Angela Merkel’s Unintended Message

Looking for the Germans in data and "Dichtung" (fiction).

Posted Sep 27, 2019

I would rather live a wandering life in tents than bow the knee to a German idol! –Dostoevsky, The Gambler, chapter IV

Who are the Germans and what are they like? There is no easy answer. Perhaps one day we’ll have the right kind of big data and a machine-learning algorithm to figure it out, but at the present time we rely on impressions, stereotypes, and the reports of those who claim to know, and we hope that these sources contain some truth. 

Prominent exemplars can shape and reinforce stereotypes. Consider Angela Merkel, Chancellor of the Federal Republic of Germany since 2005. Would you agree that she is efficient, industrious, and scientifically-minded, while at the same time lacking emotional expressiveness and a sense of humor? I thought you might. Would you also agree that these trait characteristics have a Germanic ring to them, the ring of the Nibelungen, as it were? I thought you might. Do you now think you see the Germans as an efficient if dour bunch because this is what Frau Merkel projects? Perhaps, but recall that these trait characteristics have a long history of being associated with the Germans, with ethnographic reports ranging from Tacitus to Twain, and finding their first numerical expression in the same year that Hitler became Chancellor of Germany (Katz & Braly, 1933). Is it possible that Merkel is not so much a cause behind our perception of the Germans, but instead an expression of an underlying "national character" (Wundt, 1900)?

The assumption of national character is that certain traits are common in certain groups and that they thereby characterize these groups. If a trait is highly prevalent in a particular group, it is likely that it is less prevalent in other groups. A trait being characteristic of a group is thus a matter of both absolute and relative trait prevalence (Krueger, 1996). Once the notion of the social stereotype was introduced, most social psychologists took the view that stereotypes are false by definition, although the coiner of the term, Walter Lippmann (1922), himself was clear in stating that stereotypes are just simplifications. In a simplification, errors are tolerated so that the essential message can be conveyed efficiently. When we look at a map, for instance, we should not be troubled by the highways looking thicker than they are in reality; we should be grateful that we need not reach for the microscope when planning our route. Of course, some maps include unfortunate errors. Lippmann noted that the cartography of Bohemia should not feature a coastline. Yet, this fictional Bohemian coastline became the lens through which social psychologists began to look at our perceptions of social groups. Many still do, and in spite of the fact that some of our most influential pioneers cautioned against a commitment to the idea of stereotype inaccuracy (Allport, 1954; Tajfel, 1969).  

If social stereotypes, as beliefs about the characteristics of social groups, are declared to be false by definition, an untenable implication is introduced, for by the same logic, we would have to assume that perceptions of individual people are also mere impressions, which must be false by definition. Impressions of individuals rely on the same kinds of fallible sources, so if falsity is true for perceptions of groups, it must be true for perceptions of individuals. Since hardly anyone claims that person perception is false by definition, but many claim that group perception is, we are challenged to resolve this incoherence. Can the idea that only group perception is false be rescued by the claim that, as individuals differ one from another, aggregating perceptions over individuals within a group will always yield the same average no matter which group this is? In other words, can we say that individual differences within a group may be correctly perceived, but that differences between groups, at the mean level, must be nil? This argument may seem plausible to those who are already convinced of the falsity of group stereotypes, but its statistical point is nothing but a restatement of the belief held at the outset, namely that there can be no differences between groups. This argument therefore fails by begging the question, not by answering it.

Many personality psychologists, knowing that they lack an almighty algorithm to reveal a person’s true character, accept aggregated observer judgments as proxies for truth (Funder, 1995). When pressed, they concede that while accuracy entails agreement, the reverse does not hold (Krueger, 2017). Still, the proxy argument does have some empirical and pragmatic legs. Over a large sample of people, it is likely that aggregated person judgments are positively correlated with the true but latent trait, in a wisdom of the crowd sort of way (Larrick, Mannes, & Soll, 2012). We just won’t ever know how large this correlation is.

If psychologists studying persons can do this, why can’t psychologists studying groups? To accept the agreement heuristic in the individual case, but to reject it in the group case without articulating a compelling reason for the difference is hardly a coherent position. I hasten to add that the individual’s or the group’s self-image or self-assessment must be part of the agreement heuristic. This inclusion of the inside view protects the target, be it a person or a group, from being misjudged by a coalition of observers with motives other than description. If no one likes the Yemenites or if no likes Yousef, that is bad for Yousef and the Yemenites, but it does not prove they are truly bad. Their own perceptions must be part of the calculation. The same goes for Gerhard and the Germans.

In person perception research, the so-called round-robin design is a useful way to study how people see themselves, how they see others, and how they are seen by others (Kenny, 1994). The logic of the design generalizes to group perception. In what is called “the full accuracy design,” members of each group rate each group, and it is hoped that aggregated perceptions are partially accurate, providing cues to what groups are really like and how they differ from one another (Judd & Park, 1993).

