Skip to main content Skip to navigation


Theoretical Ideas in Sociology: the problem of explanation


I have said repeatedly that since its inception sociology has been haunted by its relationship with other disciplines.  In the first place it had to settle its accounts with  already established and venerable social scientific or humanistic disciplines from which it emerged – law, economics, history, political philosophy; secondly, a number of its founding figures sought to give sociology a status comparable to that of the natural sciences.   If this is true then one of the things it ought to be able to do that history, say, cannot, is to study societies in such a way that confident generalisations may be made about the ways in which societies work and that these generalisations might be used to construct explanations of human action that have the same dignity as those one expects in the natural sciences.    This was the project of Comte and to a lesser extent Durkheim; and although Simmel has a very different approach, even he was ready to describe his own version of sociology as a science (recall his remarks about sociology and geometry, and sociology and the new anatomy (not based on the major organs)) Marx too believes that he is doing science (week 9); the scientific ambition persisted in the work of Parsons, and more recently Garry Runciman has said ‘there is no problem of explanation in the social sciences, only a problem of description’ (meaning – social sciences can explain things as well as natural scientists explain natural phenomena they can just do something else as well) 


The questions is: it is one thing to say that sociology is a systematic discipline, even that it is a science; it is another thing altogether to say that its primary task is to answer ‘why’ questions.  Is that what sociology should be striving for?  To be able to explain social life?  If so, what sort of things is it supposed to explain?  What are the tools and techniques of explanation?  If sociology is an explanatory science then how broad should its ambitions be?  What do we want to know?  Are there things that we want to know that can be adequately known without our requiring an explanation for them?  And in any case, do the phenomena that sociologists study lend themselves to the posing of why questions of the sort posed by, say physicists?



 Classic sociological explanations:

-         Marx’s account of the dynamics of historical change (e.g. ‘Preface to a Contribution to a Critique of Political Economy’)  - laws of history

-         Durkheim’s Suicide  - general relationship between suicide rates and degrees of integration/regulation

-         Weber’s The Protestant Ethic and the Spirit of Capitalism – explain the emergence of an historical individual (not a person but an individual case – and a very large one at that, namely the dominant attitude to conduct characteristic of modern society, or the type of individual who has become dominant in our society (this idea comes from Nietzsche)) 



meaning implies cause!!


·        Nomothetic (of laws) v. ideographic knowledge (of events)  Sociology, most would agree, lies somewhere between physics (laws) and history (events)  does this mean there is no explanation in history?  No but it might mean that history does seek to explain individual events (e.g. the origins of the first world war) rather than arrive at general conclusions.  But nor – usually - are sociology’s more general concerns as general as the search laws.  As Neil Smelser put it: ‘A historical problem, generally speaking, is rooted in and emerges from the logic of events at a given place and period; for example, why did the French monarchy and aristocracy become so unresponsive to demands for social reform during the eighteenth century?   By contrast, a sociological problem, generally speaking, tends to be rooted in and is generated by some conceptual apparatus; for example, what are the relations between blocked social mobility and social protest, as illustrated in the eighteenth-century French case?’ (35-6)  We will come back to this.

·        A famous rejection of the idea that sociology in particular could ever be a science is found in Peter Winch’s book, The idea of a Social Science.  Winch’s big idea is that because human beings aer not inanimate objects or electrons or what have you, and because they use language, everything that happens in social life must be understood in terms, not of its being law-like or not, but in terms of its being intelligible or not.  As speakers of language we live in a social world that is orderly, but this is not because our movements are subject to laws or blind force, but rather because it is meaningful.  In emphasising meaning (rather than cause) Winch wanted to remove any hint of determinism or reductionism from the social sciences.  Ironically enough, though, he introduced quite a lot of it himself because he did say that social life was orderly and relatively predictable, and that the reason for this was that in speaking we followed rules and understood the difference between following rules and making a mistake.  In a way, this idea of rules is as constraining as that of laws.  ‘It’s all about understanding meanings’ can be as rigid as saying ‘it’s all about explanation’.    Winch is sometimes connected with Weber, but Weber was wary of rejecting causality this way.  As the very first sentence of Economy and Society says: ‘Sociology is a  science which aims at the interpretative understanding (Verstehen) of social action in order thereby to gain an explanation of its causes, its course, and it effects.’ (Weber, Economy and Society)  Or: meaning and cause are not opposed but the interest in both may be contained in the same project.  That indeed is what the Protestant ethic is all about, as we will see.

