No such thing as society: Thatcher, Wittgenstein and the philosophy of social science

Abstract

This article argues that society is not a thing. It is abbreviated and adapted with permission from a public lecture, titled There Is No Such Thing as Society: Margaret Thatcher, Ludwig Wittgenstein and the Philosophy of Social Science. The original was presented by Gavin Kitching to the Cuadernos de la Catedra Ludwig Wittgenstein at Universidad Veracruzana, Mexico, in March 2019.

Introduction

One of the few advantages of being an elderly, retired academic is that one is allowed to be more informal and self-indulgent than one was when striving, earlier in one’s career, to gain the respect of peers, elders and betters. It is generally regarded as both unprofessional and presumptuous for a young, or even youngish, academic to speak or write autobiographically. In fact, even using the first-person-singular pronoun, let alone speaking reflexively of one’s own ‘thoughts’ or (heaven forbid) ‘feelings’, is regarded as suspicious at best and as professionally incompetent at worst.

To speak often in this lecture in an autobiographical and reflexive ‘voice’ is by far the best way to deal with my main themes. I am going to describe how (a) as an undergraduate student in Economics, I was inducted into the most unreflectively scientistic of the social science disciplines; (b) exposure to Marxist and radical critiques of that discipline led me to an intellectual rejection of it, but a rejection which simply reproduced its scientism in a somewhat different form; and (c) a largely accidental exposure to the later philosophy of Ludwig Wittgenstein helped me to reject both ‘orthodox’ and ‘radical’ forms of scientism, and to see both the social world, and my own activity, in a completely different way.

But before I get to that, a word about this lecture’s title. It is well-known that British Prime Minister Margaret Thatcher made the assertion ‘there is no such thing as society’ famous or infamous. She made the assertion in a speech in which she argued that notions of ‘society’ were used by what she called ‘left-wing ideologues’ to (a) deny or undermine the personal responsibility people should take for their actions and inactions (as in ‘it’s all society’s fault’) and (b) to deny to ‘enterprising’, ‘hard-working’ people the full fruits of their efforts (in the name of ‘equality’, ‘social justice’, etc.) I have some (just some) sympathy with point (a) but I have no sympathy at all with point (b).

In any event, my choosing her notorious assertion for the title of this lecture has (had) nothing to do with how ‘Mrs T’ (as the English often called her) used it politically. It has to do with the word ‘thing’ in the assertion. I want to say, in this lecture, that conceiving, not only ‘society’, but all political, social and economic phenomena as things is the distinguishing epistemological protocol of social scientism, of social science as a species of scientism. Because once you conceive anything as a thing, as an object or entity (even an ‘intellectually abstract’ object or entity), then, simply in order to distinguish it from other ‘things’, you are conceptually required to

1. 1. identify its location (‘where is it, and how does it relate to other things “around” it?’),

2. 2. identify its boundaries or limits (‘where does it end and other things begin?’),

3. 3. identify its defining characteristics or attributes (‘what makes it different from other things?’).

It is also a conceptual truth about things (i.e. a truth about our use of the word ‘thing’) that it is ontologically bi-modal. Things either exist or they don’t, and since it only makes sense to be ‘scientifically’ concerned with things that exist, conceiving the social world as a world of things is to conceive it as both complete and unchanging. So there ‘are’ things like ‘society’ or ‘societies’, but also like ‘prices’ and ‘markets’ and ‘states’ and ‘governments’ and ‘supply’ and ‘demand’ and ‘power’ and ‘poverty’ and ‘corporations’ and ‘individuals’ and ‘personalities’. In fact, there ‘are’ as many possible ‘things’ as there are nouns in the language one speaks, and there are as many actual things as there are factually correct assertions ‘applying’ those nouns to the world.

