IS THERE AN ISLAMIC CIVILIZATION?

Yilmaz Esmer

Regarding religion as the foundation of civilizations is certainly not a novel or recent idea. For
thousands of years religion has formed (and to a large extent still does form) the basis of an
individual’s identity. Even today, roughly two-thirds of the population of a traditionally secular
Islamic country say they are first Muslim and then Turkish rather than the other way around.1 It
was roughly a millennium ago when the Crusaders risked their lives in “infidel lands” in the
name of God and Jesus.2 After more than a thousand years, and in the same lands, Palestinian
youths explode themselves almost daily without the blink of an eye, this time in name of Allah
and Mohammed. Plus ça change, plus c’est la meme chose, as the French saying goes!

It is, therefore, no surprise that civilizations have frequently been identified by religions. Thus,
we talk about an “Islamic civilization” or a “Christian civilization” implying that each religion
has its civilization. The assumption here is that almost all Muslims, for example, have certain
important cultural traits, values that they more or less share and that are different from those of
non-Muslims. To be sure, there are those trivial differences in behavior and beliefs that are
obvious even to the casual observer. Thus, cows are sacred for Hindus but are sacrificed annually
to please Allah by Muslims who, in turn, do not even touch pigs let alone eat pork. Christian
monasteries also functioned as wineries while the pleasure of sipping the smallest amount of duly
fermented grapes was strictly forbidden for the Muslim believer. But these are rather superficial
differences from a social science perspective3 and they could hardly be regarded as sufficient for
defining civilizations. What, then, are the criteria for defining a civilization? This seems to be the
first question one needs to address –an obvious starting point. Second, we need empirical

1
Turkish World Values Survey carried out in December 2000 and January 2001. When asked whether they regarded
themselves above all as Muslim or Turkish, 68 percent chose the former.
2
Whether or not some, most or all religious wars had more material and tangible motives at the level of ruling elites
is immaterial here. The Crusaders and others were innocent of such designs and could be not have been persuaded to
welcome martyrdom without direct reference to divine powers.
3
Among those who took a dissenting view on this topic, Marvin Harris was perhaps the most prominent social
scientist. Harris ( ) argued that all these “riddles of culture” had sound economic and materialistic explanations.
Thus he saw them as dependent variables which were only consequences of a culture’s adapting itself to the
environment for survival.

1
evidence showing that sets of societies that are classified into civilizations are indeed distinctive
and different with respect to these criteria. Many authors, classical and contemporary, have
proposed taxonomies for grouping civilizations and cultures. However, until recently, it was not
possible to test these in any meaningful way because the data simply did not exist. Fortunately,
this problem is being overcome to a large extent thanks to the World and European Values
Surveys. We are now in a position to compare societies, including predominantly Islamic ones,
with respect to a wide range of cultural values using systematic data since the last wave of the
WVS includes a number of Islamic countries for the first time ever.

This paper is an attempt in that direction: In what respects and to what extent Islamic cultures are
different from others? The answer to this question would always have had significant theoretical
import. However, Huntington’s much acclaimed and, at the same time, bitterly criticized “clash
of civilizations” thesis (Huntington, 1993;1998) and then the tragic events of September 11, made
it the “burning question” of the day. Therefore, “the clash of civilizations” is a natural starting
point for a paper on the subject of contemporary Islamic cultures.

Huntington and The “Clash of Civilizations”

Huntington’s thesis on the world’s new fault line after the Cold War and the collapse of the
Soviet Union has been discussed so widely both in academic publications and in the media that
one need not summarize it here once again. Therefore, I shall limit myself to those points which
will be helpful in deriving testable hypotheses.

It goes without saying that a definition of the concept of “civilization” is crucial to any
meaningful discussion and test of the thesis. Nevertheless, and as noted by Norris and Inglehart
(2002:note 5), Huntington does not give us a clear definition of civilization except in general
terms and uses civilization and culture interchangeably. He writes (1998:41) that “a civilization
is a cultural entity outside Germany. […] This distinction has persisted in German thought but
has not been accepted elsewhere. […] …efforts to distinguish culture and civilization […] have
not caught on, and, outside Germany, there is overwhelming agreement with Braudel that it is
‘delusory to wish in the German way to separate culture from its foundation civilization.’” Thus,
“civilization and culture both refer to the overall way of life of a people, and a civilization is a

2
culture writ large. They both involve the ‘values, norms, institutions, and modes of thinking to
which successive generations in a given society have attached primary importance’”(Huntington,
1998:41).4 Of course, this does not tell us which specific “values, norms, institutions, and modes
of thinking” we are to analyze let alone how to measure them. However, Huntington (1998:42) is
unequivocal on one point: “Of all the objective elements which define civilizations, however, the
most important usually is religion.” Thus, “religion is a central defining characteristic of
civilizations.” (1998:47)

If we are, then, to test Huntington’s thesis, religion is our crucial independent variable. Put
differently, “the values, norms, institutions, and modes of thinking” should, to a large extent, be
dependent on religious affiliation and faith. Religion is also the clue for classifying religions and
it is proposed that in the contemporary world we can delineate the following major civilizations
(Huntington 1998:45-47):
Sinic/Confucian
Japanese
Hindu
Islamic
Orthodox
Latin American
African (possibly)

We now derive testable hypotheses from these arguments:

H1. Societies, which belong to one civilization, should be significantly different from societies
that belong to another civilization with respect to their values, norms, etc.
H2. Variance in values within civilizations should be smaller than variance between civilizations.
H3. Within a given society, groups affiliated with a certain religion (that forms the “central
defining characteristic of civilizations”) should be significantly different from groups affiliated
with other major religions.
H4. Within a given society, within group variance in cultural values should be smaller than
between group variance.

4
Huntington quotes Bozeman (1975) here.

3
Of course, it is one huge step from here to require that different civilizations must clash.
Nevertheless, if the theory is to hold, at a minimum we must demonstrate that differences do
indeed exist and that a religion-based classification of cultures is plausible.

Commenting briefly on each one of the four hypotheses stated above, H1 predicts that if a given
religion is predominant in a given society, then the cultural traits, characteristics that are typical
of the corresponding civilization will be more observable among the people of that society. It
follows that the higher the proportion of individuals affiliated with the given religion, the more
pronounced the traits will be in that society.

H2 states that, whatever the important characteristics that define a civilization (culture) are, a
civilization and a society belonging to that civilization should be more or less coherent within
itself. There is much variation among various breeds of dogs but a terrier and a golden retriever
should have more in common between herself than either one with cat or a horse.

If religion is at the foundation of cultures, then individuals living in a given society and belonging
to different faiths should display certain characteristics that are thought to be typical of their
religion and thus their civilization. Therefore, H3 predicts that Muslims, for example, and
Catholics living in the same society should exhibit certain differences in their values and norms.
Needless to say, each group should be closer to the mean of its respective civilization. A finding
to the contrary would have devastating effects on the theory.

H4 corresponds to H2 when the unit of analysis is individuals in one society. To repeat, if we are
to talk about two different groups in a society, we cannot expect as much difference within the
group as there is between the groups.

In this paper, I limit myself to a comparison of the Islamic culture with each of four major
cultures only: Protestant, Orthodox, Catholic and Hindu. Furthermore, the analysis uses a within
society approach and deals mainly with H3 although H2 and H4 are also addressed to some
extent.

4
Before presenting the findings, the all-important question of “what is it that we are comparing”
should be dealt with. Put differently, the dimensions for defining and grouping civilizations
should be made specific.

Dimensions of Culture and Cultural Values

This subtitle has been the subject of a great number of volumes and the literature is replete with
proposals for different value dimensions (many of them theoretical and some tested with data)
from different perspectives –anthropological, sociological, social psychological and others. It is
beyond the scope of this paper to even begin to summarize or discuss these various approaches
here and I certainly would not be doing justice to them within these limits of space. However,
Ronald Inglehart’s scheme and the modifications proposed to it by Halman and Pettersson are
particularly relevant to the analysis that follows and, therefore, deserve attention.