·        The structure of explanation: All explanations have the same structure and consist of two main components:

                                                               i.      explanadum (the thing to be explained)

                                                             ii.      explanans (the things that explain it)


·        Explanation in the natural sciences: hypothetico-deductive method (sometimes known as the deductive-nomological method or the ‘D-N model’ or the ‘covering law model’)

o       Explanadum is regularities or types of event (e.g. water boiling);

o       explanans consists of two components:

                                                            iii.      initial conditions (e.g. water being heated)

                                                           iv.      laws (e.g. that when water is heated to 100deg. Celsius it boils)


o       this law holds whether or not these events take place




·        Lower level ambitions even of a sociology that seeks to explain: empirical generalisations about large scale matters:

                                                             v.      Static models – e.g. Robert Michels – why do even liberal/socialist political parties tend to become oligarchies?  (‘the iron law of oligarchy’) (Smelser, p.13, diagram) 

                                                           vi.      Process models within a static social structure (social mobility, voting rates, types of social control) – Nuffield College, Oxford.

                                                          vii.      Change models – changes in social structure itself.   

  • Comparative sociology between general laws and specific events:
    • Durkheim – begins with official statistics on suicide and asks: why does the suicide rate vary between countries, between men and women, married and unmarried, between religious denominations?  Produces an aeteoligical classification of suicide (classification according to causes) – 4 types only.  And ‘cause’ here is stretching the term because they are all at the end of an entirely theoretical spectrum; the whole study is one of degrees – the tendency of the rate of suicide to rise is dependent on the degree of regulation or integration. 
    • Weber: protestant ethic: why is it that, in the west and only in the west has there emerged that particular set of attitudes – the spirit of capitalism (i.e. not the whole of capitalism) to everyday conduct so characteristic of the modern entrepreneur but not only of the modern entrepreneur but of all people who work in a society centred on the capitalist enterprise (Betriebskapitalismus)?  There are many causes of this everyday rationality – growth of trade, technology, the medieval city, development of business enterprise itself – BUT:
      •  in The Protestant Ethic Weber chooses to look at one – a particular attitude to conduct promoted by particular forms of belief and communicated in the form of practical advice by early modern Protestant pastors. 
      • In the other studies in the sociology of religion he writes about the absence of this set of attitudes –innerworldly asceticism in other great religious traditions.  So China, India, are ‘control evidence’ for PE. 

  • Barrington Moore – Social Origins of Dictatorship and Democracy:
    • why did certain modern states take one route to modernity – democracy (France, Britain, America) others take the route of dictatorship (revolution from above - Germany, Italy – or revolution from below - Russia)?  Because of the class structures of agrarian states undergoing initial stages of economic modernisation.  E.g. ‘labour-repressive’ and market-commercial agriculture produced very different long term outcomes.  In England, market-commercial agriculture created alliances between strong bourgeoisie and agrarian elements, while labour repressive agrarian systems are unfavourable to the growth of democracy.  (table from Skocpol) 
    •  method in Moore (and Weber) - Method of difference and Method of agreement (Mill)  - TABLES
    • conclusion: ‘the English experience tempts one to say that getting rid of agriculture as a major social activity is one prerequisite for successful democracy.’
  • Skocpol – States and Social Revolutions: why were there revolutions in China, France and Russia?  Can we find some general causes common to these three cases?  Why should we want to do this?  Because revolutions and why they occur is a worthy object of inquiry – what we want to know.  The historian may ask: why did the French revolution happen?  The sociologist asks: why do modern revolutions happen and why, by the same token, do they not happen when we might otherwise expect them to?  This sounds quite general but S. thought she was addressing something more specific than more scientistic American sociologists do: for them, social revolution was subsumed uner more general categories (eg ‘rebellion’ or ‘violence’ – e.g. Ted Gurr, Why Men Rebel – danger that instead of broad generalisation you get only banality) Controversy in Skocpol:
    • a structural explanation, i.e. very little room for contingent events, individual decisions, de-emphasis on forces from below, key focus on external geopolitical pressures leading to state breakdown – modern revolution is never a domestic affair,  etc.  We are not reconstructing separate narratives, but trying to arrange large bodies of historical material into a pattern in order to get an overview sufficiently clear to offer a hypothesis about why these revolutions that did occur…occurred.   But we are absolutely dependent, as the historian is, on massive empirical evidence, and we are not looking for ‘laws of revolution’ but general conclusions that may allow us to make sense of other revolutions that are not covered in this study. 
    • For some Marxists, French revolution often called a ‘bourgeois revolution’, Russian and Chinese a proletarian revolution.  For S., they can be included in the same analysis as belonging to the same ‘type’.
    •   First claim (quotation): e.g. Russia – relatively successful industrialisation modernisation of agriculture 1890s to 1914.  IF Russia had not participated in World War I, it might have developed industrially under bureaucratic guidance (like Germany) But World War I – mobilisation and politicisation of peasantry  (as soldiers, then as former soldiers – cf. Marx in the 18th Brumaire):  A smallholding, a peasant and his family; alongside them another smallholding, another peasant and another family.  A few score make up a village, and a few score of villages make up a Department.  In this way, the great mass of the French nation is formed by simple addition of homologous magnitudes, much as potatoes in a sack form a sack of potatoes.  In so far as millions of families live under economic conditions of existence that separate their mode of life, their interests and their culture from those of the other classes, they form a class.  In so far as there is merely a local interconnection among these small-holding peasants, and their identity and their interests beget no community, no national bond, they do not form a class…They cannot represent themselves, they must be represented. Their representative must at the same time appear as their master, as an authority over them…The political influence of the small-holding peasants, therefore, finds its final expression in the executive power subordinating society to itself’. (Marx, The Eighteenth Brumaire of Louis Bonaparte)

