Professional Documents
Culture Documents
Boudon (2013) The Origin of Values - Sociology and Philosophy of Beliefs PDF
Boudon (2013) The Origin of Values - Sociology and Philosophy of Beliefs PDF
Boudon, Raymond.
The origin of values : essays in the sociology and philosophy of
beliefs / Raymond Boudon.
p. cm.
Includes bibliographical references and index.
ISBN 0-7658-0043-8 (alk. paper)
1. Values. I. Title.
References
Index
Introduction
cause of the "cycles of attention" effect, but also because, while some
authors are recognized as having developed a theory of values, others
have rather developed indirectly interesting explanatory schemes on
this subject, without having properly proposed a full-fledged theory.
The case of Rawls7s Theory of Justice (Rawls, 1971) illustrates this
point: Rawls proposes among other things a theory as to whether
people consider the inequality in the distribution of primary goods
legitimate or not. While Rawls is not considered a theorist of values,
he is, in fact, one since his book contains what may be considered an
ingenious explanation of the feelings of people confronted with a given
level of inequalities in the distribution of primary goods. Moreover,
beyond this explanation, a more general explanatory scheme can be
detected in his theory: people tend to consider X as good, if and only
if they tend to consider X as good in the case where they would be
able to make an entire abstraction of their interests, in a word, where
they would consider X as good once placed under the "veil of igno-
rance." Rawls's scheme is powerful, which is undoubtedly a major
reason for his success. But it is hardly noted that Rawls's "veil of
ignorance" has its roots in an earlier and today underestimated idea,
such as Adam Smith's notion of the "impartial spectator" as opposed
to the "partial actor," as we would say, or as Kant's notion of "practi-
cal reason."
A first task for anyone interested in the theory of values and valua-
tion is to acquire an overview of all the explicit and implicit, forgotten
or popular, theories on the subject produced by the classical and mod-
em social scientists and philosophers.
The "cycles of attention" effect is also responsible for the fact that
some theories are commonly misread. Such is the case of Durkheim's.
Durkheim is currently read with the glasses of the neo-Durkeimians: a
given individual in a society would believe that "X is good," because
everybody around him believes so; the content of the "collective con-
sciousness" would be the ultimate cause of the content of any indi-
vidual consciousness. Many anthropologists among others endorse this
type of explanation. Thus, to the prominent American anthropologist
Clifford Geertz, all cultures have their idiosyncratic values; people
normally endorse them simply because those around them endorse the
same values (Geertz, 1984). As Geertz rightly recalls, Montaigne had
already noted that values vary from one society to another, and con-
tended that values should for this reason be considered as arbitrary,
Introduction 5
and that people endorse them under the mere effect of "socialization,"
as we would say. That this is true in some circumstances is beyond
doubt. We believe a number of things because others think so. But
whether this is always true is questionable. "We believe a million
things on the faith of others," wrote Tocqueville. By this he meant
that, if I am not, say, a biologist myself, I would tend to draw my
convictions on biological matters from biologists. In this case, I know
that I believe that "X is true" on the faith of others, and I perceive
clearly why I follow the opinions of other people. Geertz, however,
draws from Montaigne much more than these simple ideas: to him, all
our beliefs would be individual projections, so to speak, of collective
beliefs.
Durkheim does not use the word value in the sense Nietzsche gave
it, since it was not current in France in Durkheim's time, but he
developed a well articulated theory of values. Whether his main point
on the subject was, as often contended by neo-Durkheimians, to de-
velop in his own words basically the same views as Montaigne is
highly questionable. The careful reader of The Elementary Forms of
Religious Life is impressed, on the contrary, by the fact that Durkheim
insists repeatedly on the point that beliefs should not be analyzed as
illusions (Boudon, 1999a). Even a cat, he writes, can easily distinguish
between a ball of string and a mouse. The cat will for a while play
with the ball of string, because it believes it is a mouse. But it will
soon see that the ball is not a mouse and lose interest in it. A child
talks to his puppet, giving an observer the impression he believes his
puppet is a living being. But he would be surprised if his puppet
should suddenly punch him in the nose. Should all these people who
believe in all kinds of things in which we do not believe ourselves,
asks Durkheim ironically, be considered as being more easily the vic-
tims of illusion than a child or a cat? We normally hold something as
true in a first stage, writes Durkheim, because everybody around us
believes so; but in a second stage, we need to know it is actually true.
A belief is held by a given individual as true in a first stage because it
is collective; in a second stage, it becomes collective because it is
true.' To Durkheim, while values vary from one society to another,
they should in other words not be analyzed as illusions. They vary
from one society to another because the context, the cognitive re-
sources, the level of information of people on such and such subjects
and many other factors vary. But they are far from being mere illu-
6 The Origin of Values
sions injected into people's minds by obscure social forces. This does
not lead to the conclusion that all societies should have the same
values. Firstly, because some values may be functional in one society,
but not in another; secondly, because different values may be equally
legitimate in the eyes of social actors; thirdly, because values are often
expressed in an indirect way, through symbols, metaphors, myths or
more or less opaque concepts, and because there are obviously many
ways to symbolize verbally a value. Thus, the notion of "person," the
concept of "soul," the notion of "equality" all point to the value of the
equal dignity of all. In a word, Durkheim's theory is much more
complex and also much more acceptable than many of the more mod-
em theories, even though they state that they lean on Durkheim's
work.
The same effect of the "attention cycles" can be illustrated in the
case of Max Weber. Most interpretations of Weber describe his theory
as close to "postmodernist" views on values: values would be ulti-
mately ungrounded; they would be the effect of "free" decisions. This
view is currently characterized as the "decisionist" theory of values.
As existentialist philosophers argued, if man is free, he should be able
to act, behave, or believe without being determined by any reasons or
causes. Whether Weber endorsed a decisionist theory of values is
more than debatable. This interpretation of Weber was originally of-
fered by theorists such as Leo Strauss and Carl Schmitt. As they were
themselves men of strong convictions, they perceived Weber as a
relativist and maintained that his relativism was grounded in his sup-
posedly decisionist theory of values. On the basis of this interpreta-
tion, Leo Strauss described Weber as one the intellectuals responsible
for the crisis of values characterizing the modem Western world. The
same analysis was conducted by Erich Voegelin (1952) and by Allan
Bloom (1987). Ironically, postmodernist writers follow lazily the same
interpretation, but draw from it an opposite conclusion. They see Max
Weber as a forerunner of "postmodernism" and praise him for his
supposed decisionist-relativist theory of value^.^
What Weber actually says is very different. Yes, he states, some
ultimate values cannot be demonstrated as acceptable. It is impossible
for instance to show that it is interesting to try to describe reality as it
is. This postulate, which is the ground principle on which the whole
construction of science is built, is effectively a postulate or a prin-
ciple, as such undemonstrable, as the very words postulate and prin-
Introduction 7
ciple clearly indicate. In the same way, it cannot be shown that re-
specting the dignity of all is a good thing. Once the postulate was
proposed by the apostle Paul, however, writes Max Weber, it was not
to be forgotten again and started impregnating the history of the West-
em world for the next two thousand years (Weber, 1978 [1921]: 40).
And so, there are ultimate principles proposed on the market of ideas
by innovators. According to the interest they arouse in the audience
and to their fruitfulness, they are then selected or rejected. In the
eighteenth century, aesthetics were institutionalized as a discipline of
their own, says Weber, because the postulate which grounded the new
discipline, that the greatness of art works, the interest they are able to
arouse, in a word-their value-is the product of some features of the
works themselves, was launched on the market of ideas and selected
because it appeared as fruitful. Of course, while, according to Weber,
moral life is characterized by a "diffuse rationalization" (Durchration-
alisierung), this rationalization process can be thwarted or slowed down
by "historical forces." Equality, democracy are basic values; they have
been irreversibly selected. This does not say, though, that democracy
will never be threatened, nor that the principle of the equality of all
will not be damaged. Still, one cannot doubt the irreversibility of some
values. While slavery is still present in many parts of the world, no
one would dare any more to contend that it is a good thing. Aristotle,
by contrast, thought, as most of his fellow men, that slavery was an
indispensable social institution.
Weber's theory of values is thus far from Nietzsche's. Weber in-
sists heavily on the role of rationality in moral life and history. When
someone thinks that "X is good," he has good reasons for thinking so,
and he thinks so because of these reasons. When the Roman civil
servants felt powerfully attracted by Mithralsm, they had good reasons
of preferring it to the old polytheistic Roman religion. Human beliefs,
prescriptive or descriptive, should be considered as "understandable."
This is a basic principle of "comprehensive" sociology in Weber's
sense. How can this idea, central in Weber's work, that beliefs and
actions should be considered as understandable, be held as compatible
with a "decisionist" view of values according to which, as values
cannot be demonstrated true or false, they should be endorsed without
reasons? In other words, the current neo-Nietzschean interpretation of
Weber's theory of values misses its complexity, and is radically in-
compatible with the very idea of a "comprehensive" sociology. But it
8 The Origin of Values
category of "rational" theories that have, more than others, drawn the
attention of the public in the recent years. Their success is the symp-
tom of a demand. However, whether the supply meets the demand is
questionable. If it would, the influence of the irrational relativistic
theories of values would likely be weaker.
Another reason why it is interesting to go back to the question of
the explanation of values is that, while postmodernist views are domi-
nant on the subject, many interesting proposals are currently being
presented by social scientists and philosophers that are worth discus-
sion, although, as in the cases of the powerful theories developed by J.
Rawls, J. Habermas, or J. Wilson, major flaws can easily be detected
at first glance. Again, if the currently available rational theories of
values were sufficiently convincing, they would eliminate or at least
arouse serious doubts about the dominant irrational postmodernist views.
The question of the origin of values is both crucial and complex,
and has been the center of my intellectual interests for years. For
intrinsic reasons the notion of a "crisis of values" in our societies has
become common talk. The idea that in modern societies individuals
should and would "choose" their values freely has become an ac-
cepted idea. But I am motivated by another type of consideration. As
Durkheim, Simmel, Weber, Mead and other classical sociologists have
suggested, I can have the feeling that "X is g o o d only if I have the
feeling simultaneously that the other man should feel and think in the
same way. I cannot feel that an object is beautiful, contends Simmel, if
I do not sense that my feeling is caused by some features in the object
itself, so that I necessarily have then the feeling that other people
should think and feel the same way as I do about the object. In other
words, to Simmel as to Durkheim, Weber, or Tocqueville, value feel-
ings are not perceived by the social subject as subjective and are
effectively not merely subjective, with the exception of "tastes." But
tastes are perceived as tastes and recognized as such: I have no prob-
lem accepting the idea that many Germans do not like raw oysters,
while many Frenchmen like them. Now, leading contemporary theo-
ries of values see values, to put it in simple terms, as a matter of
"taste." The contradiction of this view with hard facts appears imme-
diately: who would consider seriously statements such as "liquidating
apartheid was a good thing," "polygamy is a good thing" as reflecting
mere tastes?
I have tried to develop and synthesize a number of intuitions pro-
Introduction 11
right hand would strike with a hammer on a nail held by the left hand,
each of the two hands ignoring what the other one is doing. I insist on
this point, because my main aim in the chapters comprising this vol-
ume has not been empirical. It has not been in other words to explain
why values are different here from what they are there, nor less to
guess what values will characterize the societies of tomorrow nor to
speculate, say, about the effects of globalization on values, but rather
to raise a problem which seems to me essential, whether it is called
methodological or theoretical, that is, how can values and valuation be
explained in a scientifically satisfactory way, and to suggest that a
good theory of values should produce a coherent synthesis between
various powerful insights derived from the sociological and the philo-
sophical literature, classical and modem, on the subject. The diversity,
the heterogeneity and the partially contradictory character of these
insights suggest that the task is both crucial and difficult. My objective
was, in other words, to defend and illustrate the principle that the
universal and contextual sides of values cannot be theorized indepen-
dently from one another. To use an analogy which seems to me able to
describe concretely this interdependence, values can be compared to
the solutions of a system of equations. The same equations produce
different solutions with different numerical values for the parameters,
so that reciprocally explaining the difference in the solutions is pos-
sible only provided one sees that the equations are the same and the
parameters different.
A few words can now introduce each of the following chapters and
essentially stress the way they are related to the questions described
above. Chapter 1 is devoted to a classification of the main theories of
values. Chapters 2 and 3 are mainly concerned with the criticism of
some particularly influential theories. The last four chapters develop
in a sequence of progressive steps the judicatory theory of values that
synthesizes in my mind a number of intuitions developed earlier, nota-
bly by classical sociologists, and propose applications of this theory.
More precisely, the main purpose of the chapter 1, "Explaining
Values and Valuation: A Question That Has Inspired Many Theories,"
is to bring some order to the numerous theories of values and valua-
tion proposed by social scientists and philosophers. A good way to
generate this order is to show that the theories can fill the cells of a
typology, a prior constraint being to identify the relevant dimensions,
the combination of which produces the cells of the typology. I hope I
Introduction 15
have succeeded, not only producing a valid typology, but also in show-
ing that its cells identify intellectual constants, so to say, in the sense
that it can be shown that theories belonging to each of these cells have
been constantly produced, from the classical to the modern age of both
the social sciences and philosophy. Given the typological objective of
the chapter, the theories evoked here are neither presented nor dis-
cussed in detail.