Inglehart (Inglehart, 1997; Inglehart and Baker, 2000) has proposed a two-dimensional scheme to
position societies with respect to the myriad data collected through World Values Surveys. The
two dimensions are named “Secular-Rational vs. Traditional” and “Survival vs. Self
Expression.”5 The first dimension, with traditional values at one and secular-rational values at
the opposite pole, include measures of the importance of God in one’s life, importance of
obedience and religious faith in child rearing rather than independence and determination, attitude
towards the justifiability of abortion, degree of sense of national pride, and respect for authority.
Clearly, traditional values emphasize the importance of religion, obedience, respect for authority,
national pride and also are opposed to abortion. The indicators of the second dimension, survival
and self-expression values, are emphasis on economic/physical security or quality-of-life,
individual happiness, attitude toward signing a petition, degree of approval of homosexuality, and
interpersonal trust (Inglehart and Baker, 2000:24).6

5
The term “well-being” instead of “self-expression” is used in Inglehart, 1997.
6
The indicators are summarized from Table 1 of Inglehart and Baker, 2000. Although the list of indicators may vary
in different publications (see, for example Inglehart, 1997 where the author uses 22 indicators instead of the 10
summarized above), the thrust of the argument that this two-dimensional scheme can adequately represent a society’s
value system does not change.

5
In various publications, it has been reported that there is much empirical support for the two-
factor model summarized above. “The items in each dimension are highly intercorrelated. The
two dimensions explain 70 percent of the total cross national variation among these 10 variables.”
(Inglehart and Baker, 2000:25) Is there a theoretical justification for the choice of this particular
set of indicators? Halman and Pettersson (2002) question this point and suggest that, from a
theoretical perspective, the inclusion of “some of the indicators are more or less ‘accidental.’ For
instance, they argue that individual happiness is “an emotion rather than a value” but, more
importantly, it is context dependent. To give another example, they are not convinced that
attitudes towards homosexuality and abortion are used as indicators of not the same but rather
different dimensions. One suggestion they make is to replace the survival-self expression axis by
what is called “civic orientation.”

At this point, I would like to explain briefly –and hopefully justify- my choice of criteria for
comparing Muslim and non-Muslim populations in a number of societies.

Summarizing a wide-range of values with a limited number of dimensions has the great
advantage of parsimony. Regardless of whether or not one likes the choice of indicators,
Inglehart’s two dimensions, for example, is extremely useful in placing a society on the world
cultural map and makes intuitive sense. It even has the advantage of lending itself readily to
graphic representation. Leaving theoretical considerations aside, summarization to such extent,
however, inevitably risks losing much detail. With that consideration, in this paper I take a
somewhat different approach and compare Islamic and non-Islamic groups with respect to four
main criteria which have received attention in the literature. In other words, in deciding the
important characteristics that should help us to distinguish cultures, I take a blindfolded approach
to data. At this point the issue is not to see which variable or question correlates with which
others but rather to shortlist broader value dimensions which have been proposed to have
significant effects on economic and political structures and change. The question then is whether
these are related to religious affiliation. Surely, these broad latent concepts will have to be
operationalized with the help of individual variables (questions in the context of surveys).
Practical data constraints aside, my choice of indicators was also motivated by theoretical factors
with no regard to empirical considerations. Only then do I carry out tests to see if the indicators

6
should form a scale. And if the answer to that question is negative, so be it. The variable is not
dropped out of the analysis –just not included in a scale and is analyzed individually.

The four value orientations on which I wish to base my comparisons are:

a. Protestant ethic
b. Social capital
c. Modernity
d. Democracy

These four dimensions can provide us with a good basis for the comparison of populations with
respect to their values and address the major arguments found in the relevant literature. Some are
(modernity, for example), nevertheless, still too broad and are divided into subcategories before
they are operationalized.

Data and Measurement

Data from the 1999-2001 wave of World Values and European Values Surveys are used to carry
out the analyses. The set used includes data from 70 societies with a total sample size of 95,341.
However, as explained in the next section, data only from eight countries are utilized.7 Although
the questionnaires cover the widest possible range of norms, values and attitudes, it goes without
saying that no questionnaire is ideal for all purposes. Furthermore, there are always problems of
missing data and some questions are not asked in some countries for various reasons.
Notwithstanding these usual constraints and limitations, we are in the fortunate position of
finding at least one –and usually more than one- indicator of the concepts that need to be
measured.

a. Protestant Ethic
Very few theories in social sciences can claim the fame and attention that Weber’s (1958) thesis
on the prime cause of capitalist development has enjoyed. Weber’s argument is so widely known
that the reader is spared yet another summary here. Thrift, determination and hard work as an
end itself are the core values of the Protestant Ethic.
7
I do not repeat here detailed descriptions of WVS and EVS projects which are readily available in a number of
publications and, more conveniently, in the website www.worldvaluessurvey.org .

7
Both WVS and EVS questionnaires contained a list of items which are thought to be important to
instill in children. The respondents were asked to choose five from among this list, which
included hard work, thrift and determination. These three indicators come as close to measuring
the concept as one can ideally hope. A scale combining the three was not formed, however, since
choosing one of the three decreases the probability of choosing one or both of the other two. All
three variables are dichotomous with ‘0’ indicating that the item was not chosen and ‘1’
indicating that the item was mentioned as important.

b. Social Capital
Almond and Verba (1963), in their now classical five-nation study, emphasized the importance of
interpersonal trust in a society as one of the forming blocks of “the civic culture” some four
decades ago. Since then, and particularly in the 1990’s, trust has become in vogue and received a
great deal of attention. Voluminous books have been written with this title (some examples are,
Fukuyama, 1995; Misztal, 1996; Warren, 1999). It is regarded as the key component of social
capital which, in turn, is argued to be a significant independent variable for sustainable economic
growth and the consolidation of democracy. Putnam’s (1993) seminal work on the role of social
capital in the development of democracy in Italy has rightfully stirred much excitement and
debate as well as criticism over the role of social capital. In addition to trust, cooperation, norms
of reciprocity and civic engagement comprise the core values of the concept of social capital.8

Three indicators of social capital are used in the analysis:

Interpersonal trust. A dichotomous variable with respondents saying most people can be trusted
coded ‘1’ and those who say one needs to be careful in dealing with others coded ‘0’.

Respect. Respondents choosing “respect for others” from among a list of items important in
raising children are coded ‘1’ and others ‘0’. This question is used to operationalize a
cooperative attitude.

8
For recent treatments of and differing viewpoints on social capital see Dasgupta and Serageldin (2000) and Rotberg
(2001).

8
Signing a Petition. As a measure of civic engagement, the question about signing a petition was
used. Although a very benign and easy way of displaying civicness, it still requires that such an
act is not likely to bring any harm on the individual. Therefore, it can be “context contaminated”
in certain authoritarian countries. Variable is coded ‘1’ for respondents who have signed a
petition and ‘0’ for those who said they might or they never would do such a thing.

c. Modernity
Modernization theories, whose origins can be traced to Karl Marx, seem to have a cyclical life.
They reach a peak in popularity only to be followed by intense criticism and then a trough. Then
they undergo some sort of adaptive mutation and regain a status of respectability. The debate
may continue but one thing is clear: modernization, modern, modernity are now household words
and are not meaningless. Modernity is a syndrome with a number of symptoms which usually are
intercorrelated. It may not be a unilineal process. It may not necessarily lead to convergence on
all fronts. It may at times have been presented as a Western and ethnocentric concept.
Nevertheless, modernization, almost always accompanied by economic growth, has certain
common discernable consequences. As Taylor (2001: 184-5) puts it “what is required by the
wave of modernity is that one come up not with identical institutions but with functionally
equivalent ones. The ‘bottom line’ is, for example, competing successfully in the international
market. More than one kind of firm and business culture can enable this.” According to him
there is “a gamut of alternative modernities” in the contemporary world. Just like we do not
expect modernization to produce identical institutions in societies, we do not expect the process
to produce individuals with identical values. We do predict, however, that the values of a
“modern” individual will tend to change in a certain general direction. By arguing that the
process will produce certain changes in cultural values “we do not assume that all elements of
culture will change, leading to a uniform global culture: we see no reason to expect that the
Chinese will stop using chopsticks in the foreseeable future, or that Brazilians will learn to polka.
But certain cultural and political changes do seem to be logically linked with the dynamics of a
core syndrome of Modernization.” (Inglehart, 1997:69; italics original)

9
It would be a monumental task and a heroic accomplishment to come up with a complete and
agreed-upon list of cultural changes that modernization is expected to produce. Certain items,
however, could serve as a common denominator. I suggest the following:

Importance of religion. Industrialization, economic growth and prosperity, advances in science

and technology, and higher levels of education are all expected to result in a decrease in
religiosity –both belief and practice. There are exceptions, no doubt, and the United States is a
case in point. But the negative correlation between modernity and religiosity generally holds and
we can make a probabilistic prediction to that effect. Our questionnaires include a number of
questions on religiosity and I use the following:

i) A 10-point scale measuring the importance of God in one’s life. The coding is reversed so that
(as with all variables) higher scores indicate higher levels of modernity. Thus a score of ‘1’
means that God is extremely important and ’10’ indicates that God has no importance in the
individual’s life.

ii) A composite scale comprising seven dichotomous variables with each one coded ‘0’ for high
and ‘1’ for low religiosity. The variables are:
Is the respondent a religious person?
Does the respondent believe in:
God?
Life after death?
Hell?
Heaven?
Does the respondent draw comfort and strength from religion?
Does the respondent choose religious faith as important in raising children?

