July 21, 2006

We feel sad because we say ü

This is old, but too good to pass up. Seed Magazine put it like this:

Germans can be grumpy, unpleasant people—and it's not because of post-Nazi guilt or a diet filled with bratwurst, says one American researcher. It's because of their vowels. Hope College psychology professor David Myers says saying a vowel with an umlaut forces a speaker to turn down his mouth in a frown, and may induce the sadness associated with the facial expression. Myers added that the English sounds of "e" and "ah" naturally create smile-like expressions and may induce happiness. Clearly the solution for the Germans, much like the solution for every other people in the world, is to become more like Americans. The German Embassy would not comment on the findings, saying they were "too scientific."
(source: BBC)

Ah, BBC science reporting!

Here's the original BBC story, from way back in 2000: "'Vowels to blame' for German grumpiness". I missed it at the time, due to the fact that the blogosphere (including Language Log) didn't exist. As usual, the article is an extraordinary display of faith, hope and charity, with a considerable admixture of creativity. We can all be grateful to Seed Magazine for reviving this extraordinary story from 2000 as "New and Notable" in January of 2006. (You'll note that I'm working hard on my New Year's resolution to take a positive attitude towards science reporting.)

The author of the reported study is David Myers, a social psychologist whose web site identifies him as "a communicator of psychological science to college students and the general public". According to the BBC story, Myers

... just finished a sabbatical at St Andrews University which involved using electrodes to manipulate the muscles of the face - research which, he said, bore out his theory.

So apparently the idea that umlauts make Germans grumpy was suppported by the results of electrically stimulating the faces of Scottish students. (As far as I know, there is no substance to the rumors that parrot telepathy was also involved.) It seems that this research was never published by Myers himself, neither in a scientific journal nor in the popular press. At least, nothing relevant is listed in his publications list, nor could I find anything on Google Scholar. And Andrew Hammel wrote to Myers back in February to ask for a pointer to a publication, and apparently got no answer. (If anyone knows a citation, please send it to me. I'll also write to Prof. Myers myself, as this may well be another Glenn Wilson case. [Indeed it was: see Prof. Myers' response.])

The idea that the expression of emotion might actually cause the associated emotions to be felt is not an unreasonable one. In fact, William James famously argued that emotion itself is simply our perception of its bodily expression: "we feel sorry because we cry, angry because we strike, afraid because we tremble". Or at greater length:

Our natural way of thinking about these coarser emotions is that the mental perception of some fact excites the mental affection called the emotion, and that this latter state of mind gives rise to the bodily expression. My theory, on the contrary, is that the bodily changes follow directly the perception of the exciting fact, and that our feeling of the same changes as they occur IS the emotion. Common-sense says, we lose our fortune, are sorry and weep; we meet a bear, are frightened and run; we are insulted by a rival, are angry and strike. The hypothesis here to be defended says that this order of sequence is incorrect, that the one mental state is not immediately induced by the other, that the bodily manifestations must first be interposed between, and that the more rational statement is that we feel sorry because we cry, angry because we strike, afraid because we tremble, and not that we cry, strike, or tremble, because we are sorry, angry, or fearful, as the case may be. Without the bodily states following on the perception, the latter would be purely cognitive in form, pale, colorless, destitute of emotional warmth. We might then see the bear, and judge it best to run, receive the insult and deem it right to strike, but we should not actually feel afraid or angry.

A good example, in my opinion, of how it takes a really smart person to have a really stupid idea.

I love the way the BBC story ends:

A spokesperson for the German Embassy said: "We can give no comment on this as it is too scientific."

"Wovon man nicht sprechen kann, darüber muß man schweigen."

[Hat tip to Margaret Marks.]

[There's been lots of email on this topic. Mark Seidenberg wrote that Myers has written very successful psychology textbooks, and has also made a substantial donation to the Association for Psychological Science. I wrote back that I'm witholding judgment about Myers' role in the BBC story -- it seems quite likely that the original research was entirely credible, and he contributed only an off-hand remark or joke about umlauts, which the BBC's reporter spun up into a n elaborate confection of pseudo-scientific bullshit (I use that word in its technical sense here).

Fernando Pereira wrote

Is that idea so stupid? Antonio Damásio claims quite a bit of experimental evidence for it in his books. It makes sense in a model in which feelings are a higher level, conscious report of states with positive or negative associations. The states, which have bodily aspects, are in the primary, reactive control loop, while the feelings engage higher level, longer term memory and executive functions.

I observed in reponse that the theory (as William James expresses it) seems to predict the impossibility of unexpressed emotions, which contradicts the experience of everyday life. You could tell a story about covert expression at the level of hormone levels, or synaptic chemistry, or brain rhythms, or something, but at that point the idea has lost its popular/Jamesian form completely. That's not to deny the plausible idea that there is feedback from external, macroscopic emotional expression to internal states.

Caitlin Light wrote:

That has got to be a joke. I don't see how Germans can have "grumpy vowels" when a common Berliner's goodbye is "Tschüssiii!" Besides which, German's aren't especially grumpy, from my experience. I would bet I can out-grump any of my German roommates.

Well, I suppose that the claim would be about statistical tendencies, not individual roommates. But the BBC article certainly did not cite any evidence that German-speakers are any grumpier, on average, than speakers of other languages.

Robert Hellig wrote:

What makes me even more suspicious about this Umlaut story is the fact that English has similar sounding vowels as the German Umlaut, even if they are not written with two dots. To my amateur ears ä is the same sound as in BigMac, ö as in turn or turbulence and ¨ as in bureaucratic. Thus my theory would rather be that the process of writing lots of dots makes you grumpy, as the facial expressions should be the same for speakers of the two languages.

Certainly for some pairs of English and German dialects, such pairings would be pretty close. But again, there was no evidence in the BBC report that anyone involved had actually considered the relative inventory of sounds and articulations in a serious way.

