SciELO - Scientific Electronic Library Online

vol.28 issue2Social classes and symbolization in the city of Porto: theoretical elements and empirical research resultsAn archaeological outlook on public health in Brazil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Tempo Social

Print version ISSN 0103-2070On-line version ISSN 1809-4554

Tempo soc. vol.28 no.2 São Paulo May./Aug. 2016 

Dossiê - Classes sociais e desigualdades: sociabilidade, cultura e política

Entrevista com John Scott

Edison Bertoncelo1 

Virgílio Borges Pereira2 

1Universidade de São Paulo, Brasil. E-mail:

2Universidade do Porto Portugal. E-mail:

John Scott é um sociólogo inglês com diversas publicações relevantes nas áreas de teoria sociológica, estratificação social, elites e análise de redes. Entre suas principais obras, estão Corporate business and capitalist class (1991); Social network analysis (1992); Stratification and power: structures of class, status and command (1996); Conceptualising the social world (2011). Atualmente, é professor visitante das universidades de Essex e de Exeter, e professor honorário da Universidade de Copenhague.

A entrevista feita com John Scott aborda alguns temas trazidos por seus estudos sobre a formação das classes sociais e das elites britânicas.

In your book Stratification and power and in many articles, you argue in favour of a multidimensional approach to social stratification. Could you tell us what such an approach means?

I have always followed Max Weber's emphasis on stratification as a multidimensional phenomenon. This interest dates from my earliest years as a sociologist when I found myself attracted by the work of both Karl Marx and Talcott Parsons. I read some early articles by David Lockwood that convinced me that the limitations of each author could be overcome - at least in part - by drawing on the ideas of the other. This was not seen as an eclectic combination of two startlingly different positions but as an analytical and critical synthesis. Lockwood illustrated this with some general remarks about stratification and this encouraged me to consider how Marx's view of stratification as rooted in property relations could be reconciled with Parsons's "functionalist" view of stratification as rooted in values.

I found an answer to this in Max Weber's fragment on "Class, status, and party", which he wrote as a way of spelling out the relationship between a Marxist view of "class" and his own recognition, from his studies in the sociology of religion, that the strata of a society were often shaped by their values. In my PhD thesis, completed in 1976, I set out this view and began to work towards a conceptualisation of status that drew on Parsons's work and would be as systematic as Marx's analysis of class.

In the work that I did in the following years I concentrated on economic structures and class relations, re-conceptualising the idea of a capitalist class (Corporations, classes, and capitalism, 1979, revised as Corporate business and capitalist classes, 1991). I also tried to examine how, in Britain, the capitalist class merged with an established aristocratic status group to form a single "upper class" (The upper classes, 1982).

The argument that I developed was not that class and status were completely separate forms of stratification, though some critics did read the work in this way. I recognised that any social stratum had both class (economic) and status (cultural) dimensions and that the two aspects of stratification are inseparable. However, I held that we could, nevertheless, recognise them as analytically independent dimensions - ideal types in Weber's terminology - and that the relative importance of each dimension in relation to social conditions and social consciousness could vary from one society to another.

Thus, under certain historical conditions, considerations of status may be especially important, and Weber showed that this was the case in classical India and China and that they could therefore be designated as "status societies". In Capitalism, however, he saw a society in which the market and property relations had become of overriding significance and status considerations were limited to what Marxists saw as "ideological" expressions of the class relations. Capitalism, therefore, could be designated as a "class society".

I wrote my book Stratification and power (1996) to draw out the general framework of these reflections. I also recognised that Weber's own paper was incomplete and that it was necessary to extend it into a larger three-dimensional model of stratification. Many other writers had recognised this but had - incorrectly in my view - seen the third dimension as either "party" or "power". In fact, Weber saw "party" as referring to the organisation of interest groups and social movements on the basis of class and status relations. He also argued that class and status are both to be seen as aspects of a wider distribution of power, and I tried to show that the explorations of power in his writings on bureaucracy and the state were the basis on which we could recognise the third dimension as the authority and command relations through which elites and non-elites are formed. A comprehensive view of stratification, in any society, has to take account of all three dimensions.

This was the limit to my multidimensionalism. I do not go along with the views of those who argue that ethnicity, gender, age, religion, and other factors are additional bases of stratification. They are clearly important aspects of inequality and oppression, but they often operate in and through class and status relations. They are not independent bases of social stratification, understood as the formation of social strata. This was exactly what Weber recognised for ethnicity and religion in classical India, and for ethnicity, religion, and age in classical China. It is also something recognised in many studiers of women's oppression that see gender relations as a major influence on class-based inequalities through the power of a patriarchal ideology. Weber himself saw age and gender as fundamental features of status and authority in patriarchal societies.

What do you consider as promising ways of operationalizing such an approach in empirical research

There are very different problems involved in operationalizing the various aspects of the multidimensional approach.

It is the operationalization of class situation that has received the most attention in conceptual and empirical work. I can best illustrate this from discussions in Britain, but similar issues have been explored in many other countries. It is in the work of John Goldthorpe and his colleagues that important advances in the operationalization of employment relations has taken place, largely as a way of devising occupational categories for use in studies of occupational mobility. Goldthorpe looked at the official government classification schemes in order to identify a large number of discrete categories of employment that can be distinguished from each other by the market and work relations that influence the life chances of those pursuing particular occupations. These occupational grading units are then grouped into larger occupational aggregates. Goldthorpes' argument is that the level or aggregation used by a researcher depends simply on the particular empirical purposes that he or she is pursuing. From my point of view, these aggregates can be regarded as clusters of class situations and might be termed "economic classes".

