The Double Social Life of Methods

I wrote a few days ago about The Social Life of Methods, an interesting programme of research undertaken by Evelyn Ruppert, Mike Savage and John Law. This is orientated towards what the authors term the ‘methodological complex’: a dominant way of understanding method, tangled up in a particular division of labour, which precludes the investigation of methods as objects in their own right. In contrast to the ensuing view of methods as neutral tools for producing knowledge about the social world, they want to study the ‘social life’ of these methods and consider how this has shaped their emergence. In this paper they talk about the dual sense in which methods can be said to have a social life:

Methods are shaped by the social world

Methods don’t come into being without a purpose and methods don’t come to prominence without advocates. In this sense, we can link techniques of map-making and surveying to the needs of nascent nation states to map their territory. The census came about as a result of the need to constitute a governable national population, measurable and therefore susceptible to intervention. For instance, Servius Tullius, the sixth king of Rome prior to the republic, instituted a census and divided the common people (plebs) into social classes: these were defined hierarchically, with rights and responsibilities ensuing from what they could contribute to the city. All the sixth class were seen to be able to contribute was their children (source). The UK census began in 1801 and, with the exception of 1939 and 1966, has been undertaken every ten years since. The present government has discussed replacing the census: “There are, I believe, ways of doing this which will provide better, quicker information, more frequently and cheaper” opined Cabinet Office Minister Francis Maude in 2010. The technology used to conduct the census has changed radically, with a move towards handheld computers and online completion perhaps giving rise to an entirely online census supplemented by existing data or even its complete abandonment. The classifications used in the census change, existing classifications change, for instance ‘pensionable age’ increases, and new classifications are introduced, such as ‘ethnic group’. In fact this didn’t appear in the census until 1991, with Margaret Thatcher having dismissed its proposed introduction in the 1981 census, seeing it as a left-wing dogma incompatible with her planned direction for the UK (weirdly placing her in alignment with some on the radical left who were suspicious of the government collecting such data at a time of racial tensions and with a rising far-right). The purposes underlying the census change, both in terms of the intentions of the government but also the uses to which groups like academics and charities make of the census data, leaving different groups with vested interests who help shape its future direction. The vast majority of the population of England and Wales completed the census form and, while there are legal sanctions for a refusal to complete it, this degree of compliance necessitates an extensive public engagement operation and makes winning public support crucial. In this sense, the logistics of the census itself are of interest, given how all this activity is compressed into the space of a few weeks.

What Ruppert, Savage and Law are arguing is that we miss all the complexity if we simply treat the census as a neutral tool to study an object. It also helps constitute that object, ‘chopping up’ a population in various ways according to political needs and social contingencies, though the ensuing data cannot be dismissed as simply a product of the exercise. The census helps constitute a ‘population’ susceptible to intervention. We confront similar questions when considering the UK government social surveys, which began in 1941 during the second world war.

while the census made people inhabiting a territory into a national population, subsequently that population reality could be further calibrated through the technique of sample statistics in the mid-twentieth century. So there’s a history of method to be told here. In the US, for instance, sample surveys grew up with the Gallup polls of the late  1930s, and then with agencies such as the Department of Agriculture during the Second World War. Mike Savage traces their analogous though later rise in the UK. He shows, for instance, how during the Second World War, the Government Social Survey became a key instrument in generating knowledge of the circumstances and concerns of the British population. It proved popular in part because its methods of anonymous sampling avoided relying on known informants, in a way that had attracted popular opprobrium. The government proved a key player in promoting survey research into the 1960s, seeing it as part of a modernising form of government that no longer needed to rely on the views of the ‘good and the great’. It proved important in shaping educational reform in the 1960s, with surveys being used to explain the social selectivity of grammar schools. As Hilton and Savage have shown, these methods were thus embraced by a technocratic middle class, seeking to distinguish themselves from older gentlemanly intellectuals

To talk of the social life of methods entails inquiry into the purposes underlying methods and advocates for particular methods, as well as how these change over time. These purposes might not be lofty or consequential ones. The focus group was largely absent from the academy until the 1980s but had been developed and refined as a tool for market research. Likewise corporations have led the way in the analysis of consumer transactional data obtained through mechanisms such as store cards. Considering the social life of such method entails looking at the origins of these methods, their purposes and advocates, as well as how they have changed as they moved out of the commercial sphere and into the academic one. What assumptions are loaded into their use? How are their operations understood by those who us them? What status is ascribed to the data that is produced using them?