Chapter 2, "Rational Theories: Limits of the 'Rational Choice
Model,"' deals with one of the major "rational" theories used today to
explain norms and values: the so-called "Rational Choice Model"
(RCM). The uninformed reader, at this point, should be made aware of
the fact that in the expression "Rational Choice Model" the word
rational is used in an indigenous fashion, as an equivalent of "instru-
mentally rational." An action is "rational" in this sense if it can be
held as representing objectively a good way of reaching the goal the
subject is following. The chapter deals with the RCM generally, trying
to make the point that, while its supporters see it as a potentially
universally valid theory, in other words as the only one actually able
to explain indistinctly all social phenomena, and notably norms and
values, this claim seems ungrounded. The model is very effective in
explaining many social phenomena, and among them some categories
of norms and values, but definitely not all. The chapter tries to explain
why.
Chapter 3, "Relativistic vs. Naturalistic Theories: Their Interest and
Limits," deals with what I have qualified above as the dominant theory
of values, the "postmodernist" one, according to which values should
be considered as the emanation of singular cultures and, for this rea-
son, vary from one culture to another. As I said before, this model
raises two questions: Why is this theory dominant today? Is it accept-
able? In this paper I consider exclusively one of the variants of the
"postmodernist" theory of values: the "culturalist" variant. There are
also other variants, such as the previously mentioned neo-Nietzschean
or neo-Marxian theory of values. These variants have in common with
the culturalist version that they defend a relativistic view of values and
treat the perception people have of the reasons as to why they endorse
such and such values as illusory. These latter neo-Marxian or neo-
Freudian theories, however, have become more parochial and also less
influential than the culturalist variant. While they were dominant in
the sixties and seventies, they are now dominated, as far as the register
16 The Origin of Values
Notes
1. "The concept that is initially held as true because it is collective tends to become
collective exclusively under the condition that it is held as true: it has to produce
its certificate before we believe in its validity" ["Le concept qui, primitivement,
est tenu pour vrai parce qu'il est collectif tend I? ne devenir collectif qu'h condi-
tion d'Ctre tenu pour vrai: nous lui demandons ses titres avant de lui accorder
notre crCance," (Durkheim, 1979 [1912]: 624)l.
2. "Many commentators have contended. . . that Strauss completely misunderstood
Max Weber; but this seems never to have disturbed him or his followers," writes
rightly Lewis Coser (Coser, 1984).
3. See Elster, 1992 ; Schmidt, 1994.
4. "Taking collective opinions as an object of investigation and building a scientific
theory of opinion is the main objective of sociology" [ . . . on peut prendre l'opinion
comme objet d'ttude et en faire la science; c'est en cela que consiste principalement
la sociologie" (Durkheim, 1979 [1912], 626)J ; "Collective beliefs are states of
public opinion" ["les croyances . . . sont des Ctats de I'opinion" (Ibid., SO)].
Explaining Values and Valuation:
A Question That Has Inspired
Many Theories
More generally, the fideist and skeptical conclusions that are often
derived from Miinchhausen's trilemma makes it difficult to under-
stand that certain scientific statements are considered as definitely
true, or that magical practices are less efficient than scientific ones. In
other words, the two possible, most acceptable conclusions that can be
derived from Munchhausen's trilemma appear as contradictory with
evidence.
To get out from beneath these difficulties and contradictions, we
must accept the idea that scientific knowledge is gained by a dynamic
move going from principles to consequences and, in turn, from the
consequences to the principles. A scientific theory is normally grounded
on undemonstrated principles. If the theory appears wrong and if it
seems difficult to reconcile it with data, after a while, a normal step
will be to put the principles into question and possibly to replace them
by other principles. If, by contrast, some principles appear as capable
of being indefinitely maintained because a flow of adequate scientific
theories can be produced, then scientists will accept them and not put
them into question any more, provided no difficulties appear which
would cause scientists again to question their principles. What I have
presented here in a few words is the core of Popperian theory of
knowledge: principles are conjectures. They can become truths when
it does not appear relevant to put them into question. But they can
always be in principle criticized and put into question. We can sum-
marize this philosophy by saying that it endorses the third solution of
Explaining Values and Valuation 25
boat or the arrow move. This new principle, the Newtonian principle
of inertia, was from that moment definitely accepted. It seems unlikely
that it would be put into question.
This example of the principle of inertia is probably one of the most
spectacular the history of science provides of the back-and-forth pro-
cess between principles and consequences characterizing the produc-
tion of knowledge. Many others of course could be offered. To men-
tion just another equally instructive one, the Greeks drew from the fact
that the ratio S/D of the side S of a square of length one to its diagonal
D should be both even and odd the conclusion that S/D was not a
"number" but a "growth" ("grandeur"). In other words, they postu-
lated the existence of two types of quantities. It took centuries before
it was realized that the notion of number should not be restricted to
natural and relative numbers (that is, of type A/B where A and B are
integers) but should include "ir-ratio-nal" numbers as well, in other
words, numbers which cannot be expressed as the ratio of two inte-
gers. The principle, here the definition as to what numbers are, was
reformed because by so doing, it was possible to eliminate the obscure
notion of a "growth," which would not be possible to measure by a
number.
Finally, we believe in a scientific truth; we believe that "X is true,"
because X appears to us as grounded on a system of reasons, which
appear to us as strong, given the state of the discussion at a given time.
Before the principle of inertia was elaborated, the physicists had rea-
sons to believe in the theory of turmoils: they could perceive it as
strong. Once the principle of inertia was formulated, the reasons sup-
porting the earlier theory of turmoils were weakened. So, we believe
in a theory when it appears to us as supported by stronger reasons than
any alternative theories. We believe the theory that "the earth is round
rather than flat," because it is grounded on stronger reasons than the
theory "the earth is flat." Our belief is grounded on the fact that we
see that all the arguments supporting the theory "the earth is flat"
would be definitely weaker.
J. Rawls. In that sense, they can be qualified not only as rational but as
absolutist.
On the whole, it is not exaggerated to say that a good part of the
theories of norms and values appear as dominated by the objective of
solving Munchhausen's trilemma, as though the theorists of norms
and values had more or less clearly seen this theorem, and as though
they had picked up the various interpretations given to it.
I will not attempt to present exhaustively these theories, nor to
criticize them. My objective is rather, on the one hand, to support the
conjecture according to which Miinchhausen's trilemma has oriented
the construction of value theories by evoking some influential theories
of axiological beliefs, and, on the other hand, to show that
Munchhausen's trilemma can be solved by a better theory than the
fideist, the skeptical theory (either in its decisionist or in its causalist
form), or the absolutist rational theory.
"Fideist" theories start from the assumption that norms and values
can be grounded on principles of absolute validity, in other words, on
principles the validity of which need not be demonstrated.
The phenomenological theory of values developed by Max Scheler
belongs to this category. It is certainly one of its most interesting
illustrations (Scheler, 1954 [1916]).4 According to Scheler, we be-
come aware of values thanks to a sense of values that would be com-
parable to our sense of colors. Values have to him an intuitive charac-
ter that phenomenological analysis can, accordingly, reveal. Of course,
this sense of value can be mistaken. But our sense of colors can also
be mistaken, for instance, under the effect of some diseases, for ex-
ample, jaundice.
The main weakness of this theory can easily be identified: the exist-
ence of a "sense of values" cannot be identified as easily as the exist-
ence of a sense of colors. The intuitionist conception of values de-
fended by Scheler rests on the observation that some values appear as
universally endorsed. It is true that a heap of garbage is universally
seen as negative. It is true also that value statements are often con-
ceived by social subjects as being objectively grounded. When I state
that "X is good, legitimate, beautiful," I have often the feeling that I
state so because X is actually, that is, objectively, good, legitimate or
Explaining Values and Valuation 29
beautiful. For this reason, I have the impression that other people
should also be convinced of the validity of the statement. Simmel says
that we cannot find an object beautiful unless we have the feeling that
our impression is provoked by some attributes of the object itself
(Simmel, 1989).
These two observations, namely that some value statements appear
as universal and that others are normally felt by a subjebt as "objec-
tive" are probably responsible for Scheler's theory according to which
we would have a sense of values comparable to our sense of colors.
Many value statements cannot, however, be related to this conjec-
tural sense of values. Supposing we accept its existence, a value state-
ment such as "this policy, which aims at reducing unemployment, is
good (or bad)" is certainly not its mere effect. Or assuming we have a
general sense of justice or of fairness, evoking this sense cannot ex-
plain why we consider such and such distribution of goods as just or
fair, unjust or unfair.
Although I do not want to discuss further Scheler's theory, it must
be recognized that it has a unique feature: it attempts to combine the
idea of the "objectivity" of values (values would be essences which
phenomenological analysis would be able to identify) with their his-
torical and social variability. For this reason, Scheler began as a phi-
losopher but ended as the grounder of the so-called "sociology of
knowledge." But while he raised important questions, the answers he
gave them are debatable.
The important point for our discussion is that the Schelerian theory
of values illustrates the theories that give the Miinchhausen trilemma a
fideist solution: it attempts to show that value statements rest upon
intuitively evident normative data. In the same way that our sense of
colors make us sure that the leaves of trees are green, our sense of
values would tell us that such and such behavior or state of affairs is
good or bad, acceptable or not, etc.
Incidentally, I can note a point which would be worth further devel-
opment, namely that the notion of religious revelation introduces also
the idea that axiological truths, as they would have been revealed,
would have no need of demonstration. Scheler himself considered
reciprocally that revealed religions express symbolically the notion of
a sense of values which phenomenological analysis could detect. Be-
fore Scheler, Benjamin Constant and Tocqueville also defended the
idea that religions express in a symbolic fashion axiological truths,
30 The Origin of Values
Decisionist Theories
other. One point is certain, at any rate: if this theory could be reduced
to the decisionist type, one would hardly understand why Weber cre-
ated the notion of "axiological rationality" (Wertrationalitat),nor how
he could have made the idea of Verstehen a basic point of his method-
ology. Understanding why people do what they do, believe what they
believe, is a crucial task for the sociologist, he contends. This means
that, to Weber, actions and beliefs, far from being "absurd," unmoti-
vated, are grounded in the eyes of actors themselves. He never ana-
lyzes beliefs and actions either as absurd or as illusory. In several
circumstances, he criticizes Marx's or Nietzsche's theories according
to which normative beliefs would be illusions, as when I believe that
"X is good," while X is merely congruent with my psychological
interests or with my class interests. Generally, Weber assumes that
people's beliefs are grounded. The causes of the beliefs are their mean-
ing to the actor, in other words, the reasons which they have to en-
dorse them. As to the expression "axiological rationality," it evidently
evokes the Kantian notion of "practical reason."
It can be noted incidentally that this decisionist interpretation of
Weber is mainly due to the influence of Leo Strauss (1953). By fight-
ing Weber's supposedly decisionist theory, Strauss tried to reinforce
his own "naturalistic" theory. For the same kind of reason, E. Voegelin
has been negative about Weber: he would have developed a decisionist
theory of values detrimental to the only acceptable theory, the natural-
istic one; he would be the father of modem relativism. Allan Bloom
(1987), a disciple of Strauss, has even contended that Weber would
have corrupted by his ideas on values New York taxi drivers and
converted them to n i h i l i ~ m . ~
In France, Raymond Aron, who has attracted attention to the impor-
tance of Weber, has interpreted Weber's lecture, given at the end of
1918, Wissenschaft als Beruf, as a tragic message, insisting on the
mutual incompatibility of values. The tone of the lecture was quiet and
peaceful. It can be read now rather as a message of tolerance: cultures,
notably national cultures, such as the French and the German, bear
various values; one should recognize that it is often not easy to deter-
mine which are better. It seems to me unlikely that, at the end of 1918,
Weber could really have delivered a tragic message insisting on the
struggle between values. Moreover, this conference cannot be consid-
ered as containing an exposition of Weber's theory of values, which is
rather contained in his essays in the sociology of religion.
32 The Origin of Values
terizes their theories, Marx, Freud, Durkheim, Pareto and others have
contributed to making the notion of "false consciousness" in this meta-
physical sense a current one, which was to be used in an uncritical
fashion, as though the notion could go without saying. Under their
influence, an entirely new "psychology" appeared on the market, as
rightly noted by R. Nisbet (1966). To this must be added, that while a
Marx or a Freud or a Durkheim introduced the notion (or its equiva-
lents) in a very careful fashion, this was not always the case with their
followers. Thus, Durkheim makes the point that religious beliefs are
symbolic and more or less confuse expressions of realities well per-
ceived but ill identified by social subjects. He introduces here the
notion (the notion not the word of course) of "false consciousness" in
the second of the meanings I have identified above. But Durkheim's
followers have often translated his ideas into the assumption that so-
cial structures can mechanically introduce into the mind of people
beliefs which people see as grounded while they are not. Possibly, the
importance of the new "causalist" psychology has not been well real-
ized.
Affective Causes
Pareto and Freud are probably the most important writers among
those who attribute affective causes to axiological beliefs.
For the two of them, feelings of value are distorted expressions of
affective causes. I believe that X is good because an instinct pushes
me to feel so. But, as I do not want to see myself as moved by
instincts, I cover my belief with what Pareto calls a "logical varnish"
("un vernis logique" ): I construct arguments (Pareto's "derivations" )
to "demonstrate" that X is good and justify my belief to my own eyes.
