Sociology and Statistics in Britain, 1833–1979

Plamena Panayotova

• Publisher: Palgrave Macmillan
• Year: 2020
• Print ISBN:
  9783030551322
• eBook ISBN:
  9783030551339
• Topic: Ciencias sociales
• Subtopic: Filosofía social

At the beginning of the twentieth century, Britain stood at the forefront of science and statistics and had a long and respected tradition of social investigation and reform. But it still did not yet have a 'science of society.' When, in the early 1900s, a small band of enthusiasts got together to address this situation, the scene was set for a grand synthesis. No such synthesis ever took place and, instead, British sociology has followed a resolutely non-statistical path. Sociology and Statistics in Britain, 1833-1979 investigates how this curious situation came about and attempts to explain it from an historical perspective. It uncovers the prevalence of a deep and instinctive distrust within British sociology of the statistical methodology and mindset, resulting in a mix of quiet indifference and active hostility, which has persisted from its beginnings right up to the present day. While British sociology has thrived institutionally since the post-war expansion of higher education, this book asks whether or not it is poorer for having failed to recognise that statistics provides the foundations for the scientific study of society and for having missed opportunities to build upon those foundations. Ultimately, this important, revealing and timely book is about British sociology's refusal to come to grips with a modern scientific way of thinking which no discipline that aspires to an effective study of society can afford to ignore.

Part IWhy Study the Historical Relationship Between Sociology and Statistics at all?

© The Author(s) 2020

P. PanayotovaSociology and Statistics in Britain, 1833–1979https://doi.org/10.1007/978-3-030-55133-9_1

Begin Abstract

1. Introduction

Plamena Panayotova¹

(1)

Department of Sociology, School of Social and Political Science, The University of Edinburgh, Edinburgh, UK

Plamena Panayotova

Email: [email protected]

End Abstract

Anyone who has experienced the more recent relationship between British sociology and statistics will be aware that this relationship has been and remains largely unproductive. While this has been recognised in some circles (British Academy 2012; Byrne 2012; ESRC et al. 2010; Goldthorpe 2015; MacInnes 2010; Payne et al. 2004; Williams et al. 2004; and Williams et al. 2008), a solution that reaches all sectors of sociology, not just those interested in ‘statistical sociology’, has not been found. Quantitative methods have not been totally absent from the sociological curriculum but, nonetheless, the effective incorporation of these methods into sociological teaching and research in Britain has recently been limited, despite their increasing importance in a digitised and data-oriented world in which the study of human behaviour, including social human behaviour, shows promising potential.

The present study can broaden our understanding of the current situation by showing that the relationship between sociology and statistics in Britain has developed along a particular long-term evolutionary course; and by explaining why this course of development was not inevitable but was a result of the interplay of numerous context-specific factors, personal choices and institutional decisions. This study can also provide us with useful analogies that can aid us in our assessment of the present and future—without historical knowledge, it would be very difficult, if not impossible, to establish what is different and unique about the current situation and see the opportunities, the alternative possibilities, that lie ahead.

However, a study of the relationship between statistics and sociology has a more far-reaching justification. As the following sections will show, an historical study of this relationship has an enormous explanatory power that facilitates a complete re-evaluation of existing scholarship; enables us to explain why a major intellectual and institutional division has existed in British social science throughout the twentieth century and allows to investigate the very foundations on which British sociology has been built.

History—Balancing the Present and the Past

The specific question of what, historically, influenced the relationship between statistics and sociology, has not previously been asked. This question poses certain historiographical challenges.

Ever since the Whig interpretation of history was identified and condemned by Herbert Butterfield in 1931, historians have been wary of falling into its trap. The Whig or ‘presentist’ approach places emphasis on the similarities and analogies between past and present (Butterfield 1965 [1931]: 10) and extracts things from out of their historical context, judging them in relation to present concerns (Butterfield 1965 [1931]: 30). It upholds the present as absolute against which all things past are to be measured in order to reveal who is in the right, as opposed to showing how people and ideas came to differ over time (Butterfield 1965 [1931]: 16, 30, 130). The danger of this approach is that, either intentionally or unintentionally, the historian distorts the past by assuming that something they hold to be true today should guide their judgement and understanding of past events.

Some sociologists who have studied the history of British sociology have fallen into the trap of ‘presentism’, a prominent example being Abrams’ 1968 essay on the origins of British sociology discussed in Chapter 2. However, presentist writing can be avoided with careful consideration. Butterfield’s advice for historians was to focus on ‘the unlikenesses between past and present’ (Butterfield 1965 [1931]: 10) and embrace ‘the belief that we can in some degree enter into minds that are unlike our own’ (Butterfield 1965 [1931]: 9). Following Butterfield’s advice, this study emphasises changes in the contemporary social, political and scientific context of nineteenth-century statistics; changes in the understanding of science; and changes in the organisation of higher education and its rapid expansion in the post-war period. It also pays special attention to the variety of historical contexts in which the relationship between sociology and statistics developed—an historical examination is more fruitful, and faithful to the past, when it consists of detailed and contextualised examinations of particular events, processes or movements rather than attempts to construct a general overarching explanation connecting all examined events or movements. This, of course, is not to say that we should ignore the degree of continuity where it exists between one set of events and another but it is equally important to recognise where the links are missing.

