Last week's Casual Friday study attracted the most e-mails and questions we've ever received. It also attracted the largest number of responses ever: we cut it off at 400, before our Surveymonkey bill got too large (this is probably thanks mostly to our mention in the Seed Daily Zeitgeist on Monday).
What was all the fuss about? First of all, the experience of taking the survey seemed awfully random: first, you were asked your birthday. Then you were asked to think of something sad, happy, or anger-inspiring. Or you were asked about the last time you had something to eat. Finally you were shown this picture:
The picture, as many of you surmised, contains no colors: it's all different shades of gray, but you were asked to name any colors you did see. Gray, Black, and White were not in the list of choices.
So what was going on here?
The study was inspired by a study on synesthesia by Jamie Ward, which we analyzed in Cognitive Daily yesterday. Ward reported on a synesthete whose initials are GW. GW saw colors when she heard or read certain words; words associated with strong emotions were more likely to elicit a synesthetic response. One explanation of this phenomenon, put forward by neuscientist V. S. Ramachandran (as several commenters pointed out yesterday), suggests that synesthesia might be caused by hyperconnectivity of perceptual systems. In other words, all brain systems have some connections to other systems; in synesthetes, there are simply more connections between certain systems. So, for GW, it's likely that her emotion centers were more connected to centers for color perception than in other individuals. The interesting tidbit here is the fact that even in normal individuals, all brain systems are also connected.
Perhaps, we thought, with a large enough sample, we could find synesthesia even in normal individuals. The birthday question was simply a way to divide participants into even groups. We wanted four groups of individuals. We attempted to induce three different emotions: happiness, sadness, and anger, by asking participants to think about a moment that had made them happy, sad, or angry. The fourth group was a control, and we simply asked those people questions about their daily routine -- no emotion should have been induced. Finally, we showed them the gray "landscape" image, and asked what colors they saw. Here are the results:
As you can see, plenty of respondents indicated they saw colors. What's more, there were significant differences in the colors they perceived, depending on the induced emotions. This result was achieved despite the fact that nearly half the participants skipped the last question entirely. Before we get to the significant results, however, let's take a look at the general shape of the data. The most common responses were light blue, light purple, pink, and beige. While there were a some yellows and greens, very few respondents indicated seeing red, orange, or brown.
For the happy group, the top three colors chosen were light purple, pink, and yellow. For the sad group, light purple, beige, and pink, and for the angry group, light purple, light blue, and beige. Let's take a closer look at the responses for the most common colors:
The happy group saw significantly less light blue than the control group. For the happy group, significantly more participants saw light purple than saw light blue, whereas there was no difference in the amount of those colors seen by the negative groups (angry and sad). There was a trend for the happy group to see less light and dark blue compared to the negative groups, and to see more light purple and pink. These results and trends correspond to GW's response to emotional words: GW saw darker and less saturated colors like blues and black when the word described a negative emotion, and brighter and more saturated colors like purple, green, and yellow when the word described a positive emotion.
Does this mean we really do all have a little synesthesia? Based on these results, that might be taking things a little too far. Remember, GW's phenomenon was instantaneous, and directly associated with the emotional words themselves. By contrast, in our study, we had to go to some lengths to induce the desired emotion. Even then, only a bit more than fifty percent of the respondents indicated seeing any color at all. The largest result, 23 percent of the happy group reporting seeing light purple, is still less than a quarter of respondents in that group. But there is significant evidence that, for many individuals, specific emotions are associated with perceiving specific colors in a neutral-colored image.
I wonder, I wouldn't be too surprised if we did have a low-grade synaesthesia in most ppl. Personally I think I do; I get a perception of a weird spiky object in my upper left visual field when I drink Cottonwood's Low Down Brown Ale. I have a similar but lesser experience to beers that are similarly hopped. And no, that's not just from where my head hits the floor.
Given that gray, black and white, were not among the options, I propose an alternate hypothesis. You are assuming that all those who claimed to see a certain color in fact saw that color--you even state results in terms along the lines of "The happy group saw significantly less light blue than the control group." But to be precise, what was observed was "the happy group reported seeing significantly less light blue than the control group."
I propose that at least some of the people who claimed to see a color in fact did not see a color--they were vexed at being asked to name a color that, clearly wasn't there, and made something up. At best, you have shown that what color people claim to see is influenced by their emotions.
I think Chuck has a good point. I chose not to answer the last question because I didn't see any of the listed colors. However, I seriously considered answering randomly, picking a color by which name most amused me, for example. Had I been in a sillier (or more spiteful) mood, I might have done so.
So assuming that people saw colors just because they said they did seems dubious.
To add to the comments of Chuck and Twig, I must admit that I... maybe... slightly forced myself to "see" the first color that came into my head. I think it happened to be the last color my eyes had rested on before turning back to the gray landscape.
Even so -- it is interesting that there was a significant difference in the data, though this might be explained by people knowing what they were "supposed" to answer. For example, I was part of the sadness group and when I came to the gray landscape question I assumed I was supposed to say blue because blue is so often associated with sadness.
For the record, though, I "saw" pink.
I would be fascinating to see if this perception of color persisted across people of different cultural backgrounds. For example, red is an auspicious color for many cultures in the Far East, and it would be interesting to see if people from such cultures would "see red" when feeling happy.
