“Reality is infinitely diverse, compared with even the subtlest conclusions of abstract thought, and does not allow of clear-cut and sweeping distinctions. Reality resists classification.” (Fyodor Dostoyevsky ‘The House of the Dead’)

Trying to describe the world in words or numbers is a challenge. We struggle to find the language and models to represent reality and they can only offer an imperfect representation of our experience of it. To name and describe things is to classify them and to impose one perspective rather than another on our understanding of things.

We’re constantly invited to understand our social world by breaking it up into categories. People are Remainers, Brexiters, millennials, Northern, Southern, disadvantaged, privileged… From baby boomers to gen. Z, these broad-brush classifications frame much of our public debate about people’s aspirations and behaviours.

The use of these categories is often linked to a type of reification, where complex dynamic social phenomena are elevated into an organising principle (eg: ‘intelligence’ or ‘knife crime’). It’s often combined with agglomeration which takes things further by lumping together many different reified interactions and describing them as examples of the same thing. So, this ‘lump and label’ involves deciding what matters, naming it and framing it, then play it back to confirm pre-existing assumptions.

But viewing people’s characteristics and behaviours through the lens of a single category is more likely to obscure than to shine a light on reality. It can substitute for genuine debate by narrowing the range of what is considered or acceptable. Do we really believe, for instance, that ‘red wall voters’ have some unique common perspective? Or that ‘disadvantaged’ people all have the same aspirations?

We know that people are complex and have a range of views on different things and that these are changeable. We can see that these categories are crude, and yet we allow them to enter our consciousness and limit our thinking. Useful discussion is shut down and prejudices reinforced and what is presented as a fresh new idea can turn out to be a lazy old stereotype.

That said, we can’t do without categories. Naming and identifying things based on their differences or similarities can help us understand them better. At the group or population level, we need to aggregate large amounts of data about individuals in a range of ways. Social science would struggle to draw any conclusions about patterns of behaviour or social trends without being able to classify people using characteristics such as age, sex, ethnicity, income or class and the use of ‘protected characteristics’ plays an important part in helping us understand patterns of structural inequality.

We know that classifying people requires simplification and loss of detail. We need to understand the limitations and we accept them because of the benefits of being able to group things easily and to try to establish some broad general rules about them.

The process of simplifying a complex world by viewing it through a particular lens is never a neutral one. The decisions about which categories matter are loaded with assumptions and values. Defining and foregrounding particular characteristics is being done by someone, usually with power, with some purpose in mind, usually the maintenance of that power, and the choice to ignore or downplay other characteristics is just as significant.

Whether we’re dealing with heavily value-laden descriptors such as ‘dysfunctional families’, ‘unskilled workers’ and ‘deprived areas’ or less ambiguous ones such as ‘benefit claimants’, ‘non-graduates’ and ‘coastal towns’, they are selected, connected and interpreted in ways which reinforce a particular narrative or power structure. The simplistic judgements made can reinforce existing prejudices and essentialist beliefs about ‘human nature’ or ‘how things are’, correlations can slide towards half-baked explanations and definitions of ‘how things should be’.

When a single category becomes an organizing principle for policy it’s in danger of being stretched beyond its usefulness. Information about a large number of people has been split into categories and aggregated ‘upwards’ to build up broad judgements at the social level. If this high-level approach is then played back ‘downwards’ to the individual level, there is a massive loss of definition and focus and the complex blend of characteristics of real people is reduced to caricature, with the potential to mislead.

When emotive or charged language is overlaid on top of ‘established facts’, real damage can be done. For example, the poet Caleb Femi has spoken about how the racialised use of the term ‘notorious’ to refer to the North Peckham estate labelled the people who lived there, in effect blaming them for the estate’s structural and design problems.

In education too, we fall easily into the same habits. So, for example, wildly general descriptions which sum students up as ‘academic’, ‘bright’ or ‘motivated’ are routinely used to make distinctions which can influence what is offered to them and what is expected of them. When these judgements have a metric attached this can provide a veneer of validity, even if all the numbers tell us is that for a variable characteristic x there is a smooth distribution within the population from a to z with a mean somewhere around m.

So, once we have invented a measure of achievement or even ‘ability’ based on a test, there will inevitably be high and low scorers who can then be the subjects of targeted interventions and attempts at ‘levelling up’ or promoting ‘social mobility’.