This approach reached its ultimate operationalization in the work of David Rose and Karen O'Reilly, who developed the Goldthorpe scheme into a comprehensive framework of 371 class situations, with a coding book, and that has become the basis for the collection and analysis of official statistics in Britain and is also used in much empirical research.

The problems that I see in this operationalization are twofold. First, the focus on employment relations has led to a relative neglect of property relations. Hence, it is difficult to use the classification to investigate the higher levels of a class structure. This problem has been partly addressed in the United States in the work of Erik Ohlin Wright, whose approach to class has attempted to distinguish various levels and forms of property ownership and control. This work needs to be more systematically developed in ways that allow the identification of a large number of property aggregates distinguished by the type and amount of property involved. I have begun to work in this direction myself, using my earlier analysis of ownership and control and have provisionally identified 30 or so propertied categories that need to be added to the 371 employment categories.

The second problem with the approach is that its main advocates have seen the highest level aggregation, typically into seven or three categories, as a "social class" classification. As social classes are "real" groups, as distinct from the "nominal" economic categories, this approach is mistaken. The high level aggregations can be seen only as "economic classes". I will return to this point later.

Status is much more difficult to operationalize and there has been far less progress in this area. The main form of operationalization has been through studies of "occupational prestige", one of the principal aspects of status distinction in a class society. In such a society, the expectation is that there would be a high correlation between class situation, as measured by occupational employment relations, and occupational prestige. However, there are a number of other aspects of status that cannot be operationalized in this way.

Status arises also from state allocation mechanisms such as the allocation of titles and honours. These may be listed in official tables of precedence, but a proper operationalization requires also a consideration of the ways in which these are perceived and recognised in popular social imagery. Little or no research has been undertaken on this issue. Status is also generated through the mass media and their creation of "celebrity" and through various private agencies such as national scientific academies, and these need to be combined with the other aspects of status. Finally, there is, of course, much evidence that gender and ethnicity, in particular, are very important bases of status discrimination and these, too, have to be formulated into classifications of status situations. There is scope for a great deal of work in this underdeveloped area.

The operationalization of command situations is slightly more advanced, as it is often possible to use the various levels in official hierarchies of authority as the basic elements in a classification scheme. There is, however, a degree of arbitrariness in drawing lines in an authority hierarchy. There is the question of equivalences between different hierarchies (for example, between military and business hierarchies of authority) and there is the question of where to identify the "top" levels in certain areas. For example, a study of economic elites as those in top corporate positions must decide whether the "top" line is to be drawn at the 50 largest companies, the 100 largest companies, the 200 largest companies, or the 500 largest companies.

The operationalization of social strata, as real social groups, is rather different. Social strata cannot simply be identified as employment aggregates, status aggregates, and/or command aggregates. Social strata, in my scheme, are defined by relations of mobility and association and need to be distinguished by the direct investigation of these relations in order to identify the boundaries and forms of closure that actually group class, status, and command situations together.

Are there any recent or not so recent empirical studies that have fulfilled this task adequately?

There are a few studies that have tried to approach social strata in this way. One is an article by Ken Smith in the British Journal of Sociology for 2007. Smith looked at mobility relations among economic class categories in an attempt to identify social class boundaries. This work was, however, limited to class situations and did not consider status and command situations.

More recently, the work of Mike Savage in Social class in the 21st century has taken a different approach to the identification of social classes. Taking up ideas from Pierre Bourdieu, Savage used multiple correspondence analysis (mca) to investigate interaction and association among various social categories. Social classes are identified as "clouds" within the mca diagrams. On this basis he identified seven social classes in Britain today.

This approach moves in the right direction, though it needs to incorporate mobility alongside interaction and it needs to use the various operationalisations that I have referred to. It is, in my opinion, limited by its failure to properly define social classes in the way that I recommend, but it is clearly an important pointer to the way in which we can move forward in stratification research.

Finally, is it possible to conceive of a multidimensional approach outside of a Weberian framework?

Well, there have been attempts in such writers as Gerhard Lenski many years ago and in various writings on intersectionality, but I see these as only partial solutions. The advantage of my Weberian framework is that it provides the conceptual vocabulary for this work and allows them to be fitted into a larger scheme. Although the approach draws heavily on Weber's conceptualisations, however, it is not necessarily tied to a specifically Weberian sociology. Questions about the relative priority of, for example, structure and action are not prejudged by the adoption of the Weberian distinctions. As I have already suggested, I have always seen this framework as the basis for integrating aspects of Marxist theory and of the system theory of Talcott Parsons. Hence, it is only in a limited sense that I regard myself as a "Weberian". This argument, however, raises far more general issues of social theory that it is not possible to pursue here. I have, however, discussed some of these issues in my book Conceptualising the social world (2011).

Recebido: 29 de Janeiro de 2016; Aceito: 02 de Fevereiro de 2016

Edison Ricardo Bertoncelo é professor de sociologia da Faculdade de Filosofia, Letras e Ciências Humanas da Universidade de São Paulo. E-mail:

Virgílio Borges Pereira é professor do Departamento de Sociologia da Faculdade de Letras da Universidade do Porto. E-mail:

Creative Commons License Este é um artigo publicado em acesso aberto sob uma licença Creative Commons