OK, but what’s wrong with that? Well, the most philosophically profound problem with it is the ontological bimodality itself. Because if things either exist or do not, then it is redundant (conceptually otiose, one might say) to ask how things come into existence. More profoundly still, it is out of the question even to ask whether a thing going through a process of formation or dissolution is ‘the same thing’ throughout that process. In fact, one of the most intellectually profound consequences of thinking about the world in this reified way is its marginalisation of the very notion of process and all related notions. So, in this world, not only do things exist or not exist, rather than coming into or going out of existence, but there are results or consequences of actions but no actions, there are outcomes of events but no events, ‘things’ have happened but nothing is happening. In short, a whole universe of human striving and failing, of conserving and changing, of acting, reacting and interacting, a whole time-located, time-taking process of actions and happenings, is ‘frozen’ into a conceptually fixed set of ‘end states’ or ‘states of affairs’. In fact, freezing processes into synchronic ‘states of affairs’ – into ‘things’ – is what allows them to be ‘scientificised’, given quantifiable ‘parameters’ and ‘measures’.¹

In short then, the point of this article is to agree with Mrs Thatcher, but in a way which she would have found both bewildering and irrelevant. Its point is to argue that there is indeed no such ‘thing’ as society and that there is a profound philosophical and moral price to be paid for conceiving ‘it’ that way. It is not however the price Mrs T had in mind, but the price of failing to see conceptualisation as a moral act and thus of failing to understand the full extent of one’s moral agency. To put it rhetorically, the price of reifying the social world is a failure to grasp the moral responsibilities of intellectuality, a failure to live the intellectual life adequately as a moral life. And one pays that price not only when one conceives society as a thing, but when one conceives ‘markets’ or ‘governments’ or ‘social classes’ or ‘exploitation’ or ‘oppression’ or ‘interest rates’ or ‘bureaucracies’, or a million other outcomes of human history and striving, as social ‘things’.

But that’s all for later, now for autobiography.

Intellectual autobiography in three parts

Economics

Between 1965 and 1968, I studied Economics as my major subject while an undergraduate at the University of Sheffield in northern England. Although I obtained a good joint honours degree in Economics and Politics and went on to graduate study, I spent a good part of my three undergraduate years in an emotionally conflicted state. In fact, by the time I graduated, I had become deeply hostile to my major subject and spent a lot of time trying to avoid the mainstream economic theory courses I had come to despise.

However, looking back 50 years, what strikes me is the disjunction between the depth of my youthful emotional and political hostility to orthodox economics and the shallowness of my intellectual critique of it. By this I mean that, from the day of my enrolment to the day of my graduation, I never seriously questioned the fundamental intellectual postulates of what I was being taught. I saw nothing wrong, that is to say, in perceiving ‘contemporary market economies’ (I think that was the term used) as composed of ‘producers’ and ‘consumers’ interacting in a market and producing a set of ‘prices’ of commodities by those interactions. It seemed equally reasonable to see ‘producers’ as requiring a particular price to cover their ‘marginal costs’ and for consumers to be willing to pay a price which reflected the ‘marginal utility’ of a commodity to them.

I was ‘no great shakes’ as mathematician or statistician, but I was good enough to understand and reproduce the graphs and equations in which these ‘logical relationships’ were expressed and to follow the logic of changing ‘demand’, ‘supply’ and ‘technological’ parameters in mathematical models.²

In fact, looking back, it is clear that my main objection to what I was being taught was not to its claimed ‘scientificity’ (a claim, or aspiration, which I wholly endorsed), or to its obsession with quantification as the prime expression of that scientificity, but simply to the failure of neo-classical economics to locate its economic actors socially. There certainly ‘were’ ‘producers’ who ‘supplied’ markets and there ‘were’ ‘consumers’ who demanded goods and services. But these two groups of supposed ‘economic actors’ were not at all on a social par. Rather the owners and managers of economic enterprises were a small and privileged minority of the population and the ‘consumers’ were mainly a much-less-privileged majority. And that privilege and disprivilege consisted not just in large disparities in wealth and assets, but in vast information disparities about costs and price formation. Hence this was no meeting of technically equivalent ‘actors’ seeking to maximise their utilities and preferences, but the massively unequal meeting of different social groups or classes producing market outcomes which are far from ‘equal’ for everybody. (At least so long as ‘equal’ was understood in ways beyond the substantively empty formulations of ‘Pareto optimality’.)

In line with this critique, my main objective in my doctoral studies was to locate economic processes in a social setting, and indeed to reconceptualise economic actors as unequal social groups rather than abstractly postulated (and implicitly equal) ‘individuals’. It hardly needs saying how well this social reorientation predisposed me to Marxian political economy (and to radical political economy generally) when I first encountered it. But before coming to that, one final, explicitly philosophical, point about my youthful education in neo-classical economics. I was implicitly taught economics as a species of realism. It was simply assumed that economic terms functioned as names of real things. ‘Producers’ and ‘consumers’ were real human beings, or groups of such beings; ‘firms’ and ‘enterprises’ were also names of groups of such beings. But not merely were economic actors real, so were the magnitudes which their activities produced. Thus, ‘production’ or ‘output’ was as real as producers, ‘consumption’ as real as consumers, ‘investment’ as real as investors and so on.