The courses being taken by students can also be labelled, and these can  become attached the students themselves. So, we find ourselves describing students as ‘vocational’ or ‘academic’ as if these types of course represent inherently different types of student with distinct interests and aspirations.

Too often, definitions of ‘ability’ or ‘aptitude’ are used to support exclusionary or selective practices. Institutional decisions to keep certain students out are justified by extrapolating back from what are presented as essential categories, such as ‘academic students’, although they are based on fairly arbitrary measures, such as a test score at a particular age. The selective structures themselves can then become the justification for the selective practices, ie: ‘we select for this type of student therefore this makes us the right kind of provider for them.’

The type of provider students attend can also be used to define the students themselves. Is an undergraduate attending a more selective ‘high tariff’ university necessarily more capable than another who doesn’t? Is a school ‘sixth former’ different from a ‘college student’ or a ‘grammar school student’ different from a ‘comprehensive student’ if they are studying the same course? What do these descriptions really tell us about the students themselves and their experience of education?

Data on achievement and progression can provide us with insights into the many layers of inequality in our society. Measures of social class or income can provide a clear sense of the differential benefits which students get from the education system. Using Free School Meal (FSM) eligibility as a measure provides a binary categorisation; you’re either eligible or not, and FSM students are more disadvantaged in every respect. Using IDACI or POLAR quintiles provides a clearer picture of the disadvantage gradient across 5 categories, charting the clear educational benefits of being better off. So, for example, the poorest 20% have the lowest rate of progression to university and the richest have the highest rate.

None of this seems very surprising; these data are revealing the patterns of disadvantage rather than explaining them. The figures themselves tell us little about the mechanisms which perpetuate these inequalities. But it’s clear that the use of categories to classify people can provide the means to discriminate, segregate and oppress them.

When different categories are used together, the choice of association carries with it all sorts of assumptions about which dimensions to combine and which intersectionalities are of interest. In doing so, we risk a kind of multiplication of selective judgements leading to an algebra of confusion. For instance, to what extent is the experience of ‘white working-class boys’ associated with their ethnicity, their class, their sex, some relationship between any of these factors, or something else? What is the category ‘white’ adding to the mix? Is it simply a way of ignoring the experience of black working-class boys, denying white privilege or downplaying the impact of racism on life chances?

The way we define the challenges we face will shape the kinds of debates we have and the kinds of solutions we might want to advocate. The way we use categories doesn’t do justice to the multi-layered complexity of the world and it’s often a short cut to sweeping assumptions and flawed policy.

In some cases, we may just need to see the use of categories as a helpful first stage which suggests clues but begs many questions which need further research. In many cases, the answer may be to look at things in another way, to reframe the question or start with different assumptions; for instance, is it the social structure of privilege that is the problem rather than something inherent to those who lose out because of that structure? Could what appears to be a group ‘deficit’ be seen instead as an opportunity to rethink what we’re measuring?

So, before we draw policy conclusions based on the use of simple categories, perhaps we should ask:

• What is the underlying issue and why are we defining it in these particular terms?
• Why choose these categories, or combinations of categories, and not others?
• How do the people most concerned by this issue define it themselves?
• What underlying assumptions and definitions might we want to question?
• What multiple perspectives could we take to add depth to our analysis?
• What more do we need to understand before drawing any conclusions which might shape policy?
• How will we evaluate the impact of a policy, both on individuals and systemically?

It’s not so much the use of categories which is the problem, but the choices we make, the range we choose from, the relative emphasis we give them and the social impact of the conclusions we draw from them.

It must be possible to use categories in ways which do justice to the complexity of the social world but this means being more open-minded and critical, considering alternative perspectives, including those of the people most affected, being more tentative about drawing conclusions and more careful in implementing any solutions.

Illustration: p.13 of the brilliant ‘Unflattening’ by Nick Sousanis

See also:

Edgar Morin on ‘Thinking Global’ (August 2017)

Challenging IQ (August 2017)

Reducing culture to memes (August 2015)

Education as a whole and in its parts (November 2014)

Maxine Greene, resisting one-dimensionality (June 2014)

Blob and anti-blob (May 2014)