Alternatively, I can also look for theories available on the market that
would justify my feeling, and endorse them. Example: I am poor and
suffer from this situation, so I will endorse easily theories supporting
the idea that personal merits play a limited role in the determination of
social status, that birth, social relations, and social manipulation are
the basic determinants of status, etc. I am rich: I will endorse rather
.easily the idea that status is basically determined by personal merits.
Where Pareto talks of "vernis logique," Freud talks of "rationaliza-
tion."
Such processes of course do exist and are easily observable. But the
Explaining Values and Valuation 35
Socio-Functional Causes
Social Causes
the idea that beliefs would be caused by reasons, or, in other words,
that mental states could be caused by mental states; they had the
impression, on the contrary, that mental states should be explained by
non-mental causes, and consequently that, when the social actors have
the impression that their beliefs are the effect of reasons, this impres-
sion would necessarily be illusory. Among the other difficulties gener-
ated by these diffuse positivistic ideas, one is worth being evoked:
they produce a dual version of the mind in the sense that the "causalist"
theories of beliefs view scientific beliefs as produced by reasons and
ordinary beliefs by irrational causes. Durkheim, here again, appears as
transcending this difficulty: to him, ordinary beliefs, as well as philo-
sophical or religious beliefs, are generated by the same processes as
scientific beliefs. True: the latter are clearer and more precise, while
the former are more confused. But all categories of beliefs have to be
grounded on reasons in the mind of actors. Otherwise, actors would
not endorse them except temporarily. Still, these complexities of
Durkheim's thinking were less retained by his followers than the as-
sumption that society is able to structure the contents of the minds of
social actors. Generally, Freud, Marx or Durkheim owe much of their
influence to the fact that they are currently perceived as having been
able to identify the "material" (biological, psychological, or sociologi-
cal) causes of the mental states of social actors.
The influence of positivism reappears today in the attempts made
by the so-called "cognitive sciences" to explain mental states by their
supposed biological, emotional, or neurological infrastructure (see
Sperber, 1996).
Beside what I have called the fideist and the skeptical theories of
normative and axiological beliefs, another category of theories can be
identified: the rational theories. Some have been proposed by sociolo-
gists, others by philosophers, others by psychologists. But they all
have a feature in common: they suppose that subjects endorse
axiological and normative beliefs because they have strong reasons of
doing so. Moreover, these reasons are not private. They are, on the
contrary, public, in the sense that as soon as an actor has the feeling
that his beliefs are grounded on strong reasons, he has also the feeling
that other people should accept these reasons as strong. In other words,
Explaining Values and Valuation 41
Utilitarian Theories
Functional Theories
Cognitivist Theories
Kant
Rawls
Table 1.1
Typology of the Theories of Moral Feelings
1. Irrational theories 1.1. Decisionist (Sartre,
Weber's war of gods
seen by Leo Strauss)
1.2. Fideist (Scheler,
intuitionnism)
1.3 Causalist 1.3.1 Psychic causes (Pareto,
Nietzsche, psychological
cognitivism, behaviorism)
1.3.2 Social causes (Marx,
Durkheim, Nietzsche, post-
modernism, neo-Marxism)
1.3.3 Biological causes (Ruse,
Wilson)
2. Rational theories 2.1.Absolutist 2.1.1. Universal (Kant, utilitari
anism)
2.1.2. Partial (Rawls, sociologi
cal functionalism)
2.2. Contextualist
(Weber's axiological
rationality, neo-
Weberianism, moral
cognitivism)
The objections that have been raised against the cognitivist theories
which I have just evoked show that none of them is able to explain
axiological feelings as they are observed. The utilitarian theories are
Explaining Values and Valuation 47
The most profound and broadest geniuses of Rome and Greece have never been
able to come to this so general, but at the same time so simple idea, of the
similarity of men and of the equal right to liberty that each of them bears in being
born; and they have spent energy to prove that slavery was in the nature of things,
and that it would always exist. Furthermore, the available evidence shows that
those in Antiquity who had been slaves before becoming free citizens, and of
whom several have left us significant writings, had the same viewpoint on the
institution of ~ l a v e r y . ~
When the social conditions are very unequal and when inequalities are everlasting,
individuals become progressively so dissimilar, that one would think that there are
as many different types of mankind as there are social classes; one discovers them
separately and, losing sight of the general link that gathers them all in the vast
category of mankind, one pays attention to certain men, but not to man.
escape the "war of gods" and the "polytheism of values," it does not
follow that these principles cannot be rationally discussed, that they
cannot be rejected by reality. On the contrary, the competition be-
tween principles is exposed to processes of rational selection, generat-
ing the processes of "rationalization" of the moral life which Weber
evokes repeatedly.
reality. As in the case of scientific ideas, new moral ideas are devised
and proposed by innovators. That Socrates or Voltaire have become
major figures in the history of mankind comes from the fact that they
have been the vectors or moral innovations which were selected by
social evolution.
Of course, this innovation and the selection processes are thwarted
by passions and interests. Normative principles are more exposed than
descriptive principles to the passions and interests of men because
their discussion is public. For this reason, one has easily the impres-
sion that the strength of normative ideas in the social selection process
results from the strength of the groups defending them. That the selec-
tion of normative ideas is associated with social-ideological conflicts
and appears, as suggested by Coser (1956), as facilitated by social
conflicts, is one thing. That the moral truth would be nothing else than
the outcome of this struggle and would simply reflect the relative
strengths of the groups in competition supporting the various view-
points is another. "The reasons of the stronger are not always the best"
against La Fontaine's aphorism, "la raison du plus fort est toujours la
meilleure."
To come back to Munchhausen trilemma, it has never inhibited the
production of solid descriptive ideas. The same is true of prescriptive
ideas.
A last question should be raised which I cannot explore in its full
extension. Is this process of rational selection of normative ideas a
characteristic of democratic societies or does it apply to all societies? I
would rather choose the latter answer. Weber and Durkheim have
indicated, rightly, I think, that "rationality" is a fundamental feature of
man. Hence, an idea which is positive from a rational viewpoint, has
an intrinsic force. This explains for instance that magic remains a
serious competitor of science exclusively on the issues that science
has not succeeded covering. All societies and cultures recognize this
point. The same is true of axiological ideas. This is one of the impor-
tant points well defended by Durkheim in his Formes e'le'mentaires.
In virtue of the Miinchhausen's principle there are no ultimate truths,
except vague and ill-determined ("respect reality, describe reality as it
is" is the ultimate value that grounds science, "respect all men equally"
is the ultimate value that grounds morals), but the process of rational-
ization characterizes all cultures.
Explaining Values and Valuation 55
Notes
action can always be considered rational in the very special sense RCT
gives to the notion of rationality is another. If all actions and further
all social phenomena could be validly explained by RCT, this would
be fine. But can they? In the same way, it would be fine if I lost my
keys under a public light where they would be much easier to find
than in the surrounding darkness. But this does not demonstrate that I
effectively dropped my keys under the public light. Thus, an important
question is whether RCT is effectively general, as contended by its
followers. If yes, the discussion can stop at this point. If not, the next
question is whether RCT can be revised to make it more general.
Can we legitimately apply RCT to all research situations and to all
problems? The answer to this latter question is "no," for a simple
reason: RCT assumes that individual action is instrumental, namely
that it has to be explained by the actor's will to reach certain goals.
Now, action can be non-instrumental, as most sociologists have recog-
nized. Schiitz through his distinction between Weil and Wozu Motive,
Weber through his distinction between instrumental and axiological
rationality, have stressed that action is not always instrumental. If this
is true, RCT cannot claim to be a general theory of action. Thus, one
cannot apply RCT notably in cases where an actor does X because he
believes in Z and that Z implies his doing X independently of the
consequences of X. And even when action is instrumental, it can
mobilize beliefs which need being explained and which normally will
not be explained by RCT. Thus, French authorities for many years
were more reluctant than the Dutch to accept the idea of using metha-
done to curb drug addiction. Why? Because the Dutch thought metha-
done was an adequate means while the French thought it was not. Why
had the French and the Dutch different beliefs on this point? RCT is
of little help here.
Of course, these objections are not new and there are two traditional
ways out of them:
cation of his own RCTheory, assume that voting maximizes for the
voter some cost-benefit balance. At this point, the RCTheorist should
explain the "false consciousness" he attributes to the actor: why does
the subject think he votes for one reason while he really votes for
another? But how does the RCTheorist know that the consciousness of
his voter is "false"? I am not saying here that what the actor himself
thinks and says of his own motivation is the ultimate truth. It is rather
a piece of information among others. What I am saying is (1) that the
actor's statements about his motivation are facts which, as any fact,
should be taken into consideration and explained; (2) that the
Nietzschean, the Marxian, or the RCTheorist who attributes to the
behavior of the observed subject causes which the latter does not
endorse should explain his "false consciousness," notably if he rejects
the Friedmanian epistemology. Hence, in contradiction with Coleman's
main argument for RCT, RCT creates rather than eliminates black
boxes in many circumstances (Boudon 1998).
On the whole, none of the strategies currently used to make instru-
mental non-instrumental actions appears as very convincing. They raise
more questions than they solve. In other words, non-instrumental ac-
tions cannot easily be converted into instrumental actions. Consequently,
RCT cannot be held as general.
Beside this basic objection, "empirical" objections can be raised
against the RCT, namely that it has never succeeded in explaining
satisfactorily important classes of phenomena. Voting is the best known
of these classes. People vote, though any individual vote has a practi-
cally zero probability of having an effect on the outcome of an elec-
tion. In this case, the anticipated consequences of individual action
cannot easily be taken for the cause of the action. An enormous amount
of literature has been devoted to this problem, that tries to reconcile
RCT with the hard fact of voting (see for instance Overbye [1995]).
Many other puzzles in the same style could be evoked: why do people
appear so easily upset by political corruption and so sensitive to it?
This familiar observation cannot be easily explained by RCT since
political corruption has, in the case of Western democracies at least, a
negligible and invisible effect on the well-being of citizens. In other
words, the rejection here is implausibly the effect of the consideration
by social actors of the consequences of corruption on themselves.
Allais (1953) and a number of authors after him have revealed another
Achilles' heel of RCT (if I may make Achilles' heels plural) and
Rational Theories 61
unique type: not only instrumental, but egoistic. I agree with M. Somers
that such an assumption has a metaphysical flavor.
At one point in his Old Regime Tocqueville wonders why at the end
of the eighteenth century (Tocqueville [I8561 1955) French agricul-
ture remains stagnant at a time when agriculture is flourishing in En-
gland. This is particularly puzzling since the physiocrats, who develop
the view that modernizing agriculture is the main path to growth, are
politically very influential in France at this time. Tocqueville's expla-
nation: administrative centralization is the reason "civil servant" posi-
tions are more numerous and hence more easily available in France
than in England. Also, French centralization makes serving the King
in France a unique source of prestige, influence, and power; conse-
quently, other things equal, landlords are more easily incited in France
than in England to leave their ground and buy a royal position. In
England, by contrast, being an innovative landowner can not only
produce local respect and prestige, it can also open the way to
Westminster. This macroscopic difference between England and France,
summarized by Tocqueville in his notion of "administrative central-
ization," explains why landlord absenteeism is much greater in France
than in England. Further, landlord absenteeism is the cause of the low
rate of innovation: since their interests are at Court, landlords them-
selves have little motivation to innovate; as to the farmers who run the
landownerships, they would have a motivation to innovate, but hardly
the capacity for doing so. Finally, the low rate of innovation is respon-
sible for the stagnation.
Tocqueville uses here "methodological individualism" (MI). The
macroscopic difference between France and England is explained as
the effect of individual decisions made by the landlords. The indi-
vidual decisions are analyzed as taken, not by "angels," but by men
belonging to social contexts. The parameters characterizing the French
and the British contexts are themselves the product of a long history.
This point gives me the opportunity of stressing that MI does not
imply solipsism as soon as individual decisions are analyzed, as here
by Tocqueville, as affected by the parameters characterizing the con-
text. Finally, Tocqueville uses here what we call RCT: by leaving
their land and serving the King, the landlords gain in influence, pres-
tige, etc. In England by contrast, it is a better strategy to appear locally
64 The Origin of Values
more than soldiers (Smith, 1976 [1776])? Why did the people of Lon-
don not try to exert pressure on political power by street gatherings as
frequently as the Parisians (Root, 1994)? Why are magical beliefs
more widespread in some societies than in others (Boudon, 1994b)?
Why was methadone used much earlier in Holland than in France
(Bergeron, 1999)? All these questions have been convincingly an-
swered in works which use MI and an open theory of rationality rather
than the special figure of rationality used by RCT. In particular it can
be noted that, in his Wealth of Nations, Adam Smith (1976 [1776]),
RCT's spiritual father, solves the above question about miners and
soldiers using what I call the CM rather than RCT (Boudon, 1996,
1998).
The second remark, which I cannot develop, is that CM excludes
radically solipsism: I cannot perceive as strong the reasons leading me
to endorse a statement, "X is good, legitimate, right, true, etc.," with-
out conceiving these reasons as grounded and hence as intersubjectively
valid (Boudon, 1995b).
Finally, a crucial question is raised by M. Somers: what is a good
theory? Sometimes, good scientific theories use mathematical language,
and are derived from a general theory, etc. But such attributes are not
components of a good theory generally. Otherwise, physical theories
would be good, but biological theories bad, since the latter make little
use of mathematics and are hardly deducted from a general theory.
Celestial mechanics is not the model to be followed by all disciplines.
