Literature Is Hard to Remember—Compared to Facebook
And gossip is more memorable than the evening news, according to a new study
Steven Cherry: Hi, this is Steven Cherry for IEEE Spectrum’s “Techwise Conversations.”
Maybe it’s the death of formalism. Back in January we talked about “gamification,” the idea to incorporate elements of games into schools, to improve student engagement. Now there’s new research showing that tweets and Facebook postings are more easily remembered than sentences from books.
The study compared Facebook to novels, but if it’s true for enjoyable fiction, it’s surely true for textbooks as well. So once again, maybe we should rethink something about schools. Maybe also worth rethinking is the weird fact that conferences publish formal papers, but what you see at the conference is a PowerPoint about the paper. Maybe professional societies should be publishing the presentation, not the paper, or at least both the presentation and the paper.
My guest today is one of the coauthors of the study, Laura Mickes. She’s a senior research fellow at the University of Warwick and currently a visiting scholar at the University of California at San Diego. By presumably no coincidence, all but one of her six coauthors are also from UC San Diego. She joins us by phone from there.
Laura, welcome to the podcast.
Laura Mickes: Thank you. Thanks for having me.
Steven Cherry: Before we declare formalism dead, let’s start with the results themselves. You did three experiments. The first was a straight-up comparison of Facebook posts to sentences from published books and also to faces. I have an absolutely terrible memory, so let’s not test me. But maybe you can give us an example of what the subjects were asked to do?
Laura Mickes: Right. So the subjects were asked to memorize 100 either Facebook posts or sentences from books. And so an example of one of the posts might be “because sometimes it makes me wonder” and an example of a sentence from a book might be “Even honor had its limits.”
And so you would be memorizing, depending on which condition you were in, either the Facebook post condition or the sentences from books condition. You would be memorizing 100 of those types of stimuli, and then tested on 100 of those items that you did see, and then 100 items that you didn’t see. And then you had to tell us whether you had memorized those or not. And we found that memory for the Facebook posts was much better than the memory for the sentences from the books.
Steven Cherry: And how did the faces stand up?
Laura Mickes: Oh, the faces did even worse than the sentences from books. That was a little surprising. So these faces were neutral faces. So, often when psychologists and other cognitive scientists will study the memory or perception of faces, they use this database. So it’s a pretty standard database of faces, where the faces are shown in frontal view and they have neutral expressions. So our participants had to study 100 of those and be tested on 100 of those faces and 100 faces that they didn’t see. So it was a very challenging test for them, and you could tell by their accuracy scores and their memory for that.
Steven Cherry: Yeah, 100 sounds like a lot. And I would be terrible at faces. Now, your second experiment tested whether this result was due to “deep social encoding.” What is deep social encoding?
Laura Mickes: So we did—the first experiment we ruled out a lot of boring explanations. So it could be that the Facebook posts have these irregular characters, emoticons, and exclamation points and smiley faces, and so we ruled out that that was contributing to the memorability. So even posts without those irregular characters were remembered really well.
We also ruled out some other things, but we were left wondering: Could it be that when we read a post like “because sometimes it makes me wonder” we naturally and effortlessly encode it or memorize it with respect to someone we know? So I might think, “Well, Steven might write something like that,” so for that reason I would remember that better than certainly a face or a sentence from a book, because you don’t naturally do that with sentences from books.
So we tested that directly. We looked at—we had people rate how easy it was for them to think of someone who would write or say something, write that post or write that sentence from the book, and if people were doing this naturally with the Facebook posts, memory wouldn’t be increased, but it would be if they did it for the sentences from books, because they wouldn’t be doing that naturally. But we found that they got a boost in memory for both stimuli types, so for both the Facebook posts and sentences from books, which means that’s not really what they were naturally doing.
Steven Cherry: Right. The Facebook posts memory for them improved, so they weren’t doing it in the first place. So in the third experiment, you compared headlines to sentences.
Laura Mickes: Yes, so in the third experiment, we looked at headlines, we looked at sentences within the articles, and we looked at the comments from below. So the comments are microblogs. These are written the same way the Facebook posts are—at least they seem to be—so they’re both microblogs. And we thought maybe the posts were advantaged because they were not plucked out of context. So the sentences from these books are plucked out of context, and we thought maybe that’s giving the advantage to the Facebook posts.
