Within the European Values Study (EVS) a discussion has been going on for decades on the strategic question of whether it is better to use a grand theory to make sense of the findings of its cross-national surveys or whether the researchers should instead use partial or middle-range theories. This chapter attempts to make an assessment of this debate. In the 1990’s several EVS-researchers opted for a rather simple and parsimonious theoretical model to explain cross-national value differences and value changes. This model they derived from modernization theory. However, after some initial success the model proved to be too simple. In the following decades it made place for a much more complex one, although modernization theory remained to be the hard core of the model. Some value researchers persevered despite half-hearted successes. Others dropped out and admitted themselves to the camp of the middle-range theories enthusiasts. This raises the additional question of what the virtues and vices of grand theories and middle-range theories are. This question too is examined in this chapter. Finally, an alternative grand theory is proposed as a possible viable alternative, cognitivist rational choice theory.
In the early days of the European Values Study (EVS), from its founding in 1978 till sometime after the first cross-national survey in 1981, its goals were highly descriptive and of an applied nature. However, pretty soon it appeared to the researchers involved that several important things were lacking. It turned out that to give direction to the choice of items to be included in the questionnaire explicit social-scientific research questions were urgently needed. What was also missing were theoretical notions that could help to interpret or even explain the empirical findings. It is on these theoretical notions that this chapter is dedicated.
In 1984, Loek Halman made an appearance in the EVS community. At first as the data analyst of the Dutch group, but soon he became the indispensable secretary of EVS’ steering committee. After he finished his PhD thesis in 1990, his position within the project became increasingly a pivotal one. He was the key figure in designing the questionnaire, in keeping everybody informed, in data analyses, and in publishing books and book series. In the corridors of EVS conferences and workshops, he was affectionally called Mr. European Values. As such he had a dream: to have a grand theory that could explain all or at least most of the outcomes of the different waves of EVS. At first, his favorite grand theory, modernization theory, seemed to be rather successful. However, soon the odds turned.
Halman, however, did not want to give up his dream so easily and persevered. At this point, I have to admit my complicity. Somewhere in the second half of the 1990’s, I became chair of EVS’ Theory Group. He and I have several times cooperated in an endeavor to reconstruct and test more sophisticated versions of modernization theory in order to make his dream come true. Now Halman is retiring it seems to me high time to take up stock of our endeavors. Did we succeed? If not, should EVS then satisfy itself with only middle range theories? Or is there perhaps an alternative grand theory available that can successfully replace modernization theory?
In Halman’s earliest publications (Halman et al., 1987; Halman, 1987, 1991), modernization theory already appeared on the scene. It seemed to him an ideal grand theory for EVS because it provided insights regarding the transformation of traditional societies into modern ones. Not only as far as structural changes such as industrialization and market formation and expansion are concerned, but also or particularly with regard to its effects on cultural phenomena, i.e., a tendency that traditional and religious norms and values are replaced by more secular, instrumental and individualized ones. In a chapter in Halman et al. (1987), he and his co-authors Dorenbosch and Heuks (Dorenbosch et al., 1987) reconstructed modernization theory and empirically tested hypotheses derived from it. Without mentioning Ockham’s razor, they applied its law of parsimony that states that the simplest explanation is usually the right one. Their first hypothesis was apparently as a precaution formulated in probabilistic terms: ‘Structural modernization will be positively correlated with cultural modernization’. However, they realized that this was a too simple hypothesis. To refine it, they used some auxiliary theories, more specifically insights from cultural lag theory and social diffusion theory. The second hypothesis stated that structural modernization is after some delay followed by cultural modernization in such a way that it starts in the industrially and technologically most advanced centre of the modernizing world and then spreads with still more delay to its less developed periphery. Both hypotheses found ample support in the macro-level analysis of the data of the 1981 wave of EVS. Nevertheless, in a chapter in the same volume Halman (1987) seemed to realize that not only the explanatory model but also the empirical test they used was a too crude one because it based itself only on central tendencies in the country samples. Therefore, he also looked at the variation at the individual level. He found that, at least in the Netherlands, individuals too could be distinguished according to their degree of modernity.