These seven variables are all positively correlated and a scale is constructed by adding them. The
resulting measure has a minimum value of ‘0’ (extremely religious) and a maximum of ‘10’ (not
religious at all). The scale has high reliability with a Cronbach’s alpha of 0.83 for WVS countries
and 0.88 for EVS countries.

10
Secularism. Modernization is expected to encourage secular values which I define here not in
terms of individual faith but rather with respect to attitudes toward the separation of the church
and the state. A scale is constructed with two questions in order to measure secularism:
Politicians who do not believe in God are unfit for public office.
It would be better if more people with strong religious beliefs held public office.
Both questions have five response categories (strongly agree, agree, neither, disagree, strongly
disagree) so the resulting scale has a minimum value of ‘2’ (extremely non-secular) and a
maximum value of ‘10’ (very secular). The scale has an alpha of 0.65 for WVS and 0.76 for EVS
countries.

Efficacy. A 10-point scale is used to measure efficacy, that is, how much freedom and control
individuals think they have over their lives. At the lower end are those respondents who believe
“what they do has no real effect on what happens to them” and at the higher end are those who
think they have complete freedom over their lives.

Marriage and family. Traditional cultures attach great importance to the family and the
institutional arrangement that bring it to existence. Inglehart and Baker (2000:25) write that “The
importance of family is a major theme: In traditional societies a main goal in life is to make one’s
parents proud –one must always love and respect one’s parents regardless of how they behave.
Conversely parents must do their best for children even if their own well-being suffers.” Seven
indicators have been added to construct the “importance of family and marriage scale.” They are:
Children must love and respect parents unconditionally
Parents are dutybound to sacrifice their well being to do the best for their children
Child needs a home with both father and mother to grow happily
Women have to have children in order to be fulfilled
Marriage is an outdated institution
Approval of woman having a child without marriage/stable relationship
More emphasis on family life would be good/bad

11
All of these seven questions are coded ‘0’ or ‘1’ with the former indicating emphasis on
traditional family values. Thus a maximum score of ‘7’ on the scale means the respondent does
not attach importance to any of the family values included. The scale has an acceptable reliability
with a Cronbach’s alpha of 0.48 for WVS and 0.46 for EVS countries.

Female employment. Widespread participation of women in the workforce is one of the most far-
reaching consequences of industrialization that had profound impact on the whole social
structure. Rates of female employment continue to increase and to affect people’s (women and
men) outlook on life. In their pioneering study of modernization in six countries, Inkeles and
Smith (1974:18) proposed “to classify as modern those personal qualities which are likely to be
inculcated by participation in large-scale modern productive enterprises such as the
factory…[italics original]”9 Employment changes a woman’s role as a wife and a housewife and
it changes her values, her general outlook on life. Work for pay is not a role assigned to women
in traditional societies. Therefore, acceptance of this new role is a good indicator of one
important aspect of modernity. Two variables are added to measure attitudes towards working
women:
Working mothers can establish warm and secure relationship with their children
Both the husband and wife should contribute to household income
Both questions have four answer categories (strongly agree, agree, disagree, strongly disagree)
and both are coded so that higher scores indicate approval of working women. The combined
variable has the lowest reliability scores of all the scales constructed (alpha=0.39 for WVS and
0.31 for EVS countries).

Sexual tolerance. Sexual liberation is a characteristic more of the postmodern, postindustrial

rather than the modern, industrial society. More liberal, permissive sex norms tolerant of sex
between unmarried couples and homosexuality have been gaining increased approval in the West.
Some countries, like the Netherlands to be followed by others, have legalized homosexual
marriages. Linking this development to the increase in the proportion of postmaterialists,
Inglehart (1990:195) writes that “Postmaterialists are far more permissive than Materialists in
their attitudes toward abortion, divorce, extramarital affairs, prostitution, and euthanasia.” This
9
Inkeles and Smith (1974) studied a sample of males only in six countries, but their basic conclusion is, I believe,
even more valid for women.

12
correlation between sexual norms and materialism/postmaterialism is confirmed seven years later.
(Inglehart, 1997:276-80)

The last dimension of modernity that I analyze is sexual norms. It has already been observed that
Inglehart (1997), and Inglehart and Baker (2000) have treated attitudes towards homosexuality
and abortion as components of two different dimensions and that Halman and Pettersson (2002)
disagree. In this paper, the scores of three 10-point scales are added to construct a sexual
tolerance scale:
Justifiability of
Homosexuality
Abortion
Prostitution (or casual relationships in Europe).
However, approval of prostitution question was only optional in the EVS questionnaire and was
asked in a very limited number of countries. Therefore, in those countries it is replaced with
“casual relationships.” The two, obviously, are not the same thing but perhaps functionally
equivalent for our purposes since all of these four questions are positively correlated with
coefficients ranging between 0.38 and 0.65. In both cases (that is using prostitution or casual
relationships as the third item of the sexual tolerance scale) reliabilities are high with alphas of
0.56 and 0.67 for WVS and EVS countries, respectively.

d. Democratic Culture
What is broadly referred to as the “democratic culture” is the last dimension of our comparisons
between Muslim and non-Muslim populations. Democracy needs democrats. It is difficult, if not
impossible, for any system (even at the micro level) to survive unless an appreciable proportion
of its members have internalized the basic values and norms of the system. Is it possible that a
family can survive for long if husband and wife believe that it is an outmoded, useless,
dysfunctional institution that does not serve the interests of any member?

No one disputes that commanding high levels of legitimacy and enjoying wide support of its
citizenry is helpful to the survival of a political system and that this is particularly true for
democracies. However, what is important here is whether the core values of the system are

13
shared by the citizens. Expressed support for this government or that regime can change
dramatically within relatively short periods depending on perceived performance. Such support
is superficial and fickle if it is not sustained by a deeper value structure that is in congruence with
the political system. In this respect, Klingemann’s (1999) concept of “dissatisfied democrats” is
of particular importance. As he explains, dissatisfaction with how the system is functioning does
not pose a threat to democracy. “Quite to the contrary they [dissatisfied democrats] may well be
the hope for the future of democratic governance.” (Klingemann, 1999:56)

This is very important for the specific purposes of the present article because it tells us that
questions about government or system performance are less than ideal to gauge public support for
democracy. Similarly, general and abstract questions of the type “do you believe democracy is
the best system of government?” have low validity for at least two reasons. First, being a
democrat is very fashionable in the contemporary world (Klingemann, 1999) and it is very easy
to be a “democrat” in the abstract. Thus the answers can be expected to be biased towards
democracy because of global trends and the operation of a social desirability effect. Second, such
a question will suffer from context contamination. An individual living under an authoritarian
system that does not deliver to her/his satisfaction is more likely to say “democracy is best” that
an individual who is very satisfied with the performance of the system regardless of its
authoritarian or totalitarian nature. I propose that values related to, for instance, tolerance,
freedom, or participation are better measures of democratic culture.

The following analysis utilizes a five dimensional scheme for assessing and comparing
democratic values in different cultures.