A good scientific explanation of a phenomenon P is rather a set {S) of
statements meeting three requirements: (1) that all s belonging to S are
acceptable, (2) that {S}=>P, (3) that relevant facts are not arbitrarily
ignored. They are satisfied for instance in Tocqueville's above ex-
amples: all statements are acceptable (simple psychological statements,
empirical statements congruent with observation, no "black box" con-
cepts in the components of the statements, etc.); moreover, {S)=>P
(P=stagnation of French agriculture, enthusiasm of the peasants for the
idea of Reason); finally, the theory explains not only the observed
behavior but the verbal statements of actors as they have reached us.
Notes
1. Firstly, that national, ethnic and other social groups tend to have
a distinctive culture of their own;
2. Secondly, that these cultures and subcultures include systems of
values which cannot be ranked, for the reason that values cannot
be grounded objectively;
3. Thirdly and consequently, that endorsing multiculturalism would
imply endorsing value relativism, and reciprocally.
Communitarianism
The need for identity exists, beyond doubt. Moreover, it is true that
personal identity can only develop within a social context. As R. Hardin
(1995) contends, Kurt Tucholsky probably committed suicide at the
age of thirty-four because, after leaving Germany for political reasons,
he emigrated to Switzerland and felt that he could not express his
identity in the Swiss intellectual, social and political context, at least
as he saw it.
Or to take another example, also mentioned by R. Hardin,
communitarians insist on the importance of the fact that Russia and
Yugoslavia exploded into a plurality of ethnical entities as soon as the
communist regime they had been subjected to collapsed.
Do not such big facts demonstrate, claim communitarians, that be-
longing to a community is a basic human need? Do not they show that
communities are totalities, that people identify themselves to these
totalities? That cultures are incommensurable systems of values? That,
as they are incommensurable, these values are "arbitrary"? But, while
values are ungrounded, they are vital to individuals, since personalities
can only develop within a definite cultural context.
I am (by far) not the first to raise this question. On the contrary: the
discrepancy between the relativistic theories according to which val-
ues would be illusions, and the conviction of people that they are not,
has produced an interesting reaction on the part of some philosophers,
sociologists and anthropologists, who have proposed in recent years
naturalistic theories of values. I cannot consider these theories in de-
tail. But it may be useful to spend a short time on one of them in order
to check the potentiality of this naturalistic orientation.
moral convictions change over time, but it can explain this change
more easily than other types of theories. The fact that science is his-
torical, that a statement that was treated yesterday as false is treated
today as true was never held as an argument against the possibility of
reaching truth in scientific matters; in the same way, in moral matters,
the fact that some institutions were held as bad yesterday and are now
considered as good is not an argument against the fact that moral
evaluations are grounded on strong reasons in the minds of people.
Moreover, it is hard to see how normative irreversibilities, as scientific
irreversibilities, could be explained in a satisfactory fashion, if not
rationally. The Newtonian principle of inertia is considered as irre-
versibly valid because it is objectively better than the principles it
replaced. In the same fashion, as noted by Tocqueville, we will never
again hear somebody explaining that he or she enjoyed being the
spectator at a capital execution. Capital punishment can be reintro-
duced depending on political circumstances; but we will never be able
to experience and express the feelings of Madame de SCvignC.
By so saying, I do not contend that there are no historical contin-
gencies. On the contrary, for, if there were no contingencies, there
would be no innovations, either scientific or moral. Tomorrow, totali-
tarian regimes can reappear. But unless human memory is destroyed,
the idea that democracy is better than despotic regimes will remain
present in human minds. "I1 n'y a pas plus de recul d'idCes que de
recul de fleuves," wrote Victor Hugo ["as rivers, ideas never go back-
wards"].
The argument that change in moral values confirms relativism rests
finally on a fallacy. The research of truth, positive or normative, is
historical. But truth, either moral or positive, is not historical. The fact
that science has a history is not an argument against the possibility of
scientific truth. The fact that morals have a history is not an argument
in favor of moral relativism.
If our moral convictions rest upon strong reasons, why then is the
similarity in this respect between the positive and the normative ill-
perceived?
The main reason is that it contradicts many influential traditions,
which I have already mentioned.
84 The Origin of Values
On the whole, the idea that moral and generally axiological state-
ments cannot be objectively grounded was treated as evident by a host
of influential thinkers who differ from one another in all other re-
spects. Strangely enough, empiricism converges in its interpretation of
moral feelings with the irrational sociological theories inspired by
Marx and Nietzsche or derived from a misreading of Durkheim, and
also with the irrational psychological theories inspired by Freud. All
these theories contend that moral convictions cannot be grounded ra-
tionally. To them, the reasons the subjects see as grounding their
beliefs should not be considered as their genuine causes, but as mere
"rationalizations," in other words, as false reasons, as the effect of
"false consciousness." Existentialists could be added to the list, since
they consider that moral beliefs cannot be grounded and propose to
treat the reasons given by subjects as illusory. Moral decisions would
be "absurd" in Sartre's vocabulary. In other words, they would be
ungrounded and ungroundable.
Thus, empiricism, Marxism, Freudianism, neo-Durkheimianism, ex-
istentialism, postmodernism and other-isms, different as they are from
one another in most respects, agree on one point, namely that moral
and generally normative convictions cannot be rationally grounded.
Consequential Reasons
But the idea that normative beliefs cannot be grounded has also
been reinforced by another factor worth being stressed, namely the
relative weakness of the rational theories of moral and normative feel-
ings developed notably by the social sciences. This will be my second
main theoretical point.
These theories are rational in the sense that, to them, the reasons on
which the normative beliefs of a social actor are grounded in his mind
are their causes. But, with the exception of Max Weber and a few
other writers, most rational theories of axiological beliefs developed
by the social sciences propose to explain the moral evaluation of an
action or of a state of affairs by their consequences. Some of these
theories are powerful, but, because of this consequentialist restriction,
none of them can be considered as a general theory of normative
feelings. Consequently, they cannot efficiently counterbalance the pow-
erful influence of the culturalist and generally of the irrational theories
of normative beliefs.
Relativistic vs. Naturalistic Theories 85
Axiological Reasons
from the fact that they serve the interests of the group and its mem-
bers, while the universal values would have no direct relation to inter-
ests. Hence, the competition between the two types of values would be
unequal. Universal values would have a strength, to him, exclusively
in the case of dyadic relations, where instrumental and axiological
rationalities appear as convergent: I keep the promise I have made to
B, because I have an interest in doing so: preserving my future ex-
changes with B. But in the general case, instrumental and axiological
rationality would be divergent.
This analysis, although profound, is, in my opinion, unfinished.
While Hardin is himself a follower of the so-called Rational Choice
Theory, he recognizes-at least implicitly-that universal values also
exist, and that they are not grounded in individual interests. In Weber's
vocabulary, he recognizes that social actors are moved not only by
instrumental rationality but also by axiological rationality. But, as a
follower of Rational Choice Theory, he assumes that, in the last in-
stance, individual interests are the main factor accounting for behav-
ior, beliefs, attitudes, and actions: to him, actors are basically moved
by their individual and group interests. Hence, he concludes that
axiological rationality would be structurally weak, while instrumental
rationality would be structurally strong.
Hardin, however, refutes himself unwillingly. Namely, he has no
doubt that he is himself on the side of Antigone, not Creon, and that
he has negative feelings against the institution of the duel that he
analyzes beautifully, and against the massacres in Yugoslavia. He also
sees clearly that most people would have the same feelings. No spec-
tator of Antigone has ever been on the side of Creon. No spectator of
the Merchant of Venice has ever been on the side of the judge Portia.
And, if universal values are often hurt in the short term, they can also
win in the long term. The institution of the duel has disappeared, as
well as the hierarchical social organization it reflected.
At this point, we are confronted with a question symmetrical to
Hardin's question. To the question, "where does the strength of par-
ticularistic values come from?" he answers : individual interests medi-
ated by group interests. But he has no answer to the question: "where
does the strength of universal values come from ?"
It comes actually from two sources: this will be my fourth theoreti-
cal main point. The strength of universal values comes firstly from
"axiological rationality": we are all convinced that Portia and Creon
Relativistic vs. Naturalistic Theories 91
were wrong. We are all convinced that they were wrong on the basis
of strong reasons, of strong axiological reasons. Secondly, one should
be aware of the fact that while in some situations the interests of social
actors are relevant parameters, in many other situations they are not.
When the sixteenth-century aristocrats had to have an opinion on the
duel, their class interests and, through their class interests, their indi-
vidual interests were a crucial parameter: they were normally con-
cerned with the question of how to maintain their social superiority in
a situation where they were threatened; and, consequently, they ap-
proved generally an institution which served their interests. By con-
trast, the interests of, say, the nineteenth-century bourgeois did not
play any significant role in their evaluation of the duel: they were not
concerned by the institution, and they evaluated it on the basis of
universal axiological reasons. So, after a while, once the aristocracy
had lost any significant social role, the duel appeared as meaningless
to a majority of people and became, moreover, ridiculous, because it
was more and more frequently practiced by parvenus, who wanted to
demonstrate, by fighting in duels, a purely imaginary social superior-
ity. Moreover, in the last decades of its life, the duel entailed neither
big risks nor big courage. In the more than twenty duels in which
Clemenceau was involved, only one of the fighters was slightly hurt.
In other words, the social actors are not only actors pushed on all
subjects by their interests. They are also spectators. But spectators are
also actors-as spectators-on the social and political stage. This is one
of the basic weaknesses of Rational Choice Theory. It does not take
into account the fact that, in many situations, social actors have on
such and such subject an opinion that is not dictated by their interests,
for the very simple reason that they are not directly concerned by the
subject, that the subject does not cross their interests. In other words,
Rational Choice Theorists forget-or ignore-the existence of public
opinion. Now, public opinion does exist and has beyond a doubt a
strong influence on political, social, and historical processes. And it
appears as being guided by axiological rationality (though it is obvi-
ously not clearsighted in all circumstances). For public opinion is
nothing else than the aggregation of opinions of actors who, on many
issues, are not directly concerned in their interests by the issues in
question and who, consequently, do not form their evaluations in ac-
cordance with their own class and individual interests, but rather by
axiological reasons. That I am not concerned in my personal or in my
92 The Origin of Values
Notes
Weber as Nietzschean?
Weber as Kantian?
universal practical reason. Nor can he probably accept the idea that
Weber's Kantianism could be literal. Consequently, he fails to see
what the notion of axiological rationality really means. This can be
easily sensed in his tone: a strange, un-understandable, on the whole, a
confused notion, he suggests.
First type: I wear a coat because the weather is cold and because
a coat is an easy way of struggling against cold.
Second type: I do not steal because I believe one should not steal.
Third type: I shake his hand because he is French and because
Frenchmen shake hands when they meet.
Fourth type: She protected him because she liked him.
black box. The sociologist who learns that some ritual can only be
explained by the fact that the members of a far tribe are governed by a
"primitive mentality" would perhaps ask the biologist to explain to
him why the brain of the so-called primitive is wired in a different
fashion from ours. At any rate, explanations using concepts as "primi-
tive mentality," "national spirit," "socialization," etc., are not "final." I
am not saying they should be rejected, merely that they suggest imme-
diately further questions: what are the mechanisms behind the words
"socialization," "primitive mentality," etc.
One of the main appeals of the set of methodological principles
advocated by Weber is that it produces "final" explanations without
black boxes, as well as being realistic.
political authorities do not care enough about enforcing the basic prin-
ciples of the social link. All these puzzles cannot be explained without
the category of axiological rationality.
Thus, the category of "axiological rationality" invites the develop-
ment of a theory which would make the functionalist theory, the Ra-
tional Choice Theory, but also the so-called "exchange theory" (Heath,
1971; Homans, 1958) or the contractualist theory (the two important
latter theories will be only mentioned here without further develop-
ments), particular cases of this theory.
and also the fact that, because knowledge is circular, the trilemma is
not contradictory with the possibility of reaching truth and objectivity.
This example of democracy suffices to show that a value statement,
"X is good," can be as objective as any positive statement. If the
feeling that "democracy is a good thing" were not objectively grounded,
one would not observe a consensus on the subject. One would not
understand that against the principles-basic in international relations-
which require respect of the sovereignty of foreign states, pressures on
foreign governments to the effect of installing or developing democ-
racy is generally well understood and approved by the public opinion.
How could these collective feelings be otherwise explained? Theory
and empirical sociology converge here. (Of course, I am not saying
that consensus is a proof of truth, but only that when consensus ap-
pears, it has to be explained by making it the product of reasons likely
to be perceived as objectively strong).
An objection can be made here: namely that democracy was cer-
tainly not always considered so. Even before the First World War, the
universal voting right was criticized. A. Pareto, for instance, saw in
this right another of these symptoms of human craziness he liked to
collect and prophesied that it would generate social chaos. Does not
this show that our belief that democracy is good is a product rather of
socialization than of reason and that it has little to do with our beliefs
in scientific statements? The fact that moral truths are historical, how-
ever, is far from being a deadly objection against the theory of
axiological beliefs which I develop here on the basis of Weber's "value
rationality."