In his PhD thesis, Halman (1991) addressed the question of whether and, if so, to what degree in the Western world a compartmentalization of values had taken place. In the meantime, the survey had also been fielded in the USA and Canada. From the structural-functionalist version of modernization theory, he derived the hypothesis that in countries in which the structural modernization of society has progressed further, people’s values in the different domains are less interconnected than in countries in which the process has not gone so far yet. He tested whether this hypothesis made empirical sense. He concluded that the result was at best mixed. As expected, in advanced modern countries, religious values appeared to be less important for the values on other life domains than in less advanced ones. However, this does not mean that a country being structurally more modern implies that the interconnectedness of value domains is lower there than in less structurally modern countries. Halman also concluded that there are big differences in value preferences in the different countries. It is not a simple task to explain these cultural differences by referring only to their degree of modernization, he argued (Halman, 1991).
In 1990, a second wave of the EVS survey was fielded. Now the possibility arose to subject hypotheses not only to cross-national tests but also to longitudinal ones. The Dutch team did exactly that, they analysed the dynamics of value change between 1981 and 1990 in the West and published their results in a book (Ester et. al., 1994). Once again, the grand theory used was modernization theory. In a chapter co-authored with Ruud de Moor (Halman & de Moor, 1994), one of the questions addressed was whether the populations of modernizing societies showed a shift from traditional towards individualized values. The authors concluded that modernization did not lead to a uniform replacement of traditional values by individualized ones in all domains of social life. In an epilogue to the book, his co-author Ruud de Moor (1994), one of the founding fathers of EVS and chair of the steering committee, concluded that most hypotheses derived from modernization theory and tested in their book were not supported by the data. In his opinion, instead of grand theories, like modernization theory, empirically founded partial theories were needed. Roma locuta causa finita, one might assume. Exit modernization theory?
The answer is in the negative. In the 1990’s, modernization theory made a remarkable comeback. Several social theorists such as Beck, Giddens and Inglehart argued that the relatively simple modernity of industrial societies had in the meantime been replaced by a different kind of modernity. This led to amendments to modernization theory. Arts and Halman (2002, 2004) used their suggestions as a heuristic devise for formulating a number of hypotheses pertaining to the supposed effect of late or post modernization on decreasing control over life, diminished interpersonal and institutional trust, and the rise of post-materialist values. Now institutionalism was used as an auxiliary theory, at least as far as the institutional arrangements of welfare regimes were concerned. We could profit not only of the data of the first two waves of EVS but also of the third wave of 2000. Post-materialism appeared not to be on the rise in the countries investigated, nor did the analyses unequivocally support the idea that differences in trust and control over life are based in differences in welfare state regimes.
In 2008, the fourth wave of EVS followed. Arts and Halman (2011, 2014) did one more attempt to save modernization theory from oblivion. The EVS-dataset now offered an opportunity to test Inglehart’s various amendments to modernization theory that had led to a much more complex and sophisticated version of this grand theory (e.g., Inglehart, 1977, 1990, 1997; Inglehart & Baker, 2000; Inglehart & Welzel, 2005). Inglehart suggested why there is a time lag between technological innovation and economic growth on the one hand and value changes on the other. He argued that value changes most of the time take place through intergenerational population replacement, i.e., younger birth cohorts replace older ones in the population. This is, by its very nature, a slow process. He assumed that people’s basic values are largely fixed when they reach adulthood, and remain more or less stable thereafter. He also assumed in the so-called socialization hypothesis, that people’s basic values to a large extent reflect the conditions that prevailed during their pre-adult years. From these two assumptions followed that intergenerational change will occur if younger generations grow up under different conditions from those that shaped earlier generations. Another assumption Inglehart made was that not only long-term developments such as technological innovation and economic growth, but also short-term changes, such as different phases of the business cycle, and short-term events, such as wars and revolutions, have an impact on people’s values. This assumption is connected with the so-called scarcity hypothesis, which states that people tend to attach the greatest value to the most pressing needs of the moment. These hypotheses can be tested by looking for age, period, and cohort effects. Inglehart also argued that value patterns are the products of not only modernization processes but also of country-specific patterns of the past, in other word, of cultural traditions. Historical value patterns are therefore interwoven with modern and post-modern ones. Thus, not only do technology and economy matter, but history does as well. Why cultural traditions are persistent, is explained by the theoretical notion of path dependence, which is the idea that cultural traditions create forces to sustain themselves even though the circumstances that gave rise to and reinforced them in the past may now no longer be relevant.