Tolerance. It goes without saying that a democrat is a tolerant person –tolerant of different ideas,
different lifestyles, different religions and ethnicities, etc. WVS/EVS questionnaires have an
excellent and time-tested battery for measuring the level tolerance. Respondents are presented
with a list of potentially “undesirable” groups and are asked to pick the ones that they would not
like as neighbors. The scale of tolerance used in the analysis includes the following items:
People with a criminal record
People belonging to a different race

14
 Emotionally unstable people
Muslims (or relevant religious minority)
Immigrants/foreign workers
People with AIDS
Jews (or another visible minority group)

Two groups that are included in the battery are not included in this scale: homosexuals and heavy
drinkers. Homosexuals are not a scale item because, as already discussed, sexual tolerance is
treated as a separate dimension and it is felt that this question should belong to that specific
dimension rather than a measure of general tolerance. Heavy drinkers are excluded because
Islam specifically forbids the use of alcohol. Therefore, this item does not have the same status
as the others. For a Hindu, being tolerant of, say, Muslims or former criminals or immigrants and
being tolerant of killing cows are not the same thing.

The seven items in the tolerance scale are all dichotomous and higher scale values (maximum is
7) indicate higher levels of tolerance. In some countries, questions about certain groups
(particularly religious minorities) were not asked. In those countries, the scale is constructed
using six (for example, Montenegro) or five (for example, India, Nigeria) items.10 The scale has
high reliability with an alpha of 0.74 for WVS and 0.77 for EVS countries.

Support of freedom. Support for freedom is measured with the help of a variable constructed from
two questions included in the materialism/postmaterialism battery. Respondents were asked to
choose the most and the second most important aims of their countries from a list that contained:
Maintaining order in the nation
Giving people more say in important government decisions
Fighting rising prices
Protecting freedom of speech
The variable is coded ‘1’ if the respondent chooses “protecting freedom of speech” either as the
first or the second aim and ‘0’ otherwise.

10
Needless to say, these five or six-item tolerance scales were only used for within country comparisons.

15
Participation. Participation is an important dimension of “civic culture.” An apathetic and
nonparticipating citizenry that is as little involved in politics as possible is the dream of many
dictatorships. Democracy requires the opposite –a point emphasized by all students of
democratic culture. A comparative analysis of democratic values should not ignore the
participation dimension. Our measure of political participation includes three items:
How much the respondent is interested in (in EVS talk about) politics?
Would the respondent join a boycott?
Would the respondent attend a lawful demonstration?
All three items have three response categories resulting in a scale with a maximum of 9 (very
high participation). Alpha coefficients are 0.56 both for WVS and EVS countries.

Problems with democracy. Three items that are collected under this heading tap opinions on
problems that are frequently associated with democratic systems. Unfortunately they, too, can be
influenced by context. If an individual is living under a democratic system where stability is a
serious problem and governments come and go once every few months (recall France before the
5th Republic or Italy in the not too distant past) he/she is more likely to agree that indecisiveness
is a problem for democratic systems. Similarly, if internal security and terror are serious
problems, one is expected to accept the evaluation that democracies have difficulties in
maintaining order. With these caveats, the items that form the “problems of democratic systems”
scale are:
In democracies, the economic system runs badly
Democracies are indecisive and have too much quibbling
Democracies are not good in maintaining order
All three items are scored on a four-point scale (strongly agree, agree, disagree, strongly
disagree). Higher scale values (maximum 12) represent pro-democracy attitudes. Alpha
coefficients are 0.76 for WVS and 0.79 for EVS countries, respectively.

Alternatives to democracy. It is interesting to note that an appreciable number of people who

think democracy is the best system of government, can also accept solutions that are incompatible
with a democratic system: another reason for arguing against “is democracy good?” type of direct
questions. Why would someone who expresses agreement with a statement like “democracy may

16
have its problems but it is better than any other form of government” also think “it would be good
to have the army rule?” And this is not a hypothetical question. Even in the sample of European
countries with levels of education much higher than the world average, close to 10 percent
(unweighted sample from 32 European societies) fall into that category! Thus, alternatives to
democracy scale, comprising three items listed below, is another measure indicative of the
support for democracy. The items are:
Having a strong leader who does not have to bother with elections or parliament
Having experts, not government, make decisions according to what they think is best
Having the army rule
Respondent were asked to rate these as “very good,” “good,” “bad,” or “very bad.” Again with
higher scores indicating higher degrees of support for democracy, alpha’s are 0.57 for WVS and
0.52 for EVS countries.

The scheme developed for the comparison of Islamic groups with others is summarized in Table
1. The results reported in the next section are based on analyses following this scheme.

(Table 1 about here)

Results: Global Perspective

Having completed the discussion of our measurement model, we can now take up the hypotheses
derived from Huntington’s thesis. It will be recalled that the first hypothesis, H1, was:
Societies which belong to one civilization should be significantly different from societies that
belong to another civilization with respect to their values, norms, etc.

Inglehart and Norris (2002) report a test of this hypothesis. To my knowledge, this is the first
published test, based on extensive comparative data, of Huntington’s thesis. Briefly, the
approach taken by Inglehart and Norris is to code everyone living in a society as belonging to the
same civilization (determined by the religious group that has the plurality in that society) and then
comparing individuals living in predominantly Islamic societies with those living in Catholic,
Protestant, Orthodox, Central European, Latin American, Sinic/Confucian, Sub-Saharan African,

17
Hindu and Japanese societies. Individuals are compared with respect to -surprisingly, not the two
dimensions developed by Inglehart but rather- “democratic political values” (approval of
democratic performance, democratic ideals, religious leaders, strong leaders) and “liberal social
values” (approval of gender equality, homosexuality, abortion, divorce). Evaluating a series of
multiple regression models with civilization as the independent variable and including a number
of control variables, they conclude that “Any black-and-white ‘Islam versus the West’
interpretation of a ‘culture clash’ as conveyed by the popular media is far too simplistic.”
(Inglehart and Norris, 2002:15) Their findings indicate that there is indeed a difference between
Islam and the West but the difference concerns, in their words, “the eros and not the demos.” In
other words, the observable differences between the Islamic world and others pertain to values
related to gender differences and sexual liberalization rather than support for democracy.

Inglehart and Norris do not test H2 about the variances within civilizations. This hypothesis is
important because it will give us an idea about the consistency of civilizations with respect to
cultural values. This point has been made frequently by scholars who object to the idea that there
is “one Islam” and that the variances within it are insignificant. If the objection is not valid, we
should expect low variances within the Islamic civilization. –certainly lower than the total
variance that includes all other civilizations in addition to Islam.

A standard ANOVA test is not appropriate in this context because of the huge sample size which
yields thousands of degrees of freedom for within groups estimates. Statistical significance under
these circumstances is not very meaningful since it is impossible not get significant F ratios.11
Instead, we take an informal approach and simply examine the standard deviations for four
civilization groups. Societies that are included in the 1999-2001 WVS/EVS surveys are
classified following Huntington’s scheme and Inglehart and Norris’s (2002) criteria with minor
modifications (the full list is given in Table 2). For purposes of brevity, results for four
civilizations (Protestant, Catholic, Islamic and Orthodox) are compared in Table 3 which tells a
very interesting story.

(Table 2 about here)

11
Technically, statistical tests are only appropriate for probability samples. Since a number of countries employed
quota sampling, statistical tests should only be regarded as an approximate guide.

18
Figures in Table 3 measure only the coherence within civilizations with respect to the measures
employed but certainly do not give any clues about the existence or direction of differences.
Therefore, it should be interpreted with reference to the means. For example, we may find that,
on the average, a given civilization has comparatively lower levels of sexual tolerance as
indicated by group means. If, however, the group has a very high variance for this variable, then
we know that although on the average members of Civilization X are less tolerant, there are great
differences among them and one would be hard pressed to refer to them as if they all display their
“civilization characteristic” –in this example, low sexual tolerance. Conversely, a low mean and
a small variance would be an indication that low tolerance is a consistent characteristic of that
culture. In other words, when defined in terms of the dimension in question, we could speak
about the existence of a civilization.