None of my economics teachers ever discussed the epistemological status of the categories they were using, and I have no memory of any student (including myself) enquiring about that. Moreover, I am certain that had anyone asked, ‘but is consumption real in the way consumers are?’ they would have been laughed out of court (or classroom). In short, I was taught economics as realism not in contrast to ‘idealism’ or ‘constructivism’, or any other ‘ism’ but as the only possible option. Of course ‘investment’ is the name of ‘a real thing’. What else could it be? And why would anyone teach or learn about it if it wasn’t?

This unreflective realism³ also formed my understanding of what a ‘radical’ critique of economics entailed; what intellectual attributes it would have to possess. Around about 1969 or 1970 I was looking for a social-scientific species of realism that would have socially situated individuals and groups as economic and political actors. At this point, and right on cue, enter one Louis Althusser.

Althusser

In the early 1970s, I read English translations of Pour Marx (1965) and Lire le Capital (1968) and was, as they say, ‘blown away’. But again, in retrospect, it is significant what it was in Althusser (or rather in Althusser’s account of Marx) that did ‘blow me away’. It was the extension and deepening of the realism to which I was already committed, that I had taken in holus-bolus with my neo-classical economics. For Althusser’s Marx told me that the realism of neo-classical economics was merely ‘apparent’. The ‘real things’ economics taught about – markets, prices, production, consumption, investment and so on – were not real at all, they were the mere ‘forms of appearance’ of underlying ‘value relations’ and ‘class relations’. Neo-classical economics, it was now revealed to me, was an intellectual discourse in which forms of class exploitation appeared as formally equal relations of monetary exchange and were thus legitimised and naturalised. In fact, neo-classical economics (along with the rest of ‘bourgeois’ social science) was an ideology. It was an ideology (according to Althusser’s Marx) because through its categories and explanations, forms of class exploitation were redescribed in ways that rendered them invisible both to their victims and their perpetrators, and ‘naturalised’ them as ‘market relations’ – as the only ‘efficient’ way of organising modern economies.

But this was not the most philosophically significant part of Althusser’s account of Marx for me. That lay in his claim that what I had been taught as a species of realism was actually a form of ‘empiricism’ (a new ‘ism’ to which I was now introduced). This shallow ‘empiricism’ took the naively perceptible ‘forms of appearance’ of the world as the only reality and failed to grasp the ‘really real’ (as one might say) ‘deep structures’ of class and value relations.

I have to say, at this point, that because I possessed no formal philosophical education at all, I had no clear idea what ‘empiricism’ actually was, or how, precisely, it differed from what I now learned to call ‘structuralist realism’. For example, had anybody suggested to me that neo-classical economics was not a realism at all but a form of rationalism, I would have been completely bewildered. I could not have begun to say how empiricism differed from realism, or either of them from rationalism, and I certainly had no inkling what a variety of subtly different orientations to the world might be advanced and referred to using these blunderbuss terms.

But in any case, I had no time to explore any of these issues in depth, or even superficially, because I faced an altogether more pressing task. I had to write an Oxford doctorate in development economics!

Now the difficulty with intellectual autobiography is to distinguish those parts of it which reflect a broader zeitgeist (and which will thus be echoed in the experiences of others) from those parts which are merely idiosyncratic and personal, and thus of no interest to anyone but oneself. On the whole, the story of my doctoral research (among peasant coffee farmers in northern Tanzania) and of the book Class and Economic Change in Kenya (Kitching, 1980) which finally emerged from that fieldwork and further study in Kenya, belongs in the latter category, so are of no concern here.

In fact, the only point of philosophical significance in nearly a decade of researching and writing the book is that I found myself writing a long-period history of agricultural development in colonial and post-colonial Kenya using an Althusserian ‘structuralist’ framework. I took the historical narrative of Class and Economic Change in Kenya to be an ‘empirical’ narrative describing the ‘forms of appearance’ of Kenyan political economy. The underlying determinants of these ‘forms of appearance’ were ‘revealed’ in the concluding ‘theoretical’ section of the book devoted to a ‘structural analysis’ of these forms.