If we look at a more sophisticated version of Ockham’s razor that says that explanations should be not only as simple as possible, but also as complex as necessary, we have to conclude that the simplicity of the original modernization theoretical explanation of cross-national value differences and value changes had made place for a much more complex one. The pressing question that could be asked was for us whether modernization theory was still a progressive research program or sooner had become a degenerating one. Had it not become top-heavy because of the introduction of all kinds of auxiliary assumptions and hypotheses?
In 2019, Halman wrote a state-of-the-art article that looked like his EVS swan song (Halman & Gelissen, 2019). At last, he seemed to have woken up from his beautiful dream about grand theory. The conclusion is hard and straightforward. Modernization theory falls short when it comes to explaining the often considerable differences in value orientations between populations in various countries. There is more needed than economic growth and technological innovations to explain these differences. Institutions, culture, history, policies, all appear to affect people’s values. Nevertheless, context is not enough. It is essential to include individual-level characteristics, at least as controls. Quite often, individual attributes appear differently distributed in different countries, which may be the main reason why differences in value orientations between countries remain. Multi-level analysis is the appropriate tool for separating such composition effects from true contextual effects and multi-level theories are needed to explain what is going on with regard to cross-country value differences and value changes.
When Ruud de Moor (1994) concluded that modernization theory was far too general to explain the findings of EVS, he avoided the term middle-range theories. Why this was the case is not entirely clear. Perhaps to prevent that he would get bogged down in a long, drawn-out debate in sociology about the concept of theory. The discussion about grand versus middle-range theory started at the annual meeting of the American Sociological Society in 1947 dedicated to a stocktaking of the discipline. As chairman of the section on theory, Talcott Parsons delivered a paper on The Position of Sociological Theory and Robert Merton was his discussant. These papers were published in the next year (Parsons, 1948; Merton, 1948). Parsons returned to this discussion in his presidential address on The Prospects of Sociological Theory to the annual meeting of the same society in 1949 and in the meantime Merton (1949) had elaborated on his short discussion paper. Parsons’ address was published in the following year (Parsons, 1950). He argued that we need theory among other things to make in empirical research a selection among the enormous number of possible variables. Whereas Parsons emphasized the importance of high levels of generality in constructing sociological theories, Merton defended theories of the middle range: theories that lie between working hypotheses and a unified theory that tries to explain all the observed uniformities of social life. Twenty years later, Merton (1968) looked back at how the discussion proceeded. He noted that the responses to his plea were polarized. Many empirical sociologists gave assent to a middle-range theoretical strategy because this was what they already practiced. Many theorists, however, found it a retreat from properly high aspirations. The third response was a combination of the two others. An emphasis on middle-range theories does not mean exclusive attention to this kind of theorizing. Instead, it sees the development of more comprehensive theory as coming about through consolidations of middle-range theories rather than emerging, all at once, from the work of individual theorists on the grand scale.
If we look at the many publications generated by EVS and its daughter, the World Values Survey, it becomes clear that the overwhelming majority of them falls within the limits of the second and third response noticed by Merton.
Loek Halman and I belonged to the few who persevered and tried for a long time to save grand theory within EVS from oblivion following the methodological guideline that one must treat budding theoretical research programs leniently. It may take decades before they get off the ground and become empirically progressive. However, later on, we gave up and resigned in our defeat (Arts, 2011; Halman & Gelissen, 2019).