I once attempted to find clues as to what the Germans, the Italians, the (U.S.) Americans, and the English are like by using a full accuracy design (Krueger, 1996). There was partial success, in that the American and Italian judges strongly agreed on group characteristics, and their perceptions were similar to what Katz and Braly had found 60 years earlier. They saw the Germans as aggressive, industrious, and efficient, and they saw the Italians as passionate, pleasure-loving, and talkative. There was also an interesting failure. I could not get data from England when I realized that the one contact I had was Welsh, and I could not get data from Germany when prospective respondents at a Berlin university refused to complete the survey. They smartly but incorrectly inferred that the true goal of the study was to see if they could be persuaded to make judgments about national groups. In other words, they thought this was a study of obedience. This incident sharply contradicted the Germans’ stereotypic inclination towards obedience, but it was—as a small consolation to me—consistent with a certain rebelliousness and recalcitrance that I had observed from time to time. Does Frau Merkel not offer a closer fit with the German trait profile than with the Italian one?

The Gambler and the Germans

This brings us to Dostoevsky and his gambler. Dostoevsky is widely credited for his subtle insights into the human mind, and many consider his work to have co-inspired Freud to develop psychoanalysis. In The Gambler, however, first published 1866 in Russian and in 1887 in English [what took them so long?], Dostoevsky shares some of his views on national character, and many of his observations of the French, the Germans, and the Russians are only thinly varnished. In chapter IV, Dostoevsky sketches the portrait of the typical German family that, while antiquated in many respects, still has some ring of truth.

Alexei Ivanovich, a.k.a. Dostoevsky, gambles away his money in Roulettenberg, a.k.a. Wiesbaden, among a crowd of jaded international aristocrats of varying means and ambitions. Roulettenberg/Wiesbaden was, and some say still is, a good place to observe how international characters work out their idiosyncratic natures. Dostoevsky’s most penetrating depiction of the German character does not, however, focus on an individual aristocrat, but on what he portrays as the typical family.

The epigraph to this essay is the overture to the gambler’s anti-German screed. He affirms his passion for gambling, knowing that it will probably ruin him, as a rejection of the conservative Protestant ethic of laborious wealth accumulation he observes among the Germans. He has “seen” this ethic and he has “verified” his impression, and it makes his “Tartar blood boil.” “Every German family,” he explains, “is bound to slavery and submission to its Fater [sic]. They work like oxen and amass wealth like Jews.” They mostly succeed, but it is “not a beautiful spectacle—the spectacle of a century or two of inherited labor, patience, intellect, rectitude, character, perseverance, and calculation.” Dostoevsky notes the self-delusion of the German culture and the German’s conviction that the world would be better off if more nations followed the German example. ”They think there can never be anything better than this; wherefore, from their point of view they begin to judge the rest of the world, and to censure all who are at fault—that is to say, who are not exactly like themselves.”

I would argue that this sentiment survives today after the horrors between 1914 and 1945, and in the current era, it leaves a footprint on the relationships Germany has with its partners in Europe. The policies—and the ideology—of austerity, reflect the values of labor and rectitude diagnosed by Dostoevsky. Under Angela Merkel’s leadership, the Germans have done very well economically, but the attitudes of some of their neighbors have soured, not just as a consequence of envy, but in response to the Germans’ perceived air of self-satisfaction. One hopes the Tartar blood will not boil over.  


Allport, G. W. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley. 

Funder, D. C. (1995). On the accuracy of personality judgment: A realistic approach. Psychological Review, 102, 652–70..

Judd, C. M., & Park, B. (1993). Definition and assessment of accuracy in social stereotypes. Psychological Review, 100, 109-128.

Katz, D., & Braly, K. (1933). Racial stereotypes on one hundred college students. Journal of Abnormal and Social Psychology, 28, 280-290.

Kenny, D. A. (1994). Interpersonal perception: A social relations analysis. New York: Guilford.

Krueger, J. (1996). Probabilistic national stereotypes. European Journal of Social Psychology, 26, 961-980.

Krueger, J. I. (2017). Reverse inference. In S. O. Lilienfeld & I. D. Waldman (Eds.), Psychological science under scrutiny: Recent challenges and proposed solutions (pp. 110-124). New York, NY: Wiley.

Larrick, R. P., Mannes, A. E., & Soll, J. B. (2012). The social psychology of the wisdom of crowds. In J. I. Krueger (Ed.), Social judgment and decision making (pp. 227-242). New York: Psychology Press.

Lippmann, W. (1922). Public opinion. New York: MacMillan. 

Tajfel, H. (1969). Cognitive aspects of prejudice. Journal of Social Issues, 25, 79-97.

Wundt, W. M. (1900). Völkerpsychologie. Eine Untersuchung der Entwicklungsgesetze von Sprache, Mythus und Sitte. Leipzig: Kröner.