(Table 3 about here)

The standard deviations reported in Table 3 can be summarized as follows:

a. With respect to values related to Protestant Ethic, standard deviations for all four
civilizations can be considered equal.
b. For all indicators of democratic culture, populations of Islamic societies are as likely to
differ around their means as the other three groups.
c. Efficacy and social trust variables (with the possible exception of signing a petition)
have roughly similar standard deviations for all four groups. Islamic civilization can be said to be
somewhat more consistent about signing a petition.
d. Interestingly, Islamic culture has the largest variance in attitudes towards female
employment. Apparently, the Islamic world is divided on that issue.
e. The standard deviation for Islamic cultures is somewhat low for the marriage and
family scale.
f. The most striking difference in standard deviations between the Islamic culture and
others is sexual tolerance. This scale has a variance of 2.67 for people living in Islamic societies.

19
The next lowest standard deviation is 6.00 (Orthodox) and Protestants have a standard deviation
of 7.69. This finding warrants a look at the scale means:
Protestant 14.5
Catholic 11.8
Orthodox 9.3
Islamic 4.3.
Out of a possible high of 30, Muslims or non-Muslims living in Islamic societies score a mean of
4.3 on sexual tolerance scale. This is remarkable indeed. And coupled with a very low standard
deviation, we find a telltale sign for the existence of a civilization here.
g. The next group of variables that stand out in Table 3 are related to religiosity. God is
extremely important in the lives of Islamic populations who are also considerably more religious
that others. Furthermore, the variance for these measures in Islamic culture is much lower than
any of the other three cultures.

These conclusions are very much consistent with the findings reported by Inglehart and Norris
(2002). Two factors gain prominence in defining an Islamic civilization: high religiosity and low
sexual tolerance. And with respect to both of these characteristics, Islamic culture is much more
coherent and consistent then Protestant, Catholic or Orthodox cultures. It seems that one cannot
speak about a clear and distinct Islamic cultural zone if one is concerned with values related to
Protestant Ethic, social capital or democratic culture. But it makes sense to refer to an Islamic
civilization within the context of religiosity, treatment of women, and sexual tolerance (or rather
the lack thereof).

Results: Country Analysis

The conclusions of the previous chapter were based on global level analyses where it was
assumed that the predominant characteristics of a society tend to determine the value systems of
individuals living in that society regardless of religious faith. To test our third hypothesis, we
have to change our unit of analysis and turn our attention to comparisons of groups within
countries. If we regard religion as the defining characteristic, the foundation of civilizations, then
H3 should be confirmed such that the values of different religious groups living in one society
should differ in the predicted direction. Referring to the previous section, Muslims in a given

20
country should have significantly higher levels of religiosity and lower levels of sexual tolerance
than, say, the Protestant citizens of that country. To test this hypothesis, we have to have data on
multi religion societies with sizeable Muslim populations. At the time of this writing, we have
data for eight such societies:
Bulgaria
Macedonia
Bosnia
Montenegro
Nigeria
Uganda
India
This set allows us to compare Islamic populations within a country with Protestants (Nigeria and
Uganda), with Roman Catholics (Uganda), with Hindus (India) and finally with the Orthodox
populations of Albania, Bosnia, Bulgaria, Macedonia and Montenegro. The Muslims in these
societies are indigenous populations and not immigrants (who could still carry the traits of their
home countries) and this is an advantage for our purposes. Nevertheless, it is clear that this set of
countries presents severe data limitations. First, five out of eight are in the Balkans and allow
comparisons only with the Orthodox culture. Second, this is an area where ethno-religious
conflicts reached war proportions in the very recent past. Third, all eight are low –and mostly
very very low- income countries. And finally, these countries score so much lower than average
on a number of value dimensions that it may be difficult to find differences between groups
within a society.

H3 was derived from Huntington’s thesis. But, among others, Coleman (1990) and Putnam
(1993) would also make the same prediction since they argue that a correlation exists between
religious affiliation and values (in their case values related to social capital). Halman and
Pettersson (1998) analyzed the relationship between social capital and religious denomination in
Germany, Spain, Sweden and the U.S. They concluded that, with the individual as the unit of
analysis, one could not readily confirm the correlation. Inglehart and Baker (2000) reached a
similar conclusion with regards to interpersonal trust. They found Catholics to have about the
same levels of trust as Protestants within given societies. They have an explanation: “…once

21
established, the cross-cultural differences linked with religion have become part of a national
culture that is transmitted by the educational institutions and mass media of given societies to the
people of that nation. Despite globalization, the nation remains a key unit of shared experience,
and its educational and cultural institutions shape the values of almost everyone in that society.”
(Inglehart and Baker, 2000:37)

In Tables 4 through 7, I present within country comparisons for all value dimensions that have
been described in previous sections and summarized in Table 1. Entries in bold indicate a
significance level of 0.05 or better (two-tailed difference of means test with no assumptions about
equality of variances). As explained above (note 11), because of sample properties, significance
levels should not be regarded as exact probabilities.

a. Protestant Ethic.
With respect to the first two indicators (hard work and thrift) one does not find any consistent
differences between Muslims and non-Muslims in our sample of eight countries. In fact, Muslim,
Orthodox, Catholic and Hindu populations attach similar (almost equal) levels of importance to
instilling values of hard work and thrift in children with two exceptions: Orthodox groups score
low on hard work in Macedonia and on thrift in Montenegro. But these indeed are exceptions
and overall similarity of groups prevail for these two variables. The findings for the third
indicator (determination, perseverance) are more interesting: in six of the eight countries,
Muslims score significantly lower than other groups and in the remaining two (Bulgaria and
Uganda) the differences are in the same direction although they do not attain statistical
significance.

(Table 4 about here)

Can this finding be generalized to the rest of the world? The answer is ‘yes’ when we examine
the means for civilization groups. Those living in the Islamic zone attach significantly less
importance to determination and perseverance in children than Protestants, Catholics or the
Orthodox. Is there a determination and perseverance problem in the Islamic culture? Following
Weber, could this finding give a clue as to why Islamic countries have been painfully slow on the

22
road to capitalist development and industrialization? It is certain that more research –both
theoretical and empirical- is needed on this topic.

b. Social Capital
An examination of the means reported in Table 5 shows that there is no consistent pattern for any
of the indicators of social capital. Some differences are significant but are usually in opposite
directions in different countries. By looking at the data from these eight countries, one certainly
cannot say there is a correlation between religion (in our case Islam) and social capital. Thus, the
conclusions reached by Halman and Pettersson (1998) for four developed Western countries are
confirmed. They found no consistent differences between Catholics and Protestants and we find
no consistent differences between Muslims and others.

(Table 5 about here)

c. Modernity.
Religiosity. It has already been observed that, at the global level with countries as the units of
analysis, religiosity is one of the two important characteristics that may define Islamic
civilization. By and large, this finding is confirmed at the micro level. In Albania, Bulgaria,
Macedonia, Bosnia, and Nigeria non-Muslim populations attach significantly less importance to
God in their lives and score significantly lower on religiosity scale than Muslims. Importance of
God scale is not available for India. In the remaining countries, t values do not reach 5 percent
significance level but the differences (for both indicators) are always in the predicted direction
(the one and only exception is Protestants in Uganda whose mean religiosity is very slightly
above Muslims).

In short, at this level of analysis too, it is confirmed that Muslims are more pious than others.
Could this finding be related to the higher proportions of fundamentalists and dogmatics found in
Islamic populations since unquestioned faith should be a prerequisite for fundamentalism?
Surely, there is no evidence of such a relationship in our analyses but the question seems worthy
of exploration.