Sixteen years ago, I published a description of the insoluble conceptual puzzles into which this attempt to write ‘structuralist economic history’ led me. Since it is just as valid today as it was then, I have simply reproduced a long section of it below. Before doing so though, I should say that in 1981, Class and Economic Change in Kenya won the Herskovitz Award of the US African Studies Association and was widely reviewed and praised both for the quality of its historical research and its ‘theoretical rigour’. The problem for me, however, was that, at the very time the book was being lauded, I was having increasing doubts about its entire epistemology. Without wishing to be overly presumptuous, I suppose it was rather as if, in about 1934 or 1935, Wittgenstein were to have received a major prize for the Tractatus!

At any rate, here is a shortened version of what I wrote in 2003:

One of the aims of the book was to create a ‘seamless web’ in which theoretical concepts and historical and empirical description would be tightly woven in the text [. . .]

In this objective I failed abysmally. In fact, so total was the failure that I did the only decent thing any social scientist could do in the circumstances. I extolled failure as a virtue. In the introduction to the book I explained that it had a ‘descriptive’ or ‘historical’ part (dealing with the development of the Kenyan political economy from 1905 to 1970) and a ‘theoretical part’ (the last three chapters) which, I said, discoursed on the ‘theoretical implications’ of the historical narrative.

The only thing that caused me to acknowledge this division [. . .] was the language of the book. I noticed that in the last three theoretical chapters [. . .] the language I was employing was in some way different. However hard I tried, the sentences and paragraphs there just read and sounded different from the sentences and paragraphs of the previous thirteen chapters. By the time the book was published I had managed to identify two of these differences more precisely. They were:

1. 1. A marked increase of passive forms of the verb in place of transitive or active forms. Things tended to ‘happen’, ‘be related to’, ‘have consequences for’, rather than being done by active subjects.

2. 2. A loss of any clear context of time or place . . . both chronological and spatial ‘marker’ terms almost entirely disappeared in the final three chapters. In the familiar theoretical pattern, the text, although nominally still concerned with a particular state (Kenya) at a particular time, manifested a timeless, placeless, universalist tone . . . Indeed I turned necessity into virtue by claiming explicitly in the book’s introduction that ‘theory is timeless and proceeds in a sort of eternal present’ and that the ‘theoretical discourse’ was a ‘recapitulation’ of the historical narrative from ‘a theoretical perspective which essentially situates itself at the end of the process described in the narrative and looks at certain logical and structural relationships within the economy and society at that point’.

. . . [W]ith the book published I worried more and more about this ‘linguistic’ issue, and the more I worried the more confused I became . . . It became clear to me that there were several features of the book that I could . . . not explain at all. These included:

1. (a) Why the language of the book had not been obedient to my will. Why, no matter how hard I tried, could I not construct my ‘seamless web’ of narrative and theory? The Althusserian notion of ‘levels of abstraction’ consoled me for a while, but the more I thought about that, the more unclear I became about;

2. (b) What such ‘levels of abstraction’ actually were, and indeed about what one does when one abstracts. For example, a clear sense of time or place had apparently ‘been abstracted’’ from the theoretical discourse. But when I looked at my final chapters closely it was clear that all I had done was to linguistically contract or collapse time and space. I had presented certain concepts and processes as ‘timeless’ when, on reflection, they clearly were not. This was especially true of my ‘theorisation’ of capital accumulation in Kenya which I had only been able to present as structurally ‘fixed’ because it was both a repetitive cyclical process occurring throughout the colonial period and expanding in scale through those cycles. In any case;

3. (c) By the time I had reached this point in self-questioning I realised that I simply could not justify calling one part of the narrative ‘historical’, ‘empirical’, ‘descriptive’ or ‘concrete’ and the other part ‘theoretical’, ‘abstract’ or ‘analytical’ except in terms of my changing use of language . . . I had no clear idea of the meaning of any of these terms, though I had been happily passing them off on myself and my readers. In short, I was completely conceptually puzzled. I had wanted to construct a ‘seamless web’ but my own use of language had prevented me. But now I could not see why, apart from the language, I had not been able to do it.