Consider scientific beliefs. Aristotelian physicists believed that any
physical move is produced by some force or set of force^.^ This sheet
of paper moves because I apply force to it. If I would not apply force,
it would not move. This point seems so trivial that insisting on it can
easily appear bizarre. What I want to say by evoking it is that Aristote-
lian physicists had strong reasons for believing that any move is the
effect of some force. But they drew from this statement conclusions
that appeared acceptable to them and are unacceptable to us, for in-
stance that when a ship keeps on sliding after the wind has suddenly
fallen, some force should be responsible for this move. They tried
consequently to figure out what this force could be and introduced the
assumption that the move of the ship produced a turmoil. This turmoil
was for its part supposed to produce force pushing the boat, which was
114 The Origin of Values
finally held responsible for the fact that it kept moving. But after a
while, Buridan came and said: "if the argument were right, the hypo-
thetical turmoil would have the effect that the straw on a straw heap
should fly in opposite directions depending on whether the heap is
located at the front or at the back of the deck."9 As the direction where
the straw flies does not actually depend on the location of the straw
heap on the deck of a ship, Aristotelian physicists came-slowly-to
the conclusion that the principle according to which there would be no
move without force producing it was false. Finally, physicists came to
a new principle, which we now consider as evident, namely, that a
body that moves needs a force to be stopped, exactly as a body not
moving needs a force to be brought into move. This is the so-called
"principle of inertia." The feeling of obviousness that it produces to-
day in our mind is well the product of history.
The same kind of story could be related on normative as well as
positive statements.
As reported by George Trevelyan (1993), Voltaire, before he came
to England, could not conceive that a society could function orderly
when writers were allowed to publish what they wanted. And, to re-
turn to my earlier example, as long as actual democratic regimes or at
least political regimes embodying some of the features of what we call
democracy did not exist actually, they were not conceived; no one
could imagine them, nor a fortiori give them a positive value. Then, at
the occasion notably of civil struggles in England in Cromwell's time
(I follow again Trevelyan here), the principle of the separation of the
executive and legislative powers appeared and its effects started being
evaluated and positively appreciated. Much later, theories of democ-
racy were developed by analysts such as Montesquieu, John Stuart
Mill, and others who presented the principle of the separation of pow-
ers as crucial. At this point, it started being perceived as evident, in the
same fashion as the principle of inertia appeared as evident after it was
understood that it solved many physical puzzles.
But the story does not end at this point and further objections were
raised toward other principles of democracy that today we consider
obvious. As I mentioned before, at the time of the First World War,
the argument that universal voting right would produce chaotic politi-
cal effects was being developed. But this right was introduced in many
places and produced no chaotic effects. And so, an argument which
was strong before became weakened under the attack of experience.
A Seminal and Difficult Notion 115
Notes
6. Bell (1997) shows that the negative accent Weber puts on Gesinnungsethik has to
be related with G. Lukacs, who was present at the private discussions Weber
organized in his home and who frightened him by his fanaticism, as he frightened
Thoman Mann, since he appears as the Jesuit Naphta in Der Zaubeberg (Lukacs
was, it seems, proud of this portrait). The conceptual distinction transcends obvi-
ously these circumstances, however.
7. I have left aside here the consequentialist arguments in favor of democracy (as it
makes economic development easier). They have been developed again by Olson
(1993).
8. I follow here Duhem (1954), tome 1: 371-372: "Aucun corps inanimt ne peut Ctre
en mouvement s'il n'est soumis B l'action d'un moteur qui soit distinct d e lui et
extkrieur B lui; il faut que ce moteur, pendant toute la durte du mouvement, lui
soit constamment appliqut, soit sans cesse en contact avec hi."
9. J. Buridan, Questions sur la physique, develops the so-called "theory of impetus,"
according to Duhem a first formulation of the principle of inertia as we know it.
Question 12 of Book VIII in particular criticizes the principles of Aristotelian
physics using this example of the straw heap.
10. The ideas I have presented are developed in a more extensive fashion, notably in
(Boudon, 1997a).
Generalizing the ''Rational Choice Model"
into a Cognitivist Model1
can hold the irrational cause E as the cause of P, even though I cannot
show its existence directly. Scheler (1954 [1916]) has rightly stressed
the point that an action or a belief has always a meaning to the actor
and that assuming that this meaning is not the cause of the action is a
very strong and hence in many cases questionable assumption. A basic
point of Freudianism and Marxism is, on the contrary, that the mean-
ing to the actor of his actions or beliefs, in other words the reasons on
which he thinks they are grounded, would in many cases be mere
"rationalization" as Freudians would contend, or the product of "false
consciousness," as Marxians would say: these reasons would not be
the actual causes of the actions and beliefs. The lasting and diffuse
influence of Freudianism and Marxism explains to some extent at least
that social scientists introduce so easily the idea that the meaning to
the actor of his beliefs or actions is false and consequently should not
be treated as its cause.
3. When a social actor has the feeling that "X is good" or that "X is
true," he also has the feeling that others should endorse these state-
ments. Beliefs are not seen by social actors as preferences. I prefer
white to red wine, but I accept easily that my neighbor has the oppo-
site preference. By contrast, if I believe that "X is good" or that "X is
true," I will normally expect that my neighbor holds the same belief.
This normally "trans-subjective" dimension of beliefs is easily ac-
counted for in the framework of the cognitivist model: if beliefs are
endorsed once they are perceived by the social actor as grounded on
solid reasons, he should expect that these reasons also convince his
neighbor. This "trans-subjective" dimension of beliefs is much less
easily accounted for by irrational theories. It is, however, an essential
character of beliefs.
4. The "cognitive" model does not ignore socialization. A pupil
believes that "2 and 2 are 4" certainly because he was taught this, but
the belief "holds" in his mind because it is grounded on solid reasons.
Socialization is a channel which should not be confused with its con-
tent, and which can be effective only for some contents. As Durkheim
himself stresses, even a cat cannot be conditioned to believe that a ball
of string is a mouse. Socialization is in other words never the only
cause of beliefs: this corollary is immediately derived from the prin-
ciples of the cognitivist model.
As I noted above, the "cognitivist" model is accepted as far as Type
1 beliefs are concerned. It is easily recognized that I believe that "2
Generalizing the "Rational Choice Model" 125
plus 2 are 4" because there are strong reasons for believing that the
statement is valid. It is easily accepted that this belief is "trans-subjec-
tive," in other words that I would normally expect that the generalized
Other also believes that "2 plus 2 are 4." It is easily accepted that
socialization is not the only cause of the belief. But the claim of the
"cognitivist" model that the same type of analysis can also be con-
ducted in the case of beliefs of Types 2 and 3 is exposed to more
resistance.
I will now show with the help of some examples that the "cognitivist"
model can be applied successfully to the case of beliefs of Types 2 and
3. I will illustrate the application of this model to Type 2 beliefs by
examples drawn from cognitive psychology. I will explain further why
I chose this field rather than another. Then, I will illustrate Type 3
beliefs by examples drawn from the sociology of knowledge and the
sociology of norms.
theory, the "cognitivist" theory explains more of the observed data and
explains them with lighter and more acceptable assumptions than the
irrational phylogenetic theory of the naturalistic type originally pro-
posed by the authors of the study. The "cognitivist" explanation could
also inspire experimental designs which would make direct testing of
it possible. An easily devisable design could check, for instance, whether
the modal value is changed with a change in the parameters introduced
in the question or with a change in the wording of the question where
the ambiguity of the notion of "rate" would be eliminated, etc. In other
words, the "cognitive" theory has not only a greater explanatory power,
it has also a greater heuristic value.
Table 5.1
A Negative Correlation Perceived as Positive
- -
Disease
Symptom Ill Not ill Total
Symptom: yes 37 33 70
Symptom: no 17 13 30
Total 54 46 100
Table 5.2
A Causal Presumption Can Be Derived from One Piece of Information
Disease
Symptom Ill Not ill Total
Symptom: yes 4 16 20
Symptom: no 16 64 80
Total 20 80 100
The hostility toward the Church is due to the fact that this institu-
tion is seen, not only as the product, but as the gate-keeper of Tradi-
tion, and hence as one of the main obstacles against the necessary
political renovation of the country. The anticlericalist stance of the
Enlighteners derives again from a practical syllogism, the premises of
which are perceived as analytical or empirical obvious statements.
Anticlericalism becomes a collective belief because each ideal-typical
citizen in particular has good reasons to give it a positive meaning.
The collective belief results from the aggregation of individual beliefs
grounded on strong reasons. This explains why it spread so quickly,
says Tocqueville. If its diffusion would have resulted essentially from
a contagion effect, it would have spread much more slowly.
Tocqueville's analysis uses up to this point the "cognitivist" model
(people have strong reasons to believe what they believe, to do what
they do). But it contains also an additional part worth being mentioned
in the context of this discussion, where Tocqueville suggests that some
other people moved toward anticlericalism rather on the basis of utili-
tarian reasons in the style of the "Rational Choice Model" than on the
cognitive reasons evoked up to this point. Anticlericalism was also
endorsed by many people, he says, because they found it useful to
follow what they perceived to be the dominant mood. In other words:
the cost-benefit balance of conformism appeared to them as more
favorable than the balance of anti-conformism. In most cases, they
kept their Christian faith, since religious conversions rest rarely on
opportunistic decisions, but they understood that it was better for them
not to raise their voice against the anti-clerical propaganda. On the
whole, some had "cognitive" reasons to be anti-clerical, others "utili-
tarian" reasons not to oppose the anticlerical actions and declarations:
Ceux qui niaient le christianisme Clevant la voix et ceux qui croyaient encore
faisant silence, il arriva ce qui s'est vu si souvent depuis parmi nous, non seulement
en fait de religion, mais en tout autre matibre. Les hommes qui conservaient
l'ancienne foi craignirent d'Ctre les seuls B lui rester fidbles, et, redoutant plus
l'isolement que l'erreur, ils se joignirent B la foule sans penser comme elle. Ce qui
n'Ctait encore que le sentiment d'une partie de la nation parut ainsi I'opinion de
tous, et sembla dks lors irrksistible aux yeux-m~mesde ceux qui lui donnaient cette
fausse apparence (Tocqueville, 1986 [1856]: 1045).
138 The Origin of Values
more moral than doctrinal in the USA, explains plausibly why moral
conformity is greater in the USA than, say, in France. A surveyl1 has
shown for instance that Frenchmen tolerate more easily than Ameri-
cans that positions are filled on the basis of personal relationships
rather than on impersonal rules. Current observations show also that
the former are more tolerant about corruption and privileges. They
seem in other words to have a more tolerant attitude toward the viola-
tion of basic moral values. My guess (which would have to be con-
firmed by adequate research) would be that moral values are stronger
in the U.S. because they are more loosely connected to religious dog-
mas than in France. By contrast, moral values would suffer in France
from their connection with religious dogmas, which some people en-
dorse, while many others do not. Again, this is just a guess. I mention
it for methodological reasons: it suggests that even macroscopic dif-
fuse cultural differences such as the ones I am evoking here can be
explained as the aggregate result of reasons.
The example from Marx has shown that normative evaluative be-
liefs could, as "positive" beliefs, be analyzed as the effect of reasons.
This point is so important that I would like to reinforce it by two
additional examples that illustrate the fruitfulness of the "cognitivist"
model in the important field of norms and values.
C. W. Mills
French Doctors
I will discuss briefly a final example. A few years ago, two French
doctors were condemned in a trial dealing with the transfusion of
blood contaminated by the AIDS virus. A petition was signed by
numerous doctors and scientists throughout the world, including thirty
Nobel Prize winners, making the point that the trial had been unfair
and entreating the French president to use his prerogative of presiden-
tial pardon, which could effectively be applied in this case, to the
effect of extracting the doctors from jail. A strong majority disap-
proved the proposal. Already before, influential members of the gov-
ernment made clear they would not recommend the pardon. This is an
example of these current collective beliefs which the social sciences
should be able to explain. The example is taken because the collective
belief in this instance appears unusually strong. Can the "Rational
Choice Model" explain this social fact satisfactorily? Can the socio-
logical classical "internalization of norms" model explain it? Evidently
not, while the "cognitivist model" can: the petition rests on the argu-
ment that no one should be condemned until those equally guilty are
condemned. If this rule was applied, no one would ever be condemned.
As this consequence is unacceptable, so are the principles on which it
is grounded. The collective rejection on the petition was grounded
neither on the advantage to be derived from this rejection (which
advantage?), nor on an aversion against the condemned doctors, but
on the reason that it was contradictory to the very existence of a
system of justice.
I evoked briefly this last example because it suggests that the
"cognitivist model" easily explains the collective moral feelings that
are continuously expressed by people in routine political and social
life.
142 The Origin of Values
Notes
faster rather than slower in city traffic, and for this reason consider
traffic lights as a "good" (though unpleasant) thing. The value state-
ment "traffic lights are a good thing" is the conclusion of a strong
argument grounded on the empirical undebatable statement that traffic
is more fluid with traffic lights than without. Though simple, this
example is typical of many normative arguments. It shows that a nor-
mative argument can be as convincing as a descriptive one. Whether
A is a cause of B is a non-normative question. Now I will accept the
statement "A is a cause of B" if it is a consequence of a strong
argument. If, moreover, I prefer B to non B, A is demonstrated to be a
"good" thing for me. If all prefer B, A is demonstrated to be a good
thing for all. Thus, the gap between ought and is exists undoubtedly.
But when axiological statements can be interpreted consequentially,
that is, when "A is g o o d means "A leads to consequences considered
as good by people," their validity can be demonstrated as solidly as
the validity of descriptive statements. As "A leads to consequences
considered as good by people" can be translated into "A is an appro-
priate means to reach a state of affairs considered as good by people,"
the argument can be called instrumental as well as consequential.