So that’s why we tested headlines, which are complete ideas unto themselves. So we also looked at the idea that possibly the posts were gossipy in nature, and that would improve memorability for those items. So we looked at breaking news, CNN breaking news, and we looked at CNN entertainment news. So the entertainment news would be gossipy, and the headlines would be complete ideas. And so we did find—we were surprised that the headlines weren’t remembered better. Actually, the comments at the end of the articles were remembered the best. So even headlines, which are written to grab your attention, and for that reason should be more memorable, weren’t.
Steven Cherry: You also found that the more editing that’s been done to a piece of writing, the harder it was to remember. I have to say that’s a pretty depressing result to a magazine editor.
Laura Mickes: [laughs] I know. I joke with my coauthors that I’ll never edit another piece of writing again, and they don’t find that funny at all. So that’s, in a way, speculation. Some people disagree with me, but when I see people posting on Facebook, you know, you’re there watching them post on their phones, and they do it with hardly any editing.
So it may be, of course, large individual differences, where some people might really labor over a post. But I think for the most part, people are certainly doing a lot less editing with their Facebook posts and with the comments than the people who are working on those headlines and who have gone to school for journalism and done a lot of practicing and honing their skills to write really well. So I think there is something about natural language, that sort of spoken language, and microblogs in a way gets us back to a spoken language.
Steven Cherry: Now, you call this kind of casual writing, the Facebook posts and the comments at the end of news articles, “mind ready.” What does that mean?
Laura Mickes: Right. So, mind ready. That’s where we think the natural language, so the fact that a Facebook post and the comments at the end of these articles, these microblogs, are remembered better because they’re closer to natural language. So you and I having a conversation, we don’t really edit our thoughts. It’s more free flowing, and the idea that if something springs to mind, it comes to mind easily, and then it’s easily remembered. And we think that microblogs are more like that than something that’s heavily edited or something that’s really well thought out, that we see in sentences from books or news articles.
Steven Cherry: Our podcasts, often the transcripts, do better in terms of page views or time spent on page than other articles here. Do you think that might be because being conversations they’re more mind ready?
Laura Mickes: That’s possibly the case. That’d be a question I could turn back on you: Would you think the blogs are written in a more conversational style? Are they edited less?
Steven Cherry: Well, they’re edited less than the hard-core articles, but they are edited. But maybe we should rethink that practice, of course, now.
Laura Mickes: Right. That’s fascinating. That’s another empirical question we could test.
Steven Cherry: A lot more people have read Harry Potter than To the Lighthouse. Many people find great literature hard to understand or at least harder and slower to read. Do you think that also might be a question of mind readiness?
Laura Mickes: Yeah, I did. I do think so. We ran an experiment—it wasn’t in the paper that we’re talking about now—but we ran a follow-up experiment testing the idea of mind-ready language. And what we did was we took sentences from highbrow books and lowbrow books. So those highbrow books would be those that won awards, the Pen/Faulkner Award, and the lowbrow is we took bestselling books.
So we took sentences from each of those, thinking that it would get at your question, but we didn’t find much difference in terms of memorability for those. But it is hard to tell. We looked at those sentences and thought, “Well, gosh, I can’t even decide. I wouldn’t be able to say if that were high- or lowbrow.” But I think if you made that difference bigger, we would find a difference in memorability.
Steven Cherry: You mentioned the idea before that the sentences in books were pulled out of context, and I was wondering what might happen if you compared, say, a literary short story, something that was heavily edited, presumably, versus say an entire Facebook timeline. You know, I’m thinking the short story might be more memorable because it has a narrative structure. It’s well developed in its characters, just generally might have more cohesion.
Laura Mickes: Right. That’s a good question. So, another empirical question that you could answer. I’m not quite sure which would be more memorable. If our mind-ready hypothesis is correct, it would be the timeline just because of the way they were written more spontaneously than the short story.
Steven Cherry: Exactly. It seems there’s plusses and minuses to both.