    • Second claim (quotation on breakdown of societal controls) ; defeat due to military inferiority, demoralisation of army: ‘The Russian revolution occurred in 1917 because Russia was too inextricably entangled with foreign powers, friend and foe, economically and militarily more powerful than she.’  Traditional landed upper classes had no more investment in state structures than peasantry and could therefore engender (often unwittingly) military/administrative crisis/disorganisation: e.g. France in 18th century – tax and property system defended by parlements was inadequate to a French state that needed to modernise, and which needed resources to fight the many wars it became involved in (eg American war of independence); China – reform in late 1800s involved dynastic elite trying to counterbalance ‘entrenched obstructive forces’, but each time they did, new officials were absorbed into local and regional cliques.
    • Third claim (quotation on peasant insurrections): What does this mean: large estates worked by serfs or labourers are inimical to peasant rebellion – e.g. Germany (large Junker estates east of the Elbe).  Close supervision and discipline by landlords. Vulnerability increases when sanctionisng machinery is centralised and when agricultural work and peasant life are controlled by peasant families /communities themselves – e.g. France and Russia.  Russian govt. after 1905 tried to introduce internal class differentiation into peasant communities, increase migration to urban areas, but with little success. 
    • Fourth claim (quotation) Marginal elites drawn from skilled/university educated, oriented to state (and therefore often closer to a ruling monarch than to traditional regional elites – the latter were interested in liberty, the marginal elites interested in reform – where have we heard that before?) 
    • Final claim: ‘Social revolutions have not been caused by avowedly revolutionary movements in which an ideological leadership mobilises mass support to overthrow an existing system in the name of a new alternative’ (241)






  • Cause versus meaning?   Well, causality seems to involve a relationship between dependent and independent variables, explanandum and explanans.   BUT in the case of Weber and of Moore, the attempt to explain something changes substantially our interpretation of it: e.g. by tracing some of the roots of the spirit of capitalism Weber enriches our sense of what that spirit means; by tracing the social roots of democratic political outcomes Moore invites us to redefine the meaning of democracy itself – not just a form of government but a set of social structures and class relations as well. 
  • Structural explanations and historical inevitability:  however many different factors may be invoked (e.g. in Skocpol) if all of them rule out the decisive role of movements from below, what does this say about human agency?  If revolutions only happen as a result of state weakening, what do we say about the collapse of communism?  What role was played by the opposition movements in those countries?  Was the collapse decisively the result of external modernising pressures that the communist states could no longer resist (pressures that included not only the arms race but also consumer goods)    Or do we say as some do that the ruling cadres in those countries themselves saw the writing on the wall and worked out that they could pursue their own interests far more successfully under a new system? 












Theoretical Ideas in Sociology: explanation

‘For any given dependent variable in sociology, the number and kinds of conditions that potentially affect its variation are, at first sight, discouragingly great…The initial picture…is one of a multiplicity of operating conditions, a compounding of their influences on the dependent variable, and an indeterminacy regarding the effect of any one condition several conditions in combination.  The corresponding problem facing the scientific investigator at this stage is to reduce the number of operating conditions, to isolate one condition from another, and thereby to make precise the role of each condition’ (Neil Smelser, Essays in Sociological Explanation


‘A historical problem, generally speaking, is rooted in and emerges from the logic of events at a given place and period; for example, why did the French monarchy and aristocracy become so unresponsive to demands for social reform during the eighteenth century?   By contrast, a sociological problem, generally speaking, tends to be rooted in and is generated by some conceptual apparatus; for example, what are the relations between blocked social mobility and social protest, as illustrated in the eighteenth-century French case?’ (35-6)  