However, as well seen by Weber, axiological statements cannot
always be considered as the conclusion of consequential arguments.
By creating his notion of "axiological rationality," Weber wanted pos-
sibly to insist on the point that people may have in some circum-
stances strong reasons for believing that "A is good, bad, legitimate,
illegitimate, fair, unfair, etc.," without these reasons belonging to the
instrumental-consequential category. My contention is that by so do-
ing he introduced a powerful idea, crucial to my understanding of
axiological feelings and of justice feelings in particular.
To summarize, I would define a feeling or a statement as
"axiologically rational" if people would consider it as grounded on
strong arguments which can be but are not necessarily of the conse-
quential-instrumental type. I propose, in other words, to define
axiological rationality as a form of cognitive rationality characterized
by the fact that it deals with arguments where at least one statement is
axiological, since ought cannot be derived from is, or, more precisely,
since an ought-statement cannot be exclusively derived from is-state-
ments.
While Weber was probably the first author to suggest conceptualiz-
ing the notion of "axiological rationality" as we understand it, he is
not the first to use it.
152 The Origin of Values
were conducted with the respondents have shown that they accept the
idea that income inequalities should reflect functional inequalities: I
am entitled to get a higher income if the tasks I am supposed to fulfill
given my occupational role are functionally more important. But given
the context of the experiment, they could not answer the question of
whether or not the income distributions reflected functional inequali-
ties: they missed the relevant information which would have been
necessary to make the question meaningful. They probably started
from questions they could answer: is it a legitimate goal for a govern-
ment to try to make the mean of an income distribution as high as
possible? They answered obviously yes to this question. Is it a good
thing for a government to try to make the standard deviation lower?
With the functional theory in mind, it is easily checked that trying to
make the deviation lower can be dysfunctional and have the effect of
making the mean lower. It is not implausible to accept the assumption
that the respondents endorse some rough version of the functional
theory. But, as they do not know whether the inequalities reflected by
the distributions are functional or not, the most prudent assumption
open to them is that they are and, consequently, to draw the conclu-
sion that an effort to reduce the standard deviation can be dangerous.
On the whole, given the cognitive conditions of the experiment, an
attractive answer is to reject any constraint as far as the standard
deviation of the distribution is concerned. On the other hand, it is a
good thing to introduce a constraint on the floor, since it is normally
expected of government that it try to install some protection against
the hazards of life to which citizens are exposed. Hence, the fact that a
constraint on the floor is widely accepted by the respondents. On the
whole, the solution most widely chosen is: maximize the mean, at-
tempt to provide a minimum income, and do not attempt to minimize
standard deviation since it is produced by multiple unknown mecha-
nisms and since it can be counterproductive to lower it.
In other words, the high structuration of the respondents' answers
can be analyzed as the effect of strong reasons. It can be noted, inci-
dentally, that the same strong reasons explain presumably that actual
governments make, more or less generally though implicitly, the same
choice as the respondents. They usually consider it good to take the
decisions likely to increase the mean of the income distribution, in
other words, to enhance growth. On the other hand, most governments
consider that people should be provided insurance against the hazards
The Cognitivist Model Applied to Justice 159
of life: even the less "social" governments care about poverty and
about making the proportion of the poor as low as possible. All are
concerned, moreover-though, of course, to various degrees-that no
one, ideally at least, be "excluded" from society. Finally, all recognize
that inequalities are the product of complex micro-phenomena, that
they are partly functional, that it is impossible to determine what the
ideal standard deviation of income distribution should be, etc. They
also consider that a lower standard deviation is not necessarily better.
For these reasons, most governments are prudent in this respect. Their
efforts to reduce the standard deviation of incomes are, in many cases,
marginal when they are not merely symbolic.
Governments, as people, have most likely a more or less diffuse
minimal philosophy about income distribution. This philosophy in-
cludes a set of arguments which they consider as strong. From these
arguments, they draw conclusions as to whether it is good or not to try
to make the standard deviation as low as possible, the mean as high as
possible, etc.
Reciprocally, the strong rejection of Rawls's difference principle by
governments as well as by the samples in the considered experiment is
due to the fact that there are strong reasons to reject it: notably, it is
impossible to determine effectively the threshold under which a reduc-
tion of the standard deviation of income distribution would have a
negative effect on its mean.
It should be added that the system of reasons at work in the answers
of the respondents would have plausibly been different if the experi-
mental conditions had been different. Suppose, for instance, that the
distributions proposed to the attention of the respondents would have
been presented as reflecting, not fictitious, unidentified "societies,"
but the distributions of salaries in organizations, that the organizations
would have been described as similar to one another, and that infor-
mation would have been given to the respondents on the types of
occupations and functions corresponding to the income classes. In that
case, the respondents would probably have considered the standard
deviation and tried to see whether it reflected functional inequalities.
While it is impossible to determine what the good standard deviation
would be as far as the general society is concerned, it would not be
impossible in the case of an organization. At any rate, with the
cognitivist model in mind, it would be possible to build a design of
observation where the contextual fictitious conditions and the set of
160 The Origirl of Values
workers employed in the same type of job. Even though they are un-
likely to meet them, they can be informed on the subject through unions
and other media and use a "formalist reference group theory" approach.
The two types of comparison are possible, likely, and legitimate. Both
are equally valid conjectures.
So the choice of comparison is open.
In the cases proposed to the arbitrators, the retributions to equal qualifi-
cations or types of jobs are presented as unequal.
This inequality can be interpreted in different fashions. The inequality
can be legitimate, as this would be the case for instance if it were due to
the fact that in a company where the salaries are higher, the workers,
other things being equal, would on the average have been employed a
longer time by the firm. In this case, it should not be proposed to
reduce the inequality.
So it can be unfair to apply the principle of absolute equity rather than
the relative equity principle.
Absolute equity can be preferred to relative equity under two condi-
tions: if information is available on the causes of inequality between the
firms and if it turns out that it is illegitimate.
If not, relative equity should be preferred to absolute equity.
With imperfect information on the causes of inequality, applying the
principle of absolute equity can be unfair and, moreover, can possibly
generate undesirable effects.
Absolute and relative equity are two versions of the same theory of
justice. The choice between them should be guided by the available
information on the causes of inequalities.
they may object that the views developed by the actors themselves
should be taken into consideration since, though they can be self-
interested, they are also almost certainly inspired by a familiarity with
the field which outside observers cannot equal. Nothing prevents the
men from management and unions having in mind the above argu-
ments, even if their views are biased by the consideration of their own
interests. Being confident to those "who know" is a possible strategy
in a situation where one has the impression that much relevant infor-
mation is missing. Besides being simple, this third type of solution has
the advantage of being able to eliminate the biases introduced by the
interests of the two parties by proposing a solution located halfway
between the proposals of management and the unions. This solution
entails several types of costs, however. It is mechanical and lazy. It
contradicts the rules defining the role of the arbitrator. Moreover, in
spite of the fact that it appears as prudent, it supposes much: that the
proposals of the two parties are likely biased to the same extent for
instance. Briefly, there are strong reasons to choose this third solution,
but also strong reasons to prefer the two other main solutions.
The cognitivist interpretation I have just proposed explains
Bazerman's main findings: the existence of three groups of decisions,
the unequal size of the three groups, the data about the proportion of
variance explained by the independent variables, and notably the strong
difference in the weight of the independent variables. The group larg-
est in number starts from the assumptions: that they can decide by
themselves, that there is a best decision given the constraints, that the
information is leaky, and that the most prudent hypotheses are wel-
come in such a situation. Thus, the cognitivist approach explains the
main findings in a satisfactory fashion. Moreover, it refrains from
using the very strong assumption according to which the arbitrators
would simply express a brutal preference for one of three popular
justice principles, namely, equity, absolute or relative, and equality.
Even if such an assumption would appear acceptable, it would not
explain the existence of the three groups, nor a fortiori their unequal
size. In the cognitivist analysis, by contrast, the arbitrators are pre-
sented, not as selecting a principle under the effect of unknown causes,
but as trying to find out the best solution, given notably the constraints
in matters of information they have to face. The analysis suggests that
they all have the same conception as to what justice means in this
case, namely, making the retribution as close as possible to the contri-
166 The Origin of Values
Figure 6.1
The Three Types of Arguments behind the Decisions
of Bazerman's Three Groups
Type 3
v
Can it be applied ? No Management and b
unions as experts
+
Yes
I
Workers concretely
,
-
Type 2
comparable
Reference group
acceptable and
available
Workers formally
comparable
would confirm another conclusion that can be drawn from this analy-
sis: that it is impossible to reconstruct the feelings of justice experi-
enced and expressed by people as resulting simply from the applica-
tion of general principles (for other examples, see Boudon, 1995b).
According to the cognitivist theory I propose here, these principles are
only elements among others, though important ones, in the evaluation
procedure more or less implicitly conducted by social actors. First of
all, once they are confronted by a given situation, they will have to
decide whether they should apply one principle or another.
This example entails also several general conclusions. It shows first
that rationality can be, in Weber's terms, axiological rather than in-
strumental. The arbitrators care of course about their reputation and
want to reach a decision likely to be perceived as good. But they do
not try to guess the opinion of their audience. Rather they try to reach
a good decision, namely, a decision which themselves and others would
consider good because it is grounded on strong reasons. Secondly, the
example shows that the conclusions are not reached on the basis of
subjective idiosyncratic reasons. Otherwise, the statistical distribution
of the decisions would appear as less strongly structured.
of the earth, another drawing it toward the point in the ceiling where
the thread holding the pendulum is fixed, etc. We accept Huygens'
theory because we consider the concept of "force" and the mental
construct described by the "parallelogram of forces" as legitimate. It
suffices to think of the objections raised against such totally unempirical
notions, from Descartes to Carnap (see Boudon 1997c), to check im-
mediately that there are no criteria by which we could easily deter-
mine that such concepts are acceptable, legitimate or not. Since we
have no reason to consider the case of Huygens' theory as exceptional
rather than typical, it follows that we normally cannot define criteria
by which a scientific theory can be held as valid or true. Descartes was
convinced that the notion of "force" is unacceptable. We are con-
vinced that it is acceptable. There is no easy criterion thanks to which
we would judge it acceptable, though, namely criteria which we would
have identified and that Descartes would have failed to identify. We
accept it because it explains many things and has repeatedly been
successfully used in the history of physics. So, when Kant states that
there are no general criteria of truth, he is correct.
Why then would there be general criteria associated with such no-
tions as "fair," "good," "legitimate," etc., and their contraries, while
no general criteria can be associated with the attributes "true" and
"false"? Of course there is a strong demand for such criteria. But, as
the case of Popper's "falsification" criterion shows, the criteria that
have been proposed to meet this demand are only ~ a r t i a lThey
. ~ can in
particular circumstances be necessary. They are never sufficient.
Does this lead to an "intuitive" theory of truth, fairness, legitimacy,
or goodness? Evidently not. We hold the notion of "force" as accept-
able in a scientific physical theory and the notion of " G o d as unac-
ceptable, not for intuitive reasons, but because a historical Darwinian
process has selected the first one as adequate, given the aims pursued
by scientists, and rejected the second one as inadequate.
My claim here is that the situation is basically the same, whether
we have to explain the collective feeling that "X is true" or the collec-
tive feeling that "X is fair, legitimate, etc." Such statements are en-
dorsed when they are grounded in the minds of people on reasons
perceived by them as strong, or as stronger than the reasons leading to
the opposite conclusions. To this, it must be added that the empirical
statements included in the system of reasons should satisfy the Tarskian
criterion, that is, be congruent with the relevant observations, but that
in most cases the system of reasons will also include non-empirical
The Cognitivist Model Applied to Justice 175
statements the validity of which, in the general case, will not be reduc-
ible to the verdict of some well-defined criterion.
And so, "cognitive rationality" and "axiological rationality" point
to the fact that the feelings that "X is true" in the first case, "X is fair,
legitimate, etc.," in the second one, are grounded in the minds of
social subjects on systems of reasons perceived by them as strong;
they differ by the fact that the "systems of reasons" include at least
one prescriptive statement in the latter case.
A hardware store has been selling snow shovels for 30 Swiss francs
(or 30 German marks). The morning after a heavy snow storm, the
store raises the price to 40 Swiss FrancstGerman marks. How do
you evaluate the price rise?
Notes
1. Published with Emmanuelle Betton as "Explaining the Feelings of Justice," Ethi-
cal Theory and Moral Practice. An International Forum, 1999, 2: 365-398.
2. Many exceptions can, of course, be evoked against this statement, as the work of
the Swiss Sociologist J. Kellerhals, among other examples, shows. See for in-
stance (Kellerhals et al., 1995).
3. M. Sukale writes (1995: 43): "Damit ist Webers Einteilung des rationalen Handelns
in zweckrationales und wertrationales, als gabe es zwei verschiedene Arten
rationalen Handelns, ausserst irrefuhrend" ("WeberVs distinction between
axiological and instrumental rationality, as though there would be two types of
rational action, is extremely misleading.")
4. On this point, for instance, see (Tyler, Lind, 1992).
180 The Origin of Values
5. R. Boudon has discussed these points more thoroughly in (Boudon, 1994b) and
(Boudon, 1997~).
6. For example, Hochschild (1981), where the fact that people endorse different
norms of justice depending on the characteristics of the contexts they are located
in is considered a sign of the inconsistency of their standpoints.