I have had a previous experience of being unable to see the picture you posted without studying my laptop screen from several angles. Since you didn't include an option for the gray scale I saw at first, I thought there was a problem with my screen again. So I really studied the picture, from several angles, and all of a sudden I saw a pale yellow on the horizon.
Oh, how strange! I've just had the same experience. I only saw shades of gray when I first looked at the picture, but now looking back at it I see the yellow again. (I didn't look at it from several angles again, either.) I'm seeing a color where I saw none before.
Boy, now I not only don't trust my computer monitor color rendering, but I'm not so sure about my ordinary perceptions. I think I'm just a lot more susceptible to suggestion than I'd like to be.
I felt a bit vexed by the last question, and didn't answer spitefully, but had to look and reach hard and far to determine my reluctant answer.
"I wonder, I wouldn't be too surprised if we did have a low-grade synaesthesia in most ppl."
I agree. I wondered for decades about the high prevalence of assorted mingling of perceptions in artists, musicians, even languages.
Sorry if I have missed something here but has any consideration been taken for screen calibration? No two monitors even of the same model and batch are truly neutral grey without an external calibration device.
Ryan, you do have a good point, but how would that explain the significant differences between the happy and negative groups? It appears that any differences between monitors aren't significant enough to override the effects of induced emotions.
You all may be interested in another form of "synesthesia" which appears to be present in all of us. That is the association between certain sound contours and certain visual shapes. Ramachandran and I discuss this in our Journal of Consciousness Studies paper, and again in our Scientific American article. Our demonstration (it wasn't really controlled enough to be called an experiment, but that never stopped Gestalt Psychology - if something appears in almost everyone, statistics become almost superfluous) is based on work first done by the Gestalt psychologist Wolfgang Kohler.
In short, if I present you with two shapes, one jagged and one rounded, and ask you which is the bouba, and which is the kiki 95% to 98% of people will respond that the kiki is the jagged shape, and the rounded one is the bouba. We suggest that this is a sort of implicit synesthesia (essentially a mapping between the two senses) present in all of us.
Chris Westbury followed up on this by giving it the proper statistical analysis that it deserved with lots of trials, and controls to make the manipulation less transparent, using reaction time, etc, etc, etc... the upshot is that he replicates our demonstration very nicely.
Some cognitive linguists have even suggested that mappings of this sort are the basis for "sound symbolism" in which certain linguistic sounds are mapped onto meaning, such as the English words twist, twine, twirl, twiddle and so on, that all involve a sort of twisting mouth movement. There have even been some cross-linguistic studies that show similar effects across cultures (after performing controls to rule out onomatopoeia).
So our implicit synesthesia might run beyond just colors and emotions to a large number of other domains, even if we aren't always conscious of these connections.
To back up Chuck's comment - I saw only grey. I thought it was very obvious that there weren't any other colors, but it never dawned on me that you could just skip the question without answering it. So, I felt compelled to use my imagination, and whatever colored scene came to mind first was my answer, which happened to be an orange sunset with green grass (I was part of the happy group).
Vladimir Nabokov, author of Lolita, was a synesthete who claimed that each letter of the alphabet, to him, came with its own color. In fact, there is a book out now that details Nabokov's colored alphabet, complete with his descriptions of the exact colors he sees in each letter. If you aren't already familiar with his writing -- or with his work as a world class lepidopterist -- you'll be surprised at the precision of his descriptions. His colors aren't merely reds and blues; they are exact shades.
Link to the book here: http://www.amazon.com/gp/product/1584231394/002-8981467-1662450?v=glanc…
I agree with Ryan-- there was no control of study parameters, because there were no uniform viewing conditions, especially ambient color temperature and color settings on the user's. Both of these factors significantly affect color perception.
I find it interesting that more cool colors show up overall than warm colors, which to me indicates that many participant's monitors were set for 'cool.' Those who set their monitor's color temperature for 'warm' could easily perceive yellows and beiges. Try it and see.
As for Dave's comment that this could not explain effects of induced emotions, how would he explain that light purple was the top choice whether happy, sad, or angry?
While I do agree that there is probably some degree of synesthetic experience across the population, I can't possibly agree with the premise of this "experiment." For one thing, even the slightest reading on the subject will reinforce the fact that synesthetes tend not to agree on the colors of things, so finding a statistical agreement around color among people asked to look at a greyscale landscape is not convincing of anything, other than that we perhaps tend to have cultural associations with color. For instance, I have multi-modal synesthesia, and anger the word is yellow, but the emotion itself generates a dark cloud with blue rectangles. Sadness is also yellow as a word, but a pale fuzzy brown when experienced. Another synesthete would report something entirely different, and I have come to expect that. Even among colored alphabets, there are really only about four letters that have any kind of statistically significant agreement among synesthetic subjects.
I'd encourage you to read the article on which this casual study is based. I think you'll see that there is some agreement between synesthetes on which colors are associated with which emotions. While our study is most definitely informal, I disagree with you that this "experiment" (which we acknowledge is casual) is based on a false premise.
Asking people to think of numbers might conflict with asking them to think of emotions, given that both numbers and emotions have been associated with synesthetic experiences.
Ideally, people would be asked to consider only one possible synesthetic trigger in an experiment.