‘The historian, attentive to a problem and period, interprets the causes of events as he finds them, as it were, and thus takes on the natural eclecticism of causation in an uncontrolled historical sequence of events.  Thus, the historian is prepared to admit invasions, personality characteristics of kings, population increases, changes in landownership patterns, and social protest movements as causes, if these appear to be important for his particular historical problem.  The relatively more systematic social scientist, on the other hand, attempts to hold constant various of these events by diverse means of situational and conceptual manipulations, thus isolating, simplifying, and making less eclectic his concern with causes’  (36)

 ‘The problem of explanation in history is also the problem of the nature of sociology.  The views adopted in this field are held to have profound moral and political implications.  We have recently often been reminded of this.  The simplest argument connecting a premises about the nature of historical explanation with political or ethical consequences runs as follows: if rigid, unchangeable, and wide-ranging generalizations are attainable with regard to historical processes, then an outlook which presupposes individual responsibility is basically misguided.  Having pointed out this implication, philosophers hostile to the conclusion then devote themselves to undermining the premises.  They do so either by pointing out that the required historical laws have not been found or by arguing that they could not be.’ (1)

(Gellner, Ernest, 1973, ‘Explanation in History’, in Cause and Meaning in the Social Sciences, London: RKP )













Skocpol on social revolutions:

First claim (general overview):

a)      ‘All modernizing agrarian bureaucracies have peasants with grievances and face the unavoidable challenges posed by modernization abroad.  So, in some sense, potential for social revolution has been built into all modernizing agrarian bureaucracies.  Yet, only a handful have succumbed.  Why?  A major part of the answer, I believe, lies in this insight that ‘not oppression, but weakness, breeds revolution’.  It is the breakdown of a societal mode of social control which allows and prompts social revolution to unfold.  In the historical cases of France, Russia, China, the unfolding of social revolution depended upon the emergence of revolutionary crises occasioned by the incapacitation of administrative and miltary organizations.  That incapacitation, in turn, is best explained not as a function of mass discontent and mobilization, but as a function of a combination of pressures on state institutions from more modernized countries abroad and (in two cases out of three) built-in structural incapacities to mobilize increased resources in response to those pressures’ (Skocpol, Theda, 1976, ‘France, Russia, China: explaining social revolutions’, in Social Revolutions in the Modern World, 138)

b)      ‘It cannot be argued that the cognitive content of ideologies in any sense provides a predictive key to the outcomes of the Revolutions’ (Skocpol, 1979, States and Social Revolutions, 171) 

Second claim (breakdown of societal controls): ‘Leaving aside value-orientations and individual characteristics, we must look at the class interests and connections of state officials.  The adaptiveness of the earlier modernizing agrarian bureaucracies was significantly determined by the degree to which the upper and middle ranks of the state administrative bureaucracies were staffed by large landholders.  Only state machineries significantly differentiated from traditional landed upper classes could undertake modernizing reforms which almost invariably had to encroach upon the property or privileges of the landed upper class’. (141) 

Third claim (peasant insurrections): ‘Agrarian bureaucracy has been the only historical variety of complex society with a differentiated, centralised government that has, in certain instances, incubated a lower-class stratum that was simultaneously strategic in the society’s economy and polity (as surplus producer, payer of rents and taxes, and as provider of corvée and military manpower) and yet organizationally autonomous enough to allow the ‘will’ and ‘tactical space’ for collective insurrection against basic structural arrangements’ (147)

Fourth claim: ‘…the entry of marginal elites animated by radical nationalist goals’.  These marginal elites did not come from bourgeoisie (merchants, financiers, industrialists) – for Skocpol the modernising role of the bourgeoisie was to modernise ‘the contexts within which bureaucrats, professionals, politicians, landlords and peasants and proletarians have engaged in the decisive political struggles.’ (154) Marginal elites drawn from skilled/university educated, oriented to state (and therefore often closer to a ruling monarch than to traditional regional elites – the latter were interested in liberty, the marginal elites interested in reform – where have we heard that before?) 

Final claim: ‘what changed most thoroughly in all of the historical social revolutions was the mode of societal control of the lower strata’.   This is more important than the q. of whether, say the French revolution was a ‘bourgeois revolution’.  Even if the Fr revolutions was ‘made’ by lawyers, officials and intellectuals, ‘The men who ruled France after the revolution were bureaucrats, landowners, soldiers, commercial and financial capitalists, much as before’ (158)
































Method of agreement



Positive case

Negative case








not x


not y

Method of Difference