And Lane Greene wrote:

I am very certain, because I remember it so clearly, that in my college introductory psychology 101 textbook was mention of an experiment, presumably a serious one, in which subjects made various vowel sounds for a length of time. They were told they were doing a phonetics experiment. But a post-vowel battery of mood tests showed that the ones that had made "eeeeee" were in the best mood. I wish I could cite, but I know even my fevered imagination didn't make this one up. I've told people the story more than once over the years.

That's certainly a good story, and it might be true. (I'm sure that Lane's memory of Psych 101 is accurate -- I mean that the textbook might have given an accurate account of an experiment, though incremental overinterpretation in such cases is not unheard of. I'll look around for the reference. [Update: more on this here.]) ]

[Eva Ciabattoni writes:

I found it interesting that the umlaut study was done by an American. I'm a native speaker of both German and English and have noticed that Americans have as much difficulty pronouncing umlauts as German speakers have pronouncing American vowels. It makes me wonder how the study was structured. The foundation of such research would necessarily be accurate pronunciation.

Eva, I'm afraid that you're being far too rational here. As far as it's possible to tell from the BBC report, there were neither any Germans nor any vowels involved in this research. Instead, subjects at St. Andrews (presumably Scottish university students) were made to contract various facial muscles by electrical stimulation, and their mood was subsequently tested. Either Prof. Myers or (much more likely) the BBC reporter added all the stuff about Germans and umlauts, as a sort of quasi-scientific free association ("so if frowning makes you somewhat more likely to feel sad even you're forced into it by electrical stimulation of your facial muscles, maybe the same thing happens if the same muscles happen to be used in producing certain sounds, like, oh, say, maybe German umlauts?").

Eva continues

In fact, I'm sitting here right now trying out my umlauts. Umlaut-a puts a smile on my face. Umlaut-o makes my purse my lips as though I'm blowing smoke rings, but I'm not frowning. Umlaut-u makes me jut out my lower lip like an emoting chimp, but I'm still not frowning. Of course, the true test is how I feel after the exercise.

Let's see - Vaguely anxious with a growing sense of foreboding, but that began earlier when I read the latest news from the Middle East. I doubt whether it's due to the umlaut exercise I just did.

Well, Eva, I think you've probably just done more umlaut-related research than was done at St. Andrews, and I'm certain that you've looked into the matter more thoroughly than the BBC reporter did. ]

[OK, here's some more on the "say ee to be happy" idea. I checked Henry Gelitman, Alan Fridlund and Daniel Resiberg, "Psychology" (Fifth edition). They devote a paragraph to the James-Lange theory and the "facial feedback hypothesis" (p. 479), but don't mention any studies of the effects of pronouncing different vowel sounds. The most recent review I could find (Carroll E. Izard, "Facial Expressions and the Regulation of Emotions", Journal of Personality and Social Psychology, 58(3) 487:198. 1990) concludes that "external manipulation of facial expression produce[s] only a weak effect". In more detail:

Laird (1974) proposed the hypothesis that a subject-blind, experimenter-manipulated facial expression would elicit a corresponding emotion experience. Although this hypothesis has been attributed to Darwin, James, Tomkins, Gellhorn, and Izard, it is clearly distinct from their ideas of regulating naturally occurring feelings by self-management of expressive behavior or activating feelings through goal-directed, self-initiated expressions. Explanation of the effect of such self-initiated actions and those of experimenter-manipulated facial movements may require different concepts and the assumption of different mechanisms. That investigators have confused Laird's hypothesis relating to subject-blind, experimenter-manipulated expressions and self-managed or self-initiated expressions has contributed to the FFH [Facial Feedback Hypothesis -- myl] controversy. Because the research relevant to this controversy has been reviewed several times, I will present only a brief synopsis to set the stage for a reconceptualization of this research.

Three recent reviews of the approximately 20 studies relating to FFH reached divergent conclusions. Laird (1984) concluded that the evidence overwhelmingly favors FFH, supporting the notion that experimenter-manipulated facial expressions affect emotion experience. Winton (1986) reviewed the same studies and demonstrated that Laird's conclusion applies only to a weak or dimensional form of FFH. Most of the favorable evidence comes from studies that manipulated one positive and one negative facial expression, showing only that the facial-expression manipulation changes emotion experience on a positive–negative or hedonic dimension. According to Winton, Laird's conclusion that manipulated expressions of joy and anger lead to the experience of joy and anger, respectively, is unwarranted.

Winton argued that the strong or categorical form of FFH (e.g., manipulated joy expression leads to joy experience and anger expression leads to anger experience) requires that the experimenter contrast the effects of at least two positive or two negative emotions and use a dependent measure that can adequately distinguish between the experiences of these emotions. The only published study that meets these criteria (Tourangeau & Ellsworth, 1979) failed to support the categorical version of FFH. Contrary to most of the evidence, it also failed to support the dimensional version of FFH. A series of unpublished studies (described in Izard, 1977, Chaps. 3 & 4) that met Winton's criteria also failed to support the categorical version.

By a "weak or dimensional form of FFH", Izard means that experimenter-manipulated facial expressions might affect subjects' emotional states, not because of a normal association between the expressions and the emotions, but for random other reasons (e.g. because "facial movements that increased cerebral temperature through changes in vascular blood flow triggered unpleasant feelings, whereas those that decreased brain temperature activated pleasant feelings", or "because the procedure impresses subjects as an irrational and intrusive demand, and they respond emotionally ... [which] may help account for the lack of an emotion-specific response to the facial movements"). Much more on the relevant psychology is here, including description of an experiment on the emotional impact of producing different vowel sounds.]

Posted by Mark Liberman at July 21, 2006 08:18 AM