However, to produce an historical account of the development of sociology it is not enough to commit to the general principles of non-presentist history writing. A good piece of history that has the potential to improve historical knowledge does not imply a total rejection of the use of the benefit of hindsight and of the possibility of finding long-term pervasive trends in history; on the contrary, it aims to strike a balance between interpretation based on detailed contextual analysis and interpretation based on hindsight thinking. Without the benefit of hindsight, there would be no clear way of deciding what is of historical significance and what to study in the first place. Thus, it is only through the benefit of hindsight that we could recognise the potential and long-lasting effectiveness of statistics in studying both the social and the natural world, and, consequently, consider it important to study its historical relationship with other subjects. This, of course, is different from assuming that because we can see statistics as important now, statistics was always seen as important. A balance between the present and the past is necessary and can only be achieved by reading history both backwards and forwards.

There is another reason why the benefit of hindsight is useful in history writing—we can compare the intentions of a person or an institution when they chose to act in a particular way with the consequences of their actions which we, with the benefit of hindsight, know, but which they could not possibly have known with any certainty. For instance, the Sociological Society was founded with intention to unify all existing social science specialisms into the general science of sociology, to set up a sociological journal and establish the subject academically. However, if in our study of the Society, we were only to focus on their intentions, without recognising that the consequences of their efforts fell short of fulfilling their intentions, our account would be extremely limited. This is one way of examining a particular event, in this case the work of the Sociological Society, using both their understanding of their work, as manifested in their intentions, but also our understanding of their work, based on their legacy. Therefore, although Butterfield is correct is arguing that ‘to assume the present at beginning of our study of history’ can potentially lead to distorting accounts, experience shows that he is not correct in equating this with the keeping of the present ‘as a reference’ (Butterfield 1965 [1931]: 62–63). In fact, only by keeping the present as a reference could we begin to recognise the differences between past and present that Butterfield says are so important when we write history.

A good piece of historical analysis also acknowledges that history is not entirely the product of deliberate decision-making. In Butterfield’s words, history is not only a product of ‘agency’, but also of slow and incremental processes (Butterfield 1965 [1931]: 50). This is a criterion which existing accounts have failed to adequately satisfy and on which the present study aims to improve. As I will show, the difficult historical relationship between sociology and statistics throughout the nineteenth and twentieth centuries was not the result of anyone’s careful planning.

But something fundamental that even Butterfield does not mention, is that history is also not entirely a product of rational choices. It is a well-known principle, for instance, that the choice of research methods depends on the research question at hand and that all methods are, in principle at least, equal. Although few would disagree, what we see in practice is that some methods are ‘more equal than others’ and that the choice that sociologists make regarding what research methods to use is rarely determined solely by their research question. Moreover, sociologists can abstain and have abstained from asking particular questions because that would involve using research methods they do not favour. It is not always possible to explain fully why the level of receptiveness to certain methods, or ideas more generally, has varied with time, especially when talking about ideas that are well received in our time but were less well received in the past. But it is important to try to explain what drove the presence or absence of receptiveness; and, I argue, that while sociologists did not make a rational choice or take a deliberate decision to restrict British sociology’s engagement with statistics, the difficult relationship between the two subjects resulted from the effects of a series of interrelated beliefs and opinions about statistics that spread and evolved from one generation of sociologists to the next. And, without apportioning blame or making judgements, the statisticians themselves, it seems, did little to try to turn these beliefs and opinions around.

Sociology—The British Case

The development of social science in Britain has been characterised by a peculiar divide between academic sociology and the British empirical tradition of social enquiry that developed in other academic departments and civic and government agencies. Bulmer among others has argued that it is ‘hardly an exaggeration to say that almost all of the empirical social research undertaken in Britain between the wars went untouched by what then passed for academic sociology’ (Bulmer 1985: 4). This book provides further evidence that throughout the nineteenth and twentieth centuries the empirical tradition of social enquiry has included most of the statistical research done in Britain, while academic sociology has consistently been oriented towards philosophical and theoretical work. An historical examination of the relationship between sociology and statistics offers a suitable opportunity to find out why British sociology as a whole has developed in this particular way and why British social science more generally has been split, given that this did not necessarily have to be the case.

Evidence for this comes from the experience of other countries, such as the USA, where, as historians have observed, the introduction of statistical training and thinking into sociology towards the end of the nineteenth and beginning of the twentieth centuries was a result of particular academic and institutional conditions. The fact that this integration occurred early o...

Table of contents

• Cover
• Front Matter
• Part I. Why Study the Historical Relationship Between Sociology and Statistics at all?
• Part II. Sociology and Statistics in Britain in the Nineteenth Century
• Part III. Sociology and Statistics in Britain from 1903 to the 1920s
• Part IV. Sociology and Statistics in Britain from the 1930s to 1979
• Back Matter