The Cognitivist Model Applied to the
Analysis of Public Opinion1
Adam Smith is the author of two famous metaphors. One, the meta-
phor of the "invisible hand," has been widely assimilated, whereas the
importance of the other, the "impartial spectator," has not received
nearly as much recognition. It is not my intention to launch any further
debate on what is known as das Adam Smith Problem since Knies
(1883), but I do believe that these two metaphors sum up one and the
same intuition and conceptual framework, even if this point was not
clearly developed by Adam Smith himself.
The "invisible hand7' reconciles specific interests with the general
interest. It is not from the benevolence of the butcher that we expect
our dinner. When he lowers his prices or buys better quality meat to
attract customers the butcher does so in pursuit of his own egoistic
interests, and also helps to generate an optimal structure of supply for
the benefit of the consumer.
Another metaphor, that of the "impartial spectator," is generally not
as well understood. Its fate is much the same as that of Max Weber's
"axiological rationality." In my opinion it is a highly important notion,
182 The Origin of Values
Significant research has drawn attention to the fact that individual judg-
ments and opinions are affected by all sorts of biases. The idea of a "bias,"
although extremely widespread, is really not very felicitous. All it does, as
in the work of Kahneman and Tversky, for instance, is put a label on a
problem, the problem being the origin of those biases. Other contributions,
both contemporary and somewhat older, not only evidence some biases,
but provide a set of theoretical formulations which facilitate the compre-
hension of how opinions are formed, and above all, the fact that opinions
may well be biased. These theoretical formulations represent a major con-
tribution to contemporary work in the social sciences, in my opinion, but
they have never been given the attention they deserve. Moreover, since
they were developed independently of each other, there have rarely been
attempts to make a connection between them, and even less often to view
them as what they are: complementary pieces contributing to a theory of
opinions. One of my intentions here is to attempt just that synthesis. These
formulations are important, especially because they show that the reasons
that account for opinions oscillate between the two poles of particularism
and universalism, depending on the parameters defining the person and his
184 The Origin of Values
Table 7.1
Opinions on the Thirty-five Hour Work Week
For each of the following category of individuals, tell me whether, in your opin-
ion, the shorter work week is very positive, somewhat positive, somewhat nega-
tive or very negative:
Neither
Sub-total Sub-total positive nor No
positive negative negative opinion Total
The unemployed 72 19 5 4 100
Salaried workers,
public sector
Precarious workers
Salaried workers,
private sector
Executives
Company heads
Source: BVA opinion poll, May 11,1998.
Table 7.2
Confidence in the Special Court of Justice of the Republic
Are you confident that the court of justice of the republic, composed of elected
officials and senior magistrates, will judge the three former ministers, L. Fabius,
G . Dufoix and E. HervC equitably? A: confident or relatively confident; B: not
confident or relatively not confident.
Sympathizers
Total PC PS "Green" UDF-DL RPR FN None
A 38 45 53 37 34 37 12 30
B 57 55 44 63 63 62 88 58
No opinion 5 - 3 - 3 1 - 12
Source: BVA poll, February 18,1999. (PC=Communist Party; PS=Socialist; UDF-
DL=Center Parties; RPR=Gaullist Party; FN=Rightist Party)
Table 7.3
Opinions on the Criminal Responsibility of Ministers
In your opinion, is the fact that a minister may be brought to court for decisions
made in the course of his or her duties (A) a relatively good thing, since political
leaders are not above the law, (B) a relatively bad thing, since it may make being
a minister an almost impossible job?
-- - - -- - - - -
Sympathizers
Total PC PS "Green" UDF-DL RPR FN None
A 85 78 81 91 90 88 85 84
B 11 22 15 9 7 10 12 9
No opinion 4 - 4 - 3 2 3 7
Source: BVA poll, February 18,1999. (PC=Communist Party; PS=Socialist; UDF-
DL=Center Parties; RPR=Gaullist Party; FN=Rightist Party)
Table 7.4
Forced Retirement
When jobs are scarce, people should be forced to retire early: % "agree"
with the frequency with which the simplistic answer is rejected: this
suggests that training should be responsible for an increased aware-
ness to systemic effects. With more training, people are more likely to
see that lowering the age of retirement cannot have a mechanical
effect on reducing unemployment.
If the "job cake" is finite, another way to reduce unemployment,
one may think, would be to cut down the number of job candidates by
excluding immigrants. Here we have a combination of two oversim-
plified prerequisites: the finite nature of the job cake and the infinite
interchangeability of job candidates (see table 7.5).
Countries rank grossly similarly on these two questions. At the
same time, comparison of the means shown in the two tables indicates
an "axiological" effect. It is easier to refuse to let immigrants into
one's country than to change the age of retirement. Indeed, the latter
case involves a breach of contract, entailing social, economic, and
ethical costs, under normal circumstances, whereas the former simply
implies refusal of any contractual relationship. Now, axiologically
speaking, the latter situation, that of a breach of contract, is less ac-
ceptable than the other (Kahneman et al., 1986b). It is this type of
effect that I wish to discuss here. In passing, I would simply point out
the value of a "judicatory" conception of opinion.
Table 7.5
Jobs Preference to Own Nationality
When jobs are scarce, employers should give priority to their own nationality
over immigrants: % "agree"
the long term. Now in fact, lay-offs here and now conserve jobs here,
elsewhere, and tomorrow. But it is also true that the chances of having
this lofty viewpoint adopted by the people who are being laid off or
those who fear for their jobs are extremely slim, and their discontent
makes them easy prey for demagogic politicians.12
These examples also clearly show that depending on the issue, the
saying Vox populi, vox Dei may or may not apply. The fact that the
French do not care about having taxes lowered is due to the fact that
the majority has no interest in caring about that. In this case, the
viewpoint of the "partial actor" rather than that of the "impartial spec-
tator" prevails because of both the nature of the issue and the charac-
teristics of French fiscal politics. On the basis of opinion surveys,
politicians may then be tempted to treat lowering taxes as a subsidiary
objective. And the leader of the National Assembly will be able to go
on tranquilly spending 600,000 francs a year of the taxpayer's money
on flowers.'3
Table 7.6
Confidence in Institutions
- -
Table 7.7
Proud to b e . . .
% German
"very proud" (West) French Italian English American
"those in the know," who may very well not know at all. As Pareto
tells us, the history of scientific thinking is the history of all of the
errors that humankind swallowed on the faith of specialists given the
name of scientists. The sally takes on an air of truth when applied to
all those "specialists" on whom our contemporary world relies, and
who are anything but scientists. We need only take all those "special-
ists," referred to above, who-with the blessing of political leaders, of
course-have imposed methods for teaching reading, writing, languages,
and mathematics, and not only teaching methods, but educational poli-
cies as well, the consequences of which seem to have been mostly
catastrophic.
isolated from two other notions, less famous, unquestionably, but just
as crucial to Weber's thinking: "axiological rationality" and the "dif-
fuse rationalization" of moral life. According to Weber, conflicts be-
tween values are the vector of progress-and of rationalization-in moral
life, which is the outcome of "axiological rationality," a concept not
very different from that of the "impartial spectator." A subject is guided
by "axiological rationality" inasmuch as his behavior may lay claim to
the approval of the "impartial spectator."
The case of striking transportation workers, alluded to above, may
provide concrete illustration of this point. The right to strike is in
conflict with the public interest here, and as a rule the public is not
responsible for the situation that triggered the grievances and led to
the strike. We have a "polytheism of values" here, with a "war of the
gods" on two basic principles (the right to strike and the right of the
public not to be "taken hostage"). Now the public is less and less
willing to accept this sort of use of the right to strike, which is increas-
ingly perceived as a "French exception." This trend is evidenced by
the fact that the media increasingly tend to question the legitimacy of
these practices, while the trade unions are developing and implement-
ing innovative approaches aimed at limiting the inconveniences con-
nected with the strike while maintaining the effectiveness of the pro-
test movement. The way the public reacted to the long museum work-
ers' strike in France in June 1999 is indicative of the same trend. The
strike was generally disapproved, on the grounds that tourists were
unfairly taken hostage, and that other ways of protesting could surely
be found. In all of these examples, as in those mentioned above (the
development of international law, the limitation of state sovereignty,
and so on), "rationalization of moral life" occurs in a context of "poly-
theism of values." Understandably, this rationalization takes place in a
context fraught with conflict. Trade unionists are easily led to use their
power to cause inconvenience, when the right to strike makes it legiti-
mate to do so. But the "impartial spectator" does not accept innocent
victims being heavily taxed by that right, and wants to see more ac-
ceptable "solutions" explored. The evolution of the form taken by
protest movements at the French electricity company yields another
illustration of the "rationalization of moral life" prompted by the "im-
partial spectator": the unions have abandoned power cuts, which para-
lyze the economy and inconvenience individual consumers consider-
ably.
The Cognitivist Model Applied to Public Opinion 209
The key role played by the "impartial spectator" has been stressed
by all of the classical sociologists, as well as by Adam Smith (but the
metaphor itself belongs to Smith exclusively). They apply it to the
moral and political evolution of societies, and, correlatively, to how
collective opinions and beliefs relative to values are generated.
Tocqueville emphasizes the "tyranny of opinion," but his very choice
of the word "tyranny" indicates that the impartial spectator may keep a
critical eye on any collective opinion, however prevalent. In the same
vein, Durkheim has shown that a collective belief cannot be sustained
if people do not view it as grounded: "Concepts, originally taken to be
true because they were collectively accepted, tend to be collectively
accepted only on the condition that they are viewed as true. We de-
mand their certification before giving them credence." Max Weber
stresses the fact that moral life is pervaded by a process of "diffuse
rationalization," over and beyond the "war of the gods" and the sound
and fury of "historical forces," meaning by this that moral ideas are
subjected to a selection process which retains those ideas preferred by
the "impartial spectator" (Boudon 1997a). Max Scheler has shown
that although "pharisaism" dominates social life, its very existence is
revelatory of the fact that values prevail over interests, and that they
The Cognitivist Model Applied to Public Opinion 211
and can only survive under cover, or at least illegally. Weber, too,
defended the idea of a Durchrationalisierung of moral life, meaning
that notwithstanding the conflicting action of "historical forces," some
irrevocable changes in our moral sensitivity do occur. Like Tocqueville,
he suggests that these irreversible changes are due to the "impartial
spectator." In the same vein, Simmel has stressed the fact that when
punishment for murder ceased to be indexed on the social importance
of the victim, that change turned out to be irrevocable, for it is the
nature of the act and the author's intentions in committing it that
should determine the harshness of the sentence.
The micro-mechanisms that the most fertile schools of modem so-
ciology discern, by various means, underlying data on opinions, also
foreshadow a theory of opinions still to come. While avoiding any
eclecticism, such a theory would at last move beyond that intellectu-
ally indolent model, full of preconceptions, patterned after kinetics,
according to which correlations between opinions and socio-biographic
data are necessarily explained by an occult "force," as well as the
economic model, launched concurrently by Marxists, Nietzscheans,
and proponents of the Rational Choice Model, according to which the
opinions of social subjects are exclusively dictated by their individual
andlor collective interests.
Perhaps, too, the theory of opinions whose outlines will be per-
ceived through the identification of these mechanisms will put an end
to the confused polemics on opinion polling. Some polls, those that
place the respondent in the "impartial spectator" position, yield indica-
tions that should be taken into consideration by political leaders. To
return to one of the examples discussed above, the fact that public
opinion questions the legitimacy of the special court is not indifferent.
Other polls, those that tend to place the respondent in the position of
the "partial actor," may be-and in reality inevitably are-
instrumentalized by politicians, but at their own risk.
The main conclusion to be drawn from these pages is, no doubt,
that the answers of opinion-givers depend on the nature of the ques-
tions put in any real or mock consultation. They may tend toward the
"partial actor" stance, as when the French are asked whether they care
about tax cuts, or be more of the "impartial spectator" type, as when
people are asked whether they sympathize more with Antigone or with
Creon, or when they are consulted on the contaminated blood trial.
Be this as it may, one cannot, in my opinion, develop a theory of
The Cognitivist Model Applied to Public Opinion 213
Notes
1. This paper was written for the Conference on "The explanation of social norms. A
comparison of sociological, economic and philosophic viewpoints based on the
shared concepts of rationality and cognition" ("L'explication des normes sociales.
Une comparaison des points de vue sociologique, Cconomique et philosophique B
partir de leurs concepts communs: rationalitt et cognition"), organized by P.
Demeulenaere and R. Viale at the Sorbonne on October 14-16, 1999. Translation
into English by Helen Arnold.
2. Weber's "axiological rationality" is generally interpreted as indicating the simple
coherency of the subject's acts with the values endorsed by him for no reason.
This interpretation is incompatible with the basic postulate of "comprehensive"
sociology, according to which an actor's actions and beliefs may be "understood"
by the sociologist because they are meaningful to the actor: in other words,
because the latter has reasons to partake of them.
3. Rawls's "veil of ignorance" is another version of Smith's "impartial spectator" as
well.
4. Thereis an enormous amount of literature on "networks" that explores this theme,
but networks are only one peculiar form of "community," in the broader sense of
the term as employed by Hardin and used here as well.
5. This and the following section contain a very brief discussion of the cognitive
mechanisms analyzed in Boudon (1986, 1994b, 1999d).