23
 Table 6 MODERN ITY
Importance of God Religiosity Scale Secularism Scale Fatalism vs Free Choice Family Scale Women and Work Sex. Tolerance Scale
Mean Std. Dev. Mean Std. Dev. Mean St. Dev. Mean St. Dev. Mean St. Dev. Mean St. Dev. Mean St Dev.
ALBANIA
Muslim 3.38 2.66 2.53 2.49 5.41 1.99 5.87 2.14 0.72 0.87 6.60 1.03 6.73 3.34
Orthodox 3.98 2.75 3.47 2.33 6.07 2.05 5.99 2.44 0.78 0.98 6.72 0.95 7.14 3.29
BULGARIA
Muslim 3.59 2.83 1.88 2.18 5.06 2.16 5.64 2.76 0.54 0.75 6.32 1.17 8.71 5.48
Orthodox 5.03 2.80 3.68 2.45 6.27 2.05 6.21 2.57 1.44 1.21 6.50 1.33 10.41 6.07
MACEDONIA
Muslim 1.43 1.19 0.60 0.97 4.96 2.02 5.29 3.06 1.77 1.09 6.45 1.28 4.95 2.94
Orthodox 3.83 2.87 3.47 1.94 6.92 2.19 5.83 2.69 1.40 1.19 6.48 1.16 7.94 5.42
BOSNIA
Muslim 2.48 2.26 1.41 1.77 6.45 1.94 6.23 2.15 1.05 1.07 6.45 1.06 6.63 4.05
Orthodox 3.68 2.56 3.34 2.11 6.86 1.91 6.17 2.35 1.48 1.09 6.50 1.03 7.60 4.74
MONTENEGRO
Muslim 4.34 3.53 3.45 2.51 6.66 2.12 5.75 2.34 1.14 0.99 6.21 1.14 6.85 4.22
Orthodox 4.61 2.86 3.78 2.13 6.48 2.10 5.91 2.36 1.10 1.19 6.36 1.03 6.80 4.60
NIGERIA
Muslim 1.28 1.07 0.47 0.68 3.04 1.35 6.84 2.53 0.49 0.76 6.39 1.29 4.37 3.41
Protestant 1.42 1.26 0.60 0.73 3.42 1.65 7.17 2.35 0.61 0.83 6.64 1.16 4.95 4.12
TANZANIA
Muslim 1.38 1.50 0.54 0.89 4.32 2.29 5.86 3.50 0.70 1.06 6.39 1.29 3.71 2.69
Roman Catholic 1.41 1.62 0.59 0.98 4.60 2.39 5.54 3.58 0.85 1.09 6.66 1.23 3.52 1.99
Protestant 1.22 0.96 0.47 0.75 4.54 2.39 6.15 3.63 0.63 0.79 6.56 1.22 3.75 2.42
UGANDA
Muslim 1.50 1.33 0.76 1.06 4.28 1.71 6.80 2.43 1.24 1.12 5.89 0.95 5.02 3.25
Roman Catholic 1.50 1.32 0.80 1.02 4.54 1.90 6.70 2.38 1.49 1.26 6.01 1.16 5.21 3.82
Protestant 1.72 1.58 0.72 0.99 4.81 1.82 6.85 2.64 1.42 1.17 6.02 1.12 4.66 3.37
INDIA
Muslim na na 1.52 1.59 6.75 2.45 4.77 2.69 0.94 1.25 5.64 1.20 5.78 5.36
Hindu na na 1.68 1.80 6.37 2.16 5.66 2.89 0.96 1.08 5.93 1.32 9.81 9.31
Secularism. The separation of religion and the political/legal system has always been
problematic in Islam. The Holy Quran is not restricted to matters of faith but is also a
constitution, a legal code and a civil law for a society of believers. At least, this was the original
intention although changes in economic and political conditions as well as advances in science
and technology have always forced modifications and relaxations. It should also be noted that
Islam is the only major religion with the Prophet serving as the commander-in-chief and the head
of state as well. All these attributes have made it particularly difficult for Islamic societies “to
give unto Cesar what belong to him.” Is the legacy still alive? Our eight-country analysis
indicates that there are significant differences between Muslims and non-Muslims living in the
same society and that the former are likely to be less secular. The differences (all in the same
direction) are significant at the 5 percent level or better in Albania, Bulgaria, Macedonia, Bosnia
and Nigeria. In Uganda, the differences do not attain statistical significance although, there too,
Muslims are, on the average, less secular than both Catholics and Protestants. In Montenegro, the
Orthodox seem slightly less secular than Muslims and the same holds true for India (in both
cases, the differences are not significant). Of course, secularism has a very different meaning in
the context of Hinduism as compared to the major monotheistic religions.

Efficacy. Fate is an important concept in all monotheist religions but perhaps even more so in
Islam. The slave in Diderot’s Jacques le Fatalist keeps reminding to his master that “it is written
yonder.” A good Muslim is required to believe that all are “written yonder.” The individual is,
then, powerless against fate. In our set of eight countries, Muslims in Macedonia, Nigeria and
India are significantly more fatalistic than non-Muslims. The difference is nearly significant
(significant if equal variances are assumed) in Bulgaria. In short, although certainly not as
pronounced as some other dimensions, there is a tendency for Muslims to believe they have less
control over their lives.

24
Marriage and family. Unlike some other variables, there is no reason to believe that Islam
emphasizes family values more than other religions. Divorce has been possible and relatively
easy in Islam from the very beginning. Adultery, especially for women, was punishable by death
but it is one of the Ten Commandments after all. Islam’s emphasis on having children is
certainly not greater than that of the Catholic Church and Muslims have some more flexibility in
birth control and abortion. Nevertheless, Muslims score significantly more conservatively (i.e.
more emphasis on marriage and family) on family values scale in Bulgaria, Bosnia, and Nigeria.
It is interesting to note that the difference between Muslims and Catholics in Uganda is also
significant (for Protestants it is nearly significant). However, in Macedonia, the Orthodox
population is found to possess more conservative family values compared to Muslims and the
difference is significant. The findings about family and marriage values are not conclusive.

Female employment. The differences with respect to working women are significant only in
Nigeria and India. In both cases, Muslims are less favorable to female employment. In the
remaining six cases, without exception, Muslims have more conservative views on the subject but
these do not attain statistical significance.

Sexual tolerance. As we already know, for distinguishing Islamic civilization from others, sexual
tolerance is the criterion to watch for. The last column of comparisons in Table 6 in general
conforms to our expectations. In five of our eight countries (Bulgaria, Macedonia, Bosnia,
Nigeria and India) Muslims are significantly less tolerant then others (Orthodox, Protestant, and
Hindu). In Albania, we observe the same difference but the significance level does not reach 5
percent. In Montenegro, the means for the sexual tolerance scale is roughly equal in the two
groups (Islamic and Orthodox). In Uganda, Catholics are more and Protestants less tolerant than
Muslims but neither difference is significant. We conclude that, even within the same society,
Muslims are less tolerant in general on matters pertaining to “the eros.”

d. Democratic Culture
Of the 45 comparisons reported in Table 7 on democratic values, only 15 are statistically
significant. But more importantly, the direction of the differences are not predictable and display
a somewhat erratic pattern. From these findings, it is certainly not possible –even in very general

25
terms- to say that Islamic culture is either less or more supportive of a democratic system of
government. Thus, we find strong support for the conclusion that “the basic cultural fault line
between the West and Islam does not concern democracy.” (Inglehart and Norris, 2002).

(Table 7 about here)

Our fourth and the last hypothesis predicted that in each country variance within a civilization
should be smaller than total variance and also variance between civilizations. Once again
ANOVA proves inappropriate to test this hypothesis because of highly inflated significance
levels resulting from large sample sizes. Due to space limitations, it is not possible to examine
the variances for all of our eight countries. Table 8 reports the within groups (Protestant and
Muslim) and total standard deviations for Nigeria. Clearly, within a single country we cannot
expect to find the dramatic differences observed at the global level. Nevertheless, the results in
Table 8 are interesting. For the following variables, within Muslim standard deviations are
smaller than total (including all religious groups in the country) and within Protestant standard
deviations:
Determination, perseverance in children
Importance of God
Religiosity scale
Secularism scale
Marriage and family scale
Sexual tolerance scale
Neighbor tolerance scale
Every single variable on this list, with the sole exception of the neighbor tolerance scale,
conforms to expectations. I do not contend that results from a single country can be generalized
but it is highly unlikely that the Nigerian results are just a coincidence. Islamic populations do
have certain cultural characteristics.

(Table 8 about here)

26
Conclusion
After the events of September 11, Huntington’s thesis on the clash of civilizations gained
exceptional prominence. It has been lauded and attacked with equal fervor but almost always
without empirical evidence. A complete formalization and testing of the theory is still to come
but a recent article by Inglehart and Norris (2002) is the first time Huntington’s theory is
subjected to the test of data. Their conclusion is that the real difference between the Islamic
world and the West has nothing to do with political values but rather with attitudes towards
women and sex.

