twenty statements, however, were repeated in all three sessions. Here are the results for the false statements across the three sessions (confidence ratings were on a scale of 1-7, with 7 being the most confident):

i-fd4edf920b8e272cdf21b257e2a4731f-HasheretalFalseData.JPG

As you can see, people’s confidence in the truth of the repeated statements went up from the first to the second and second to third sessions, but stayed about the same across the sessions for the non-repeated statements. The pattern was the same (as were the absolute numbers) for the true statements, which is not surprising, because participants were unlikely to actually know which were true and which were false. This, then, is the truth effect: repeating a statement makes it seem more likely to be true.

For a couple decades, the only explanation for this effect was familiarity. Schwartz(2), for example, conducted an experiment similar to Hasher et al.’s, but asked half of his participants to rate the “pre-experimental familiarity” of the statements, instead of their truth. He found that repetition also increased “pre-experimental familiarity” ratings. He concluded that when statements are unfamiliar, we judge their truth based on memories that suggest whether they’re plausible or implausible, and a general sense of familiarity. Thus, the more familiar the statement feels (if it’s been repeated, say), the more true it feels.

But familiarity is a pretty abstract theoretical construct. The question, then, is how does familiarity affect the perception of truth? You’ve probably guessed the answer: processing fluency. In a nifty little experiment, Reber and Schwartz(3) manipulated how easy unfamiliar statements were to read. Some of the statements were presented in a light-colored font, making them difficult to read, and others were presented in a darker font, so that they would be easy to read. The light-colored font yielded truth ratings at chance. That is, the participants were just guessing, because they had no idea whether the statements were true or false. The darker font, however, made the statements seem more true. In a recent paper, Christian Unkelbach(4) used signal detection theory to argue that this effect is likely due to “discrimination ability.” That is, when statements are unfamiliar, processing fluency makes it more difficult to distinguish between statements that are true and those that are false.

So the tale of the truth effect should probably make it clear what processing fluency is: it’s simply the sense of how easy it is to process incoming information. The easier it feels, the more fluency. And fluency judgments seem to be ubiquitous, if implicit (that is, you probably don’t know you’re making them). For example, making it easier to process a brand’s label by priming concepts related to objects on the label makes that brand seem preferable, as does presenting information about a brand in a more readable font (5). Processing fluency also affects how popular an opinion seems. Hearing an opinion repeatedly, even if only from one person, makes the opinion seem like a popular one (6). Some theorists have even argued that processing fluency underlies our experience of aesthetic pleasure (7).

And finally, as if to add insult to injury, perceptual fluency can affect how credible a face looks. Brown et al.(8) showed participants sets of faces on two occasions (either 2 days or 2 weeks apart), with some faces repeated in both sessions, and some new in the second session, and asked them to rate how honest and sincere they looked. Regardless of the interval, repeated faces appeared slightly more honest and sincere. So if you hear a false statement over and over again from a familiar face, you’re screwed. Thanks to processing fluency, it will seem like the person’s more honest and the statement’s more true and popular.

OK, well that’s probably a bit of an exaggeration. In fact, it’s probably a big exaggeration. Processing fluency is just part of the picture. The truth effect, at least as demonstrated in the lab, is really pretty small, and it only works with statements we’re very uncertain about. When we know something about a statement, we use other mechanisms to determine their plausibility and truth. So it’s just as likely that people believe that Saddam had something to do with 9/11 because it fits with their world view as it is that they believe it because they’ve heard it over and over again. But processing fluency will affect the way we interpret the world around us, and I am more and more convinced by the view that fluency monitoring is going on pretty much all the time, and informing the further processing of just about all of the information we take in. I’m just not sure that fluency monitoring is going to let the poor used car salesman who waves at me every day trick me into buying a car I don’t want just ’cause his face looks trustworthy and he tells me over and over again that a 1999 Ford Taurus with that many miles really is a great buy at that price.


1Hasher, L., Goldstein, D., & Toppino, T. (1977). Frequency and the conference of referential validity. Journal of Verbal Learning and Verbal Behavior, 16, 107-112.
2Schwartz, M. (1982). Repetition and rated truth value of statements. The American Journal of Psychology, 95(3), 393-407.
3Reber, R., & Schwartz, N. (1999). Effects of perceptual fluency on judgments of truth. Consciousness and Cognition, 8, 338-342.
4Unkelbach, C. (2007). Reversing the truth effect: Learning the interpretation of processing fluency in judgments of truth. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(1), 219-230.
5Novemsky, N., Dhar, R., Schwarz, N., & Simonson, I. (2007). Preference fluency in consumer choice. Journal of Marketing Research, 44, 347-356.
6Weaver, K., Garcia, S.M., & Schwartz, N. (2007). Inferring the popularity of an opinion from its familiarity: A repetitive voice can sound like a chorus. Journal of Personality and Social Psychology, 92(5), 821-833.
7Reber, R., Schwartz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience. Personality and Social Psychology Review, 8(4), 364-382.
8Brown, A.S., Brown, L.A., & Zoccoli, S.L. (2002). Repetition-based credibility enhancement of unfamiliar faces. The American Journal of Psychology, 115(2), 199-209.

Comments

  1. #1 Patricia Mathews
    September 19, 2007

    Repetition and familiarity is only half the story. The other half is that the things believed make a god story. They seem plausible. The eskimo one seems like a fact and not a story, but behind the fact is that we know they live in the arctic and if we were them … it fits what we perceive their environment to be. Likewise the 10% of our brains story fits our observations of what fools us mortals can be and how often.

  2. #2 Andrei
    September 19, 2007

    Interestingly, you have mentioned Matisse’s Dance. I have been to Hermitage many times. With all my respect and love to impressionism and post impressionist art, every time I was walking by the Dance, I was telling myself “What a scam”.

    More on that, I have tried to get hypnotised several times during those public sessions, you know. And every time I was sent back from the stage, because the hypnotist could not do anything to me.

    Do you think some people may have some fluency problems?

  3. #3 Pelican's Point
    September 19, 2007

    This doesn’t seem right:

    “So it’s just as likely that people believe that Saddam had something to do with 9/11 because it fits with their world view as it is that they believe it because they’ve heard it over and over again.”

    Do you really think that the truth effect is as powerful as ideology (fitting into one’s worldview) in what we choose to believe or not?

    The questions were chosen to purposely not be part of the students’ worldview – or anything they would be familiar with in any case.

  4. #4 anonymous
    September 21, 2007

    It’s Norbert Schwarz, not Schwartz.