Laura Mickes: Right.
Steven Cherry: People remembered text better when it was gossipy and you compared entertainment news to breaking news stories. I was wondering if that might get back to the deep social encoding. We kind of think of celebrities as not strangers.
Laura Mickes: So the information from entertainment news was remembered better than breaking news, and we think that one of the reasons, yes, it’s more gossipy in nature, but still, the comments remembered, really, they remembered better even with entertainment news. The comments were remembered better than the headlines and sentences from entertainment news. But it also could be that our participants were familiar with those things already. So our participants know more about the Kardashians than they do about Syria, so it could be that they already knew some of this information.
Steven Cherry: So they had a context for it, in other words.
Laura Mickes: Right. I think that could account for a bit more of the memorability for that condition.
Steven Cherry: I mentioned that your results are kind of depressing for a magazine editor. One of your coauthors was quoted as saying, “One could view the past 5000 years of painstaking careful writing as the anomaly. Modern technologies allow written language to return more closely to the casual personal style of preliterate communication, and this is the style that resonates and is remembered.” So I’m wondering if maybe the height of memorable literature was Homer’s Odyssey, and it’s been downhill ever since?
Laura Mickes: Oh, gosh, that’s depressing all right. We’ve thought about testing well-known novels, but I’m not sure. I keep going back to it’s an empirical test. His quote there that you read is really capturing the idea of mind-ready language in that we are just more attuned to spontaneous speech, not the heavily edited or well-thought-out written language.
Steven Cherry: It seems to me that the value of a classroom lecture is in part revealed here: It’s less formal and therefore maybe more memorable than the textbook, say. But you go further and also recommend that professors also tweet after a lecture.
Laura Mickes: I do [laughs], yeah. I do say don’t give up the textbooks, and certainly the lectures, I think they do help because they are more spontaneous and natural. I think some technologically savvy professors are already doing that. A lot of them have Facebook pages for their classes. I think that’s a good idea in light of our results, to be doing those things and have those students follow and engage in that way too. I mean, our students are on Facebook and are tweeting already. I think we need to join them and keep up with them there, and I think it will be meaningful and helpful based on our results.
Steven Cherry: A lot of teachers are moving to more in-class exercises and collaborative learning of various kinds. Maybe the value here is that, in part, that’s also more mind ready?
Laura Mickes: Right, yeah, I think so. I’m brought back to what you said at the beginning, that maybe conferences, that how we present our work on PowerPoint, instead of those often-hard-to-get-through articles, I think that publishing those things together might be a very good idea. I like that idea.
Steven Cherry: You didn’t study the effect of a deliberate attempt to memorize something, but I notice how people who were forced to memorize poetry in school, for example, can remember those poems decades later.
Laura Mickes: Yeah, that would be kind of neat. That would be a neat thing to have them do: see if they could freely recall them well. I think this work has opened up a lot of questions, so I think we’re going to see a lot of follow-up investigations, and that would be a very good thing to do.
So will these posts be forgotten faster? Will that forgetting rate be faster for the posts? I don’t think so, but it’s something you could answer, and, yeah, it would be nice, it would be less depressing, if the poetry were remembered for a longer time than a Facebook post.
Steven Cherry: That would make me feel better, I guess.
Laura Mickes: Yeah [laughs]. I can’t guarantee that result though.
Steven Cherry: Right. Well, Laura, it’s absolutely fascinating, and I think we could talk for the rest of the day about this and some of your other research, but my listeners wouldn’t remember it anyway, I guess, so…
Laura Mickes: Thank you, Steven. It’s been a lot of fun.
Steven Cherry: Very good. We’ve been speaking with Laura Mickes of the University of Warwick and some other school about how memory is improved the more casual—and even gossipy—our writing is.
For IEEE Spectrum’s “Techwise Conversations,” I’m Steven Cherry.
This interview was recorded Wednesday, 12 March 2013.
Segment producer: Barbara Finkelstein; audio engineer: Francesco Ferorelli
NOTE: Transcripts are created for the convenience of our readers and listeners and may not perfectly match their associated interviews and narratives. The authoritative record of IEEE Spectrum's audio programming is the audio version.