This paper starts out with four testable hypotheses derived from Huntington’s thesis. It is noted
that the first hypothesis is tested by Inglehart and Norris. A theoretical and comprehensive
scheme is developed to obtain a framework with which cultures can be compared. The scheme
has four basic dimensions: Protestant Ethic, Social Capital, Modernity and Democratic Culture.
Huntington argued that, above all, it would be the last dimension on which Islamic societies
would score miserably.

An analysis of the variances within civilizations at the global level confirms Inglehart and
Norris’s thesis: if we try to classify civilizations according to political values we are not going to
make much progress: the data tell us to look elsewhere.

With these clues in mind, we compared Islamic and non-Islamic populations in eight countries for
which data are available. Global level correlations cannot be expected to be repeated at the
national level. These are multi-religion societies and the nation-state is a powerful socializing
agent. It is natural that some of the cultural differences that have their foundations in religion are
ironed out to a great extent through schooling, mass media, etc. Nevertheless, by and large, our
expectations were confirmed even at the level of the nation. Some cultural differences do
continue to exist between Islamic populations and others within the boundaries of the same state.

27
The list of characteristics that would define the Islamic culture is very consistent. First and
foremost, for Muslims their faith is important –more so, it seems, than those who belong to other
religions. Our analyses provided strong evidence of this.

Another very significant characteristic of Islamic culture is its outlook on women and sex.
Islamic values are less supportive of gender equality and less tolerant of sexual liberalization.
This finding was repeated enough times that we would be justified to regard it as more than
tentative.

Of the three indicators of Protestant Ethic that were employed, two (hard work and thrift) turned
out to be insignificant. However, determination and perseverance consistently appeared as
valueS to which Muslims do not attach much importance.

Political culture and more specifically democratic culture variables failed to distinguish Islam
from the Protestant, Catholic, Orthodox or Hindu worlds. These cultures are not related in any
significant way to political tolerance, support of freedom, participation or the search for
alternatives to the democratic system.

To sum up in one sentence: if one wishes to define Islamic civilization in the contemporary
world, religiosity, sexual tolerance, perseverance and determination, and support of gender
equality are the key concepts to look for. If these make up the yardstick, the answer to the title of
this article then is: “yes, there is an Islamic civilization.” But why is it inevitable that this should
lead to a clash? That question will have to be tackled elsewhere.

REFERENCES

28
Almond, G. and Verba, S. 1963. The Civic Culture. Princeton: Princeton University
Press.
Bozeman, A. B. 1975. ‘Civilizations under Stress.’ Virginia Quarterly Review. 51(winter):
1-18.
Coleman, J. S. 1990. Foundations of Social Theory. Cambridge: Harvard University Press.
Dasgupta, P. and Serageldin, I. eds. 2000. Social Capital: A Multifaceted Perspective.
Washington, D.C. The World Bank.
Fukuyama, F. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York:
Free Press.
Halman, L. and Pettersson, T. 1998. “Religion and Social Capital Revisited.” Paper presented
At the WVS Workshop. Stockholm. June 11-14, 1998.
Halman, L. and Pettersson, T. 2002. “A Modified Model for the Cross-Cultural Measurement
of Two Basic Value Orientations.” Paper presented at the ISA Conference.
Brisbane, Australia. July 8, 2002.
Harris, M. 1974. Cows, Pigs, Wars and Witches: The Riddles of Culture. New York:
Vintage Books.
Huntington, S. P. 1993. ‘The Clash of Civilizations?’ Foreign Affairs. 75(6): 28-34.
Huntington, S. P. 1998. The Clash of Civilizations and the Remaking of the World Order.
London: Touchstone Books.
Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton
University Press.
Inglehart, R. 1997. Modernization and Postmodernization: Cultural, Economic and Political
Change in 43 Societies. Princeton: Princeton University Press.
Inglehart, R. and Baker, W. E. 2000. ‘Modernization, Cultural Change and the Persistence
of Traditional Values.’ American Sociological Review. 65(February): 19-51.
Inkeles, A. and Smith, D. H. 1974. Becoming Modern: Individual Change in Six Developing
Countries. Campbridge: Harvard University Press.
Klingemann, H. D. 1999. “Mapping Political Support in the 1990’s: A Global Analysis” in
Norris, P. ed. Critical Citizens: Global Support for Democratic Governance. Oxford:
Oxford University Press.
Misztal, B. A. 1996. Trust in Modern Societies. Cambridge: The Polity Press.

29
Norris, P. and Inglehart, R. 2002 ‘Islam and the West: Testing the Clash of Civilizations
Thesis.’ John F. Kennedy School of Government, Harvard University, Faculty
Research Working Papers Series.
Putnam, R. D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton:
Princeton University Press.
Rotberg, R. I. ed. 2001. Patterns of Social Capital: Stability and Change in Historical
Perspective. Cambridge: Cambridge University Press.
Taylor, C. 2001. “Two Theories of Modernity.” in Gaonkar, P.D. ed. Alternative
Modernities. Durham and London: Duke University Press.
Warren, M. E. ed. 1999. Democracy and Trust. Cambridge: Cambridge University Press.
Weber, M. 1958. The Protestant Ethic and the Spirit of Capitalism. trans. Talcott Parsons.
New York: Charles Scribner’s Sons.

30
 Table 1. Dimensions for Comparing Cultural Values

1. Protestant Ethic
a. Hard work
b. Thrift
c. Determination and perseverance

2. Social Capital
a. Interpersonal trust
b. Respect for others
c. Petitioning

3. Modernity
a. Religiosity
i. Importance of God
ii. Degee of religiosity
b. Secularism
i. Nonbelievers in public office
ii. Strong believer in public office
c. Efficacy
Fate versus Free Choice
d. Marriage and family
i. Unconditional love for parents
ii. Unconditional parental sacrifice for children
iii. Need for both parents
iv. Female need for children
v. Outmodedness of marriage
vi. Motherhood out of wedlock
vii.Emphasis on family life
e. Female employment
i. Working mother and children relations
ii. Contribution to household income
f. Sexual tolerance
i. Justifiability of homosexuality
ii. Justifiability of prostitution (or casual relationships)
iii. Justifiability of abortion
4. Democratic culture
a. Tolerance
b. Support of freedom
c. Participation
i. Interest in politics
ii. Joining boycotts
iii.Attending demonstrations
d. Problems of democracy
i. Democracy and economic performance
ii. Democracy and indecisiveness
iii.Democracy and the maintenance of order

31
 Table 2
Countries and Civilizationsa
Protestant Catholic Islamic Orthodox Central Europe Latin
America
Canada Austria Albania Bosnia Crotia Argentina
Denmark Belgium Algeria Belarus Czechia Chile
Finland France Egypt Bulgaria Estonia Peru
Germany Ireland Indonesia Greece Hungary Puerto Rico
Iceland Italy Iran Macedonia Latvia Venezuella
N. Ireland Luxembourg Jordan Moldova Lithuania
Sweden Malta Morocco Montenegro Slovakia
U.K. Netherlands Pakistan Romania Slovenia
U.S.A Poland Turkey Russia
Philippines Serbia
Portugal Ukraine
Spain

Sino Confucian SubSaharan Hindu Judaic Japanese

Africa
China Nigeria India Israel Japan
S. Korea S. Africa
Vietnam Tanzania
Uganda
Zimbabwe
___________________
a
Following the classifications of Huntington (1998) and Inglehart and Norris (2002).

32
 Table 3
Consistency of Values within Civilizations
(standard deviationsa)

Protestant Catholic Islamic Orthodox Total

Protestant Ethics
Hard work 0.46 0.50 0.49 0.45 0.50
Thrift 0.45 0.49 0.47 0.49 0.48
Determination 0.49 0.47 0.44 0.49 0.48
Social Capital
Interpersonal trust 0.50 0.45 0.46 0.42 0.47
Respect for others 0.39 0.43 0.48 0.48 0.45
Signing petition 0.47 0.50 0.33 0.41 0.49
Modernization
Religiosity
Importance of God 3.29 3.13 1.59 3.10 3.22
Religiosity scale 2.52 2.36 1.28 2.45 2.47
Secularism
Secularism scale 2.00 2.23 1.95 2.17 2.35
Efficacy
Fatalism/free choice 1.86 2.15 2.73 2.51 2.37
Marriage and Family
Family scale 1.43 1.45 0.90 1.20 1.44
Female employment
Women and work scale 1.25 1.27 1.43 1.15 1.29
Sexual tolerance
Sexual tolerance scale 7.69 7.34 2.67 6.00 7.34
Democratic Culture
Tolerance
Neighbor scale 1.49 1.78 1.64 1.79 1.81
Participation
Participation scale 1.58 1.61 1.43 1.47 1.59
Problems with Democracy
Problems scale 1.77 1.97 2.01 2.14 2.00
Alternatives to Democracy
Alternatives scale 1.81 1.97 1.95 2.02 2.04
____________________________
a
The sample is weighted so that the sample size in each country is fixed at 1,000.