One could wonder whether this was too soon. To answer this question, we should perhaps look at the work of Raymond Boudon. Looking back at the debate about grand theory versus middle-range theories, Boudon (1991) concluded that what sociologists mean by the term ‘theory’ is not always clear. On the one hand you have ‘broad theory’ that tries to determine the overarching independent variable that would operate in all social processes, or to determine the essential feature of social structure, or to find out the two, three, or four concepts that would be sufficient to analyze all social phenomena. This is according to the advocates of middle-range theory a hopeless and even quixotic enterprise. On the other hand, you have sociologists who defend middle-range theories. This does not refer to a specific kind of theory, but is rather an approach to theory construction. Sociological theories, like all scientific theories, should aim to consolidate otherwise segregated hypotheses and empirical regularities. They should help explain puzzling phenomena and create new solid knowledge about the aspects of the world it is traditionally concerned with. One could argue that modernization theory is a mixed form of both types of sociological theory, partly bad partly good theorizing.
More important, however, is that Boudon refers positively to a middle-range theory that he himself developed and which he originally called subjective rationality theory and later cognitivist rational choice theory (Boudon, 1989, 1996, 1998, 2003, 2009). He elaborated this theory choosing the notion of subjective or bounded rationality that Herbert Simon coined, as a starting point. In simple situations, individuals often act according to what rational choice theory predicts: they maximize utility. Confronted with ambiguous and complex situations, they, however, appeal to theoretical notions, heuristic devices and moral principles to master these situations. Rational choice theory is therefore a powerful theory when applied to some types of social phenomena, but it appears to be powerless when confronted with many other types. Boudon identified three of these latter types: 1) All human actions are dictated by descriptive or cognitive beliefs. Sometimes they are commonplace and need no further attention, but some other times they are non-commonplace. Then it is crucial to investigate and evaluate the beliefs upon which they rest to explain the actions involved. 2) Some human actions are based on non-consequential prescriptive or moral beliefs. These actions are not intended to generate some outcome, but to endorse a moral principle. 3) Some other actions cannot in any sensible way assume to be dictated by self-interest. Sociologists often find themselves confronted with this latter kind of phenomenon, since social actors are regularly called upon to evaluate situations in which they are not personally implicated.
Boudon (1995, 1999, 2001) published around the turn of the millennium several books in which he dealt with especially the second type of phenomena. He gave a general overview of philosophical and sociological theories concerned with the sense of values that people have and their origin. He made a distinction between culturalist theories on the one hand and naturalist ones on the other. The first group considers values as cultural features that are endorsed or rejected by people because they have been socialized to them. The second group assumes that our moral principles are innate, i.e., derived from human nature. Both groups of theories can and have been rightly criticized. The first group cannot explain why there are some values that nearly everybody seems to share and the second one cannot explain the variability of values over time and space. Boudon’s objective was to defend and illustrate that the universal and contextual sides of values cannot be theorized independently from one another and returned to his cognitivist rational choice theory. He started from Weber’s distinction between instrumental rationality (Zweckrationalität) and non-instrumental or axiological rationality (Wertrationalität). People do not so much endorse values because of considerations of self-interest, but sooner because they have strong reasons why they accept some values and not others. When individuals think that ‘x is good’, they have good reasons for thinking so and they think so because of these reasons. Nevertheless, strong reasons here and now are not necessarily strong reasons there and then. A strong system of reasons can become weak and the other way round. Therefore, Boudon’s theory is both contextual and historical.