Method contribute to shaping the social world

The notion that methods are shaped by their context is relatively straight-forward. It might be under-theorised and understudied but it’s unlikely to be a contentious claim. However the authors argue that methods also contribute to shaping that world. This notion of the performativity of method can be slightly harder to grasp. They offer the sample survey as an example:

Take an example: the sample survey. This is not any example, because the sample survey is one of the most legitimated methods in use today. In the UK University teaching is assessed by the scores academic departments achieve in the National Student Survey. Our measures of crime are dependent on the British Crime Survey, which ‘corrects’ for the under-reporting of much crime to the police. The inflation rate is determined by responses to the Family Expenditure Survey, and so on.

The survey works by first sampling people. And then it works by asking them questions about matters of fact (like age, gender, income bracket, religious affiliation, lifestyle choices) and matters of opinion (such as the performance of the government, attitudes to abortion, or meat- eating). We might think of a survey as a bit like a methodological package deal. Like all package deals it has great virtues. It tells those who advocate it something about how people are planning to vote, or, say about their attitudes to global climate change. It’s also exceedingly useful because the basic methodological thinking doesn’t have to be done again. Surveying, after all, is a bog-standard method that has been industrialised and routinised. Its standards of quality control have been set, are widely agreed, and ethical guidelines are in place to police them.

At the same time, like all package deals, it is indeed standardised. You get to see parts of social reality in particular ways, while you don’t see things that escape the package. Or more strongly (and now we’re getting to the point we want to make about constituting), it may be that you get to perform certain kinds of social realities whilst not performing others. You’re actually bringing realities into being while you’re shutting down others

I think the framing here is confusing. But the underlying point is an interesting one. On the most basic level, people (sometimes) act on the basis of research and so what the methods open up and what they close down shapes the ensuing action. In so far as research methods helps shape how people see the social world then methods have consequences, including the oversights and blind spots latent to any particular method. As Einstein (allegedly) put it, “Everything that can be counted does not necessarily count; everything that counts cannot necessarily be counted”. Things that matter may not show up and, conversely, things that don’t matter may show up’. However the authors are making a much deeper point than this. Research methods construct individuals as having capacities and properties. To ask someone about their choices presupposes their capacity to choose. To ask someone about their attitudes presupposes that their reported attitudes at one moment can tell us something about their real attitudes at another i.e. it assumes their relative stability. Methods also construct collectives as having certain properties:

Individuals are abstracted from a place and then taken as representatives of that spatially delimited place. A particularly important example is the modern nation. There is no independent nation which does not appropriate to itself the ability to conduct censuses and surveys on its national population delimited by its sovereign boundaries. It is not incidental that new and developing states see the capacity to conduct national censuses and surveys as central parts of their ‘statehood’

This process of abstraction constructs a ‘population’ composed of individuals. In measuring this (newly constituted) population, they become susceptible to intervention. Who needs what? Who can contribute what? The answers to these questions may have become far more complex but the questions themselves are still rather similar to the ones addressed by Servius in pre-Republic Rome. These abstractions enter into the lives of individuals in terms of the expectations placed upon them, the facilities offered to them and the interventions they are subject to. Consider for instance, the vast auditing apparatus involved in the tax system. Methods address subjects as beings with particular sorts of competencies and encourage them to respond as such: for example a life history interview makes a whole range of assumptions about the coherency of an individual’s trajectory and their capacity to recall it. These ontological assumptions may be explicit but they’re often implicit. The latter case makes it even more imperative that we analyse methods and extract these latent claims.

The point the authors are making is that research methods have to “pick and choose between different individual and collective realities”. Methods assume the stability of their object but they also contribute to the reproduction or transformation of that object by foreground certain aspects and backgrounding others. They offer this as an argument about all methods, as opposed to quantitative or qualitative ones. We discover things about the world through methods but we also act in and on that world in way that contributes to changes in the objects we study.

First we’re saying that they make discoveries about the world, and that those discoveries may surprise us. That’s why we conduct interviews and surveys and all the rest. But also, and counterintuitively, we’re saying that they also make more or less self- fulfilling assumptions about the character of the social world. And that in so doing they tend to constitute it, so to speak, below the radar in ways that we scarcely notice. In short, that they tend to produce what John Law calls collateral realities: that is, realities that we don’t think about very much but that we’re all busy reproducing as we go about the daily methodological work of gathering and analysing data ‘about’ the social

This suggests that studying methods as objects in their own right encompasses theory: what theoretical assumptions are loaded into methods, how do they condition the enactment of the method and the status ascribed to the results? It encompasses history, in so far as that studying the emergence of the methods, its advocates and purposes, helps us understand how that method came to be constituted in the way we find it. It involves technique, in the sense that we must attend to the operation of the method but without bracketing it off in the manner implicit in the ‘methodological complex’.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.