33
 TABLE 4
PROTESTANT ETHICS
Hard Work Thrift Determination
Mean Std. Dev. Mean Std. Dev Mean St. Dev.
ALBANIA
Muslim 0.66 0.47 0.58 0.49 0.52 0.50
Orthodox 0.63 0.48 0.54 0.50 0.66 0.48
BULGARIA
Muslim 0.90 0.30 0.37 0.49 0.49 0.50
Orthodox 0.85 0.35 0.42 0.49 0.55 0.50
MACEDONIA
Muslim 0.71 0.45 0.33 0.47 0.17 0.38
Orthodox 0.27 0.44 0.40 0.49 0.51 0.50
BOSNIA
Muslim 0.56 0.50 0.43 0.50 0.42 0.49
Orthodox 0.53 0.50 0.34 0.47 0.50 0.50
MONTENEGRO
Muslim 0.82 0.39 0.50 0.50 0.34 0.47
Orthodox 0.83 0.38 0.34 0.47 0.48 0.50
NIGERIA
Muslim 0.78 0.42 0.11 0.31 0.19 0.39
Protestant 0.81 0.39 0.10 0.30 0.25 0.43
TANZANIA
Muslim 0.84 0.36 0.56 0.50 0.53 0.50
Roman Catholic 0.83 0.38 0.50 0.50 0.60 0.49
Protestant 0.84 0.37 0.52 0.50 0.58 0.49
UGANDA
Muslim 0.84 0.36 0.10 0.30 0.32 0.47
Roman Catholic 0.87 0.33 0.09 0.29 0.35 0.48
Protestant 0.85 0.36 0.12 0.33 0.36 0.48
INDIA
Muslim 0.86 0.35 0.64 0.48 0.36 0.48
Hindu 0.87 0.34 0.64 0.48 0.49 0.50

34
 TABLE 5
SOCIAL CAPITAL
Interpersonal Trust Respect for Others Signing a Petition
Mean Std. Dev. Mean Std. Dev. Mean St. Dev.
ALBANIA
Muslim 0.24 0.43 0.76 0.43 0.21 0.41
Orthodox 0.23 0.42 0.86 0.35 0.17 0.37
BULGARIA
Muslim 0.36 0.48 0.58 0.50 0.12 0.32
Orthodox 0.25 0.44 0.58 0.49 0.12 0.33
MACEDONIA
Muslim 0.19 0.39 0.77 0.42 0.22 0.42
Orthodox 0.12 0.32 0.74 0.44 0.28 0.45
BOSNIA
Muslim 0.14 0.35 0.74 0.44 0.24 0.43
Orthodox 0.15 0.36 0.65 0.48 0.13 0.33

MONTENEGRO
Muslim 0.37 0.48 0.63 0.48 0.14 0.35
Orthodox 0.33 0.47 0.55 0.50 0.27 0.45
NIGERIA
Muslim 0.28 0.45 0.55 0.50 0.07 0.25
Protestant 0.24 0.43 0.61 0.49 0.07 0.25
TANZANIA
Muslim 0.09 0.29 0.85 0.36 0.13 0.33
Roman Catholic 0.08 0.28 0.85 0.36 0.07 0.25
Protestant 0.03 0.18 0.80 0.40 0.10 0.30
UGANDA
Muslim 0.07 0.25 0.46 0.50 0.20 0.40
Roman Catholic 0.06 0.25 0.61 0.49 0.15 0.36
Protestant 0.09 0.29 0.56 0.50 0.26 0.44
INDIA
Muslim 0.49 0.50 0.53 0.50 0.36 0.48
Hindu 0.42 0.49 0.64 0.48 0.28 0.45

35
 TABLE 7
DEMOCRATIC CULTURE
Tolerance Support for Freedoms Political Participation Problems of Democracy Alternatives to Democracy
Neighbor Tol. Scale Choosing Free Speech Participation Scale Tolerating Prob. Scale Pref. for Alternatives Scale
Mean Std. Dev. Mean Std. Dev. Mean St. Dev. Mean St. Dev. Mean St. Dev
ALBANIA
Muslim 3.73 1.65 0.25 0.43 4.96 1.50 8.95 2.03 8.63 1.49
Orthodox 3.83 1.59 0.27 0.45 4.86 1.33 9.09 2.05 8.82 1.46
BULGARIA
Muslim 4.20 1.86 0.29 0.46 4.68 1.54 8.08 2.22 7.85 1.90
Orthodox 4.36 1.93 0.24 0.43 4.91 1.43 7.92 2.18 8.07 1.73
MACEDONIA
Muslim 4.67 0.93 0.41 0.49 4.97 1.56 8.35 2.13 6.46 1.77
Orthodox 4.41 1.90 0.21 0.41 5.07 1.55 7.41 2.27 6.99 2.14
BOSNIA
Muslim 4.22 1.42 0.22 0.42 4.60 1.31 8.61 1.96 8.18 1.77
Orthodox 3.87 2.11 0.28 0.45 4.71 1.41 8.07 1.93 8.31 1.49
MONTENEGRO
Muslim 3.66 1.21 0.21 0.41 5.09 1.43 9.06 1.71 9.43 1.17
Orthodox 3.59 1.35 0.25 0.43 5.46 1.69 8.85 2.19 8.61 1.55
NIGERIA
Muslim 2.37 1.06 0.31 0.46 4.97 1.57 6.98 2.17 7.45 2.26
Protestant 2.46 0.96 0.33 0.47 4.82 1.54 7.92 2.09 8.01 2.02
TANZANIA
Muslim 4.23 1.08 0.17 0.38 5.54 1.15 8.31 2.64 9.98 1.81
Roman Catholic 4.27 1.19 0.15 0.36 5.47 1.14 8.49 2.24 10.22 1.68
Protestant 4.37 1.00 0.13 0.34 5.27 1.07 8.97 2.28 10.39 1.47
UGANDA
Muslim 4.62 1.30 0.50 0.50 5.01 1.58 8.35 1.94 9.03 2.46
Roman Catholic 4.93 1.25 0.41 0.49 4.88 1.66 8.74 1.88 9.28 2.34
Protestant 4.82 1.27 0.46 0.50 5.16 1.66 8.61 1.94 8.95 2.31
INDIA
Muslim 2.91 1.05 0.16 0.37 5.03 1.87 8.15 2.05 8.21 1.93
Hindu 2.91 1.32 0.18 0.38 4.83 1.77 7.49 1.85 7.44 1.95

36
 Table 8
Consistency of Values: Results for Nigeria
(standard deviations)

Protestant Muslim Total

Protestant Ethic
Hard work 0.39 0.42 0.40
Thrift 0.30 0.31 0.30
Determination 0.43 0.39 0.42
Social Capital
Interpersonal trust 0.43 0.45 0.44
Respect for others 0.49 0.50 0.49
Signing petition 0.25 0.25 0.25
Modernization
Religiosity
Importance of God 1.26 1.07 1.20
Religiosity scale 0.73 0.68 0.75
Secularism
Secularism scale 1.65 1.35 1.59
Efficacy
Fatalism/free choice 2.35 2.53 2.41
Marriage and Family
Family scale 0.83 0.76 0.81
Female employment
Women and work scale 1.16 1.28 1.21
Sexual tolerance
Sexual tolerance scale 4.12 3.41 3.90
Democratic Culture
Tolerance
Neighbor scale 0.96 1.06 1.52
Freedom
Freedom scale 0.47 0.47 0.47
Participation
Participation scale 1.54 1.57 1.56
Problems with Democracy
Problems scale 2.09 2.17 2.16
Alternatives to Democracy
Alternatives scale 2.02 2.26 2.12

37