The pivotal orienting statement at the core of Boudon’s theory goes as follows: If individuals subscribe to a value statement, then they have most of the time good, or strong, or acceptable reasons for believing in it. Reasons can therefore be seen as causes. There, however, are not only rational causes of an axiological or instrumental nature, but also irrational ones such as emotions and traditions. Explanations with irrational forces can be often legitimately replaced by explanations with reasons. Opp (2014) has argued that Boudon’s theory is strikingly simple and has the advantage that it contributes to the explanation of both micro and macro phenomena and that it is testable. According to him, there, however, are also disadvantages, such as that the explanatory power is very low. A selection or relevance criterion for the kind of reasons and irrational factors that are causes for the explananda is lacking. Another disadvantage Opp (2014) sees is that the empirical validity is questionable. The theory not only proposes to explain why people subscribe to normative beliefs, but also to descriptive beliefs, attitudes, preferences, and behavior. Is it possible to explain the wide range of phenomena that Boudon’s cognitive rational choice theory tries to explain with a single theory? For proponents of grand theory, the answer is in the affirmative. Perhaps this is the overarching theoretical system people such as Loek Halman have dreamt of, although proponents of middle-range theories will be skeptical towards the validity of such a claim.
Forty years ago, the first cross-national EVS-survey was fielded. Four more waves followed with an interval of nine years. Surprisingly enough, it turns out that after so long a lapse of time it is still not so easy to give an unambiguous answer to the question of whether the alliance between EVS and grand theory has been a fruitful one. Modernization theory, once the favourite grand theory within EVS, could not pass the empirical tests satisfactorily in spite of ever more complex sophisticated versions. These more complex and sophisticated versions did, however, lead to much more sophisticated data analyses. Although Loek and I (Arts, 2011; Halman & Gelissen, 2019) gave up on developing and testing new versions of modernization theory, others did not. Ronald Inglehart (2018), for example, recently presented the first full statement of his evolutionary modernization theory and the empirical findings that it generated.
In the meantime, the majority of EVS-researchers heeded de Moor’s call and decided to restrict themselves to use middle-range theories to explain and test more specific phenomena. They realized that not only modernization matters, but that history, culture, institutions and so on and so forth also do. Loek Halman did not only dream, but he also accomplished a lot by co-editing a great number of books in which their papers have been published.
So far so good, but the question remains whether there is an alternative grand theory available that can successfully replace modernization theory. The answer is yes. Boudon’s cognitivist rational choice theory could come handy in this respect, because it is based on methodological individualism, which means that the theory is supposed to contribute to the explanation of both micro and macro social phenomena. De Graaf (2008) drew our attention to this theory in his inaugural lecture. Following his lead, we (Arts & Halman, 2014) avowed that EVS researchers should put much more attention to values that nearly all Europeans seem to share instead of only focusing on differences within and between European countries in this respect. It seems worthwhile to give Boudon’s theory more attention within EVS, although I realize that neither Halman nor I will be the ones that will accomplish this feat. It is up to a new generation of EVS researchers. Nevertheless, they could profit from the work of several sociologists who have criticized and tried to elaborate on Boudon’s theory such as Hamlin (2004) and Opp (2014).
Arts, W.A. (2011). Explaining European value patterns: Problems and solutions. Studia UBB Sociologia, 56, 7-31.
Arts, W., & L. Halman. (2002). Risk and trust. Value change in the second age of modernity. In P. Chiemelski, T. Krause, & W. Wesolowski (Eds.), Kultura, Osobowość, Polityka (pp. 321-348).
Wydawnictwo Naukowe SCHOLAR.
Arts, W., & Halman, L. (2004). European value changes in the second age of modernity. W. Arts & L.
Halman (Eds.) In European Values at the Turn of the Millennium (pp. 25-53). Brill.
Arts, W., & L. Halman. (2011). Value research and transformation in Europe. In R. Polak (Ed.), Zukunft.
Werte. Europa. Die Europäische Wertestudie 1990-2010: Österreich im Vergleich. (pp. 79-99). Böhläu.
Arts, W., & L. Halman. (2014). Cross-national values in Europe today: Facts and explanations. In W. Arts & L. Halman (Eds.), Value Contrasts and Consensus in Present-Day Europe. Painting Europe’s moral landscapes (pp. 1-17). Brill.
Boudon, R. (1989). Subjective rationality and the explanation of social behavior. Rationality and Society, 1, 173-196.