6. It is against the will of a great many French politicians, and thanks to the Euro-
pean Union, that the Colbertian model of the state is gradually fading out in
contemporary France. The policy of privatization presently pursued by the gov-
ernment is mostly the outcome of a combination of international pressure and of
the requirements for building the new Europe.
7. Douglas and Ney (1998) contend that their analysis founds a theory of cultural
biases because it rests on a typology. However, their typology only accounts for
some rather specific biases and does not uncover the mechanisms underlying
biases in general. A typology is a descriptive tool rather than an explanatory one
and for that reason does not in itself constitute a theory.
8. "I have pointed out how in times of enlightenment and democracy the human
spirit is loath to accept dogmatic beliefs" (Tocqueville, 1986 [1845]: 445), which
beliefs are defined by Tocqueville as "those that men receive trustfully, without
discussing them" (ibid).
9. Legrand is apparently referring to the school of sociology led by P. Bourdieu.
10. These hypotheses are also found in the "cognitive dissonance" model, according
to which the subject attempts, in a minimalist way, to modify his belief system so
as to eliminate contradictions. Insufficient attention has been paid to the fact that
this model implies that ordinary thought processes follow the same rules as scien-
tific thinking.
214 The Origin of Values
11. In the slang of the Ecole normale supe'rieure, a college for the intellectual elite,
the term "talas" designates people who "vonT-A-LA messe" (who go to mass).
See Peyrefitte (1998).
12. A perfectly classical version of this script was played out, for example, from mid-
September 1999 on, when the Michelin corporation announced its lay-offs.
13. Information given by Canal+ TV, October 13, 1999.
14. The habit of polling has reinforced the idea that individual opinions are random,
ungrounded, or again, socially determined, since people are of course only ques-
tioned about issues on which opinions are divided.
15. This interpretation has been legitimized both by conservatives such as Leo Strauss
and Carl Schmitt and by postmodernists, all too happy to turn Weber into a post-
Nietzschean. One need only refer to Coser (1984), however, to realize that there is
nothing new about the rejection of an interpretation of this sort ("Many commen-
tators have contended, for example, that Strauss completely misunderstood Max
Weber ; but this seems never to have disturbed him or his followers" ). This is an
excellent example of the application of "Hamilton's law" (1996), according to
which some hackneyed errors may be endlessly repeated in both scholarly and
non-specialized circles. In France, R. Aron's preface to a popular edition of Max
Weber's lectures, Le savant et le politique, helped to keep Strauss' interpretation
in circulation.
16. There is also Rousseau (the "general will").
17. In itself, this remark suffices to disqualify the allegations of all those neo-Marx-
ists, neo-Nietzscheans, and "Rational Choice Model" neo-Benthamians who pro-
claim the universality of their explanations, of the utilitarian value type.
References
Abel, T. 1964. "The Operation Called Verstehen." In H. Albert, ed., Theorie und
Realitat. Ausgewahlte Aufsatze zur Wissenschaftslehre der Sozialwissenschaften,
pp. 177-188. Tiibingen: Mohr.
Albert, H. 1990. "Der Mythos des Rahmens am Pranger." Zeitschriftfurphilosophische
Forschung Band 44, 1: 85-97.
Allais, M. 1953. "Le comportement de l'homme rationnel devant le risque: critique
des postulats de 1'Ccole amCricaine." Econometrics 21,4: 5 0 3 4 6 .
Axelrod, R. 1984. The Evolution of Cooperation. New York: Basic Books.
Bazerman, M. 1985. "Norms of Distributive Justice in Interest Arbitration." Industrial
and Labor Relations Review 38,4, July: 558-70.
Bell, D. 1997. Lafin de l'ide'ologie. Paris: PUF.
Bergeron, H. 1999. Soigner ou prendre soin des toxicomanes: anatomie d'une croyance
collective. Paris: PUF.
Bloom, A. 1987. The Closing of the American Mind. Simon & Schuster: New York.
Bloor, D. 1980, Knowledge and Social Imagery. London: Routledge & Kegan Paul.
Bodin, J. 1986 [1576]. Les six livres de la Ripublique. Paris: Fayard.
Boudon, R. 1974. Education, Opportunity and Social Inequality. New York: Wiley.
Boudon, R. 1981. Logic of social action. LondodBoston: Routledge & Kegan Paul.
Boudon, R. 1982. The Unintended Consequences of Social Action. London: Macrnillan.
Boudon, R. 1986. L'idkologie ou l'origine des ide'es repes. Paris: FayardJSeuil, Col-
lection "Points." Eng. trans.: The Analysis of ideology. London: Polity Press.
1989.
Boudon, R. 1989. Subjective Rationality and the Explanation of Social Behavior.
Rationality and Society 1 , 2, oct.: 171-196.
Boudon, R. 1992. Sentiments of Justice and Social Inequalities. Social Justice Re-
search 5, 2, June: 113-135.
Boudon, R. 1993. "European Sociology: The Identity Lost?" In B. Nedelmann and P.
Sztompka, eds., Sociology in Europe. In search of identity, pp. 2 7 4 4 . New YorW
Berlin: De Gmyter.
Boudon, R. 1994a. "La logique des sentiments moraux." L'Annke sociologique 44:
19-51.
Boudon, R. 1994b. The Art of Self-Persuasion. London: Polity Press.
Boudon, R. 1995a. "A propos des sentiments de justice: nouvelles remarques sur la
thCorie de Rawls." L'Anne'e sociologique 45,2: 273-296.
Boudon, R. 1995b. Le juste et le vrai; essais sur l'objectivite' des valeurs et de la
connaissai~ce.Paris: Fayard.
Boudon, R. 1997a. "La rationalit6 axiologique." In S. Mesure, ed., La rationalite' des
valeurs, pp. 15-57. Paris: PUF, and in R. Boudon, Le sens des valeurs, pp. 137-
203. Paris: PUF, "Quadrige," 1999.
216 The Origin of Values
Kahneman, D., Knetsch, J., Thaler, R. 1986a. "Fairness and the Assumption of Eco-
nomics." Journal of Business 59: 285-300.
Kahneman, D., Knetsch, J. L., Thaler, R. 1986b. "Fairness as a Constraint on Profit
Seeking." The American Economical Review 76,4: 728-41.
Kant, I. 1785. Grundlegung zur Metaphysik der Sitten, in Kant's gesammelte Schriften.
Hrsg von der Koniglich Preussischen Akademie der Wissenschaften, Berlin: G.
Reimer, W. De Gmyter, 1910, Tome IV, 1911.
Kant, I. 1787. Kritik der reinen Vernunf: Bibliographisches Institut, LeipzigJVienna:
Meyers Volksbucher, 2nd ed., no date.
Kant, I. 1797. "Uber ein vermeintes Recht aus Menschenliebe zu lugen." Berlinische
Blatter 1: 301-314.
Kellerhals, J., Modak, M., Sardi, M. 1995. "Justice, sens de la responsabilitk et rela-
tions sociales." L'Annie sociologique 45, 2: 317-349.
Knies, K. G. A. 1883. Die politische Oekonomie vom geschichtlichen Standpunkte.
Neue, durch abgesonderte Zusatze verm. Aufl. der "politischen Oekonomie vom
Standpunkte der geschichtlichen Methode," Braunschweig: Schwetschke & Sohn,
1853.
Kuran, T. 1995. Private Truths, Public Lies. The Social Consequences of Preference
Falsification. Cambridge, MA: Harvard University Press.
Landes, D. S. 1998. The Wealth and Poverty of Nations: Why Some Are So Rich and
Some So Poor. London : Little Brown.
Legrand, L. 1982. Pour un coll2ge dimocratique. Rapport au ministre de 1'Cducation
nationale. La Documentation fran~aise.
Lukes, S. 1967. "Some Problems about Rationality." Archives europiennes de sociologie
8,2: 247-264.
MacIntyre, A. 1981. After Virtue. London: Dockworth.
Marx, K. 1968. "MatCriaux pour 1' 'tconomie."' In Oeuvres: Economie. T. 11, Paris:
Gallimard.
Mead, G. H. 1934. Mind, Selfand Society. From the Standpoint of a Social Behavior-
ist. Chicago: The University of Chicago Press.
Mills, C. W. 1951. White Collars. The American Middle Classes. New York: Oxford
University Press.
Mitchell, G., Tetlock, P. E., Mellers, B. A,, Ordonez, L. D. 1993. "Judgements of
Social Justice: Compromise Between Equality and Efficiency." Journal of Per-
sonality and Social Psychology 65,4: 629-639.
Mommsen, W. 1965. "Max Weber's Political Sociology and His Philosophy of World
History." International Social Science Journal 17, 1: 23-45.
Mommsen, W.J. 1959. Max Weber und die deutsche Politik, 1890-1920. Tubingen: J.
C. B. Mohr.
Nietzsche, F. 1887. Zur Genealogie der Moral. T.3. In Werke. Hrsg. von Karl Schlechta,
Frankfurt am MaintWien: Ullstein Verl., 1972.
Nisbet, R. A. 1966. The Sociological Tradition. Glencoe, IL: The Free Press.
Nisbett, R., Ross, L. 1980. Human Inference. Englewood Cliffs, NJ: Prentice Hall.
Nozick, R. 1997. "Why Do Intellectuals Oppose Capitalism?" In Socratic Puules,
Cambridge, MA: Harvard University Press.
Oberschall, A. 1994. "R&gles, normes, morale: Cmergence et sanction." L'Annie
sociologique 44, "Argumentation et Sciences Sociales": 357-384.
Olson, M. 1993. "Dictatorship, Democracy and Development." American Political
Science Review 87, 3, September: 567-576.
Opp, K. D. 1983. Die Entstehung sozialer Normen. Tubingen: J. C. B. Mohr.
Overbye, E. 1995. "Making a Case for the Rational, Self-regarding, 'Ethical' Voter-
References 219
and Solving the 'Paradox of Not Voting' in the Process." European Journal of
Political Research 27: 369-96.
Pareto, V. 1935. The Mind and Society. New York: Harcourt Brace.
Pascal, B. 1977 [1670]. Pense'es. Paris: Gallimard.
Peyrefitte, A. 1998. Rue d'Ulm: chroniques de la vie norrnalienne. Paris: Fayard. 5th
ed .
Piaget, J. 1985. Le jugement moral chez l'enfant. Paris: P.U.F. 6th ed.
Popkin, R. H. 1964. The History of Scepticism from Erasmus to Descartes. Assen:
Van Gorcum.
Popper, K. 1967. "La rationalitb et le statut du principe de rationalitb." In E. M.
Claasen, ed., Les fondements philosophiques des syst2mes e'conomiques, pp. 142-
150. Paris: Payot.
Popper, K. 1976. "The Myth of the Framework." In E. Freeman, ed., The Abdication
of Philosophy: Philosophy and the Public Good, pp. 2 3 4 8 . La Salle, IL: Open
Court.
Popper, K. R. 1968. The Logic of Scientific Discovery. London: Hutchinson.
Radnitzky, G. 1987. "La perspective bconomique sur le progrbs scientifique: applica-
tion en philosophie de la science de lWanalysecoiit-btnbfice." Archives de
philosophie 50, April-June: 177-198.
Rawls, J. 1971. A Theory of Justice. Cambridge, MA: The Belknap Press of Harvard
University Press.
Rawls, J. 1993. Political Liberalism. New York: Columbia University Press.
Revel, J. F. 1988. La connaissance inutile. Paris: Grasset.
Root, H. L. 1994. The Fountain of Privilege: Political Foundations of Economic
Markets in Old Regime France and England. Berkeley: University of California
Press.
Rorty, R. 1989. Contingency, Irony and Solidarity. CambridgelNew York: Cambridge
University Press.
Ruse, M. 1993. "Une dCfense de l'bthique Cvolutionniste." In J. P. Changeux, ed.,
Fondements naturels de l'e'thique, pp. 35-64. Paris: Odile Jacob.
Ryle, G. 1966 [1949]. The Concept of Mind. London: Hutchinson.
Scheler, M. 1954 [1916]. Der Fornzalismus in der Ethik und die materiale Wertethik.
[Jahrbuchfiir Philosophie und phiinomenologische Forschung, January 1913 and
July 19161. 1st Germ. ed. 1916, reprinted in Gesammelte Werke. V. 2, Bern1
Munchen: Francke.
Scheler, M. 1978 [1912]. Das Ressentiment irn Aufiau der Moralen. Frankfurt:
Klostermann.
Schmidt, V. 1994. "Bounded Justice." Social Science Information 33: 305-333.
Schweisguth, E., Grunberg, G. 1996. "Bourdieu et la mistre: une approche
rCductionniste." Revue fran~aisede science politique, 46, 1: 134-155.
Shweder, R. A. 1977. "Likeliness and Likelihood in Everyday Thought: Magical Think-
ing in Judgments about Personality." Current Anthropology 18, 4, December:
637-659.
Simmel, G. 1892. Die Probleme der Geschichtsphilosophie. Munchen: Duncker &
Humblot.
Simmel, G. 1989. Gesammelte Schriften zur Religionssoziologie. Berlin: Duncker &
Humblot.
Smith, A. 1976[1776]. An Inquiry into the Nature and Causes of the Wealth of Na-
tions. London: W . Strahen and T. Cadelll Oxford: Clarendon Press.
Sperber, D. 1996. La contagion des idLes. Paris: Odile Jacob.
Strauss, L. 1953. Natural Right and History. Chicago: Chicago University Press.
220 The Origin of Values