Boudon, R. (1991). What middle-range theories are. Contemporary Sociology, 20, 519-522.
Boudon, R. (1995). Le juste et le vrai. Fayard/Seuil.
Boudon, R. (1996). The cognitivist model: A generalized ‘rational choice model’. Rationality and Society, 8, 123-150.
Boudon, R. (1998). Limitations of rational choice theory. American Journal of Sociology, 104, 817-828.
Boudon, R. (1999). Le sens des valeurs. Presses Universitaires de France.
Boudon, R. (2001). The Origins of Values. Philosophy and Sociology of Beliefs. Transaction Publishers.
Boudon, R. (2003). Beyond rational choice theory. Annual Review of Sociology, 29, 1-21.
Boudon, R. (2009). Rational choice theory. In B.S. Turner (Ed.), The New Blackwell Companion to Social Theory (pp. 179-195). Wiley-Blackwell.
Dorenbos, J., Halman, L., & Heunks, F., (1987). Modernisering in West-Europa: Een tweedeling? In L. Halman, F. Heunks, R. de Moor, & H. Zanders, Traditie, Secularisatie en Individualisering (pp. 234-254). Tilburg University Press.
Graaf, P. de. (2008). Waarden die ertoe doen: De European Values Study in de 21ste eeuw. Tilburg University Press.
Halman, L. (1987). Op zoek naar typen in de Nederlandse cultuur. In Halman, L., Heunks, F., de Moor R. & Zanders, H. (1987). Traditie, Secularisatie en Individualisering (pp. 220-233). Tilburg University Press.
Halman, L. (1991). Waarden in de Westerse Wereld. Tilburg University Press.
Halman. L. & de Moor. R. (1994). Value patterns and modernity. In Ester, P., Halman L., & de Moor, R. (Eds.), The Individualizing Society. Value Change in Europe and North-America (pp. 155-161). Tilburg University Press.
Halman, L., & Gelissen, J. (2019). Values in life domains in a cross-national perspective. Kölner Zeitschrift für Soziologie und Sozialpsychologie, 71 (Suppl. 1), 519-543.
Hamlin, D. (2004). Beyond Relativism. Raymond Boudon, cognitive rationality and critical realism. Routledge.
Inglehart, R. (1977). The Silent Revolution: Changing values and political styles among Western publics. Princeton University Press.
Inglehart, R. (1990). Culture Shift in Advanced Industrial Society. Princeton University Press.
Inglehart, R. (1997). Modernization and Postmodernization: Cultural, economic and political change in 43 societies. Princeton University Press.
Inglehart, R. (2018). Cultural Evolution. People’s motivations are changing, and reshaping the world. Cambridge University Press.
Inglehart, R., & Baker, W. (2000). Modernization, cultural change and the persistence of traditional values. American Sociological Review 65, 19-51.
Inglehart, R., & Welzel, C. (2005). Modernization, Cultural Change and Democracy. Cambridge University Press.
Merton, R.K. (1947). Discussion of Parsons’ ’The position of sociological theory’. American Sociological Review, 13, 164-168.
Merton, R.K. (1949). On sociological theories of the middle range. In Merton R.K., Social Theory and Social Structure, (pp. 448-459). Simon & Schuster-Free Press.
Merton, R.K. (1968). On sociological theories of the middle range. In R.K. Merton, Social Theory and Social Structure, Enlarged edition, (pp. 39-72). Free Press.
Moor, R. de. (1994). Epilogue. In Ester, P., Halman L., & de Moor R. (Eds.), The Individualizing Society. Value Change in Europe and North-America. Tilburg University Press.
Opp, K-D. (2014). The explanation of everything. A critical assessment of Raymond Boudon’s theory explaining descriptive and normative beliefs, attitudes, preferences and behavior. Revista de Sociologia, 99, 481-514.
Parsons, T. (1947). The position of sociological theory. American Sociological Review, 13, 156-171.
Parsons, T. (1949). The prospects of sociological theory. American Sociological Review, 15, 3-16.