I should be clear. I had not decided that there were no differences between ‘things’ like ‘history’ and ‘theory’, or ‘concreteness’ and ‘abstraction’, or ‘description’ and ‘analysis’. In fact, I continued to think that if they were not different ‘things’ they would not have different names – have different words ‘applied’ to them. It was just that I could not satisfactorily formulate to myself what these differences were. I could not formulate a distinction between ‘abstraction’ and ‘concreteness’ (or ‘the abstract’ and ‘the concrete’) which was not either so vague as to be meaningless, or which, when made more exact, turned out to be false or partial. I could not formulate a distinction between ‘theory’ and ‘description’ which did not either produce crude empiricist formulations re. ‘facts’, ‘sense data’ and so on to make description ‘conceptually’ distinct, or end up subsuming description to theory (in which case – round in a circle – why do we have two words?). (Kitching, 2003: 9–12)

At this point, enter Ludwig Wittgenstein.

Wittgenstein

Before continuing my intellectual biography with the story of my encounter with Wittgenstein’s Philosophical Investigations (1972) (hereafter PI) – and as a necessary prelude to that – a few more remarks re zeitgeist typicality and personal idiosyncracy.

My movement from neo-classical economics to Marxian political economy is not particularly unusual. In fact, neo-classical economics has long been renowned for the volume of intellectual ‘renegades’ or ‘heretics’ it produces. It has often been observed that the more distinguished any graduate student of economics, at least in the Anglo-Saxon world, the more likely it is that she or he will end up being critical, to a greater or lesser degree, of the discipline. Of course, not all such economics sceptics end up as Marxists (many become Keynesians or Kaleckians or so-called ‘neo-Ricardians’, for example) but even so, they usually manifest some Marxian or socialistic influences in their thinking.

My movement from Marxian political economy to Wittgenstein is unusual and idiosyncratic however and it owed a lot to the philosophical form of Marxism which I had previously adopted. Because, while Althusser changed my conception of ‘the real’, he only deepened my commitment to the realist epistemology which I had imbibed with my economics training. Louis Althusser, as much as Paul Samuelson or Richard Lipsey or Michael Todaro endorsed a picture of language in which concepts functioned as names of things or objects. Of course, all these thinkers grasped, and I grasped (vaguely), that there are different kinds of things and (therefore) different kinds of names (some more ‘concrete’ or ‘empirical’, some more ‘abstract’ or ‘analytical’). But, in some sense or other, all concepts were names, and a concept was ‘the right’ concept if its definition ‘mirrored’, ‘reflected’ or ‘matched’ the ‘attributes’ or ‘characteristics’ of ‘the thing’ it named.

So with that preface, back to autobiography, and Wittgenstein – my liberator from both realism and scientism!

The serendipitous way in which I acquired a copy of the PI (as an Oxford graduate student) is again personally idiosyncratic. But it lay on my book shelves gathering dust for some 7 or 8 years, because when I first acquired it (in 1972 or so) I found it utterly incomprehensible and indeed infuriating. No orientating introduction or preface, no chapters, no overarching argument or thesis, or none that I could see. Just two very arbitrary ‘Parts’, and a host of numbered ‘remarks’ which had no obvious logic or purpose. To the best of my recollection, I stopped reading after about seven or eight pages and dumped what I had been told (by the person who gifted it to me) was ‘the greatest work of philosophy of the twentieth century’ onto a bookshelf with a contemptuous snort.

Seven or eight years later, however, drowning in the conceptual puzzlement of Class and Economic Change, I picked it up off another shelf, blew off the dust, and began to read. I only did this because I intuited that I needed to know something about language, and the PI was the only book I possessed about language.

I was utterly transfixed from page 1. I mean literally from page 1, from the opening quotation from Augustine’s Confessions and Wittgenstein’s (1972) comments on it:

These words, it seems to me, give us a particular picture of the essence of human language. It is this: the individual words in language name objects – sentences are combinations of such names. In this picture of language we find the roots of the following idea: Every word has a meaning. This meaning is correlated with the word. It is the object for which the word stands. Augustine does not speak of there being any difference between kinds of word. If you describe the learning of language in this way, you are, I believe, thinking primarily of nouns like ‘table’, ‘chair’, ‘bread’ and of people’s names, and only secondarily of the names of certain activities and properties; and of the remaining kinds of word as something that will take care of itself. (p. 2e)

One way of reading the PI is to see it as a sustained undermining of this ‘picture of the essence of human language’ – to show it to be wrong, but ‘deeply’– that is, subtly and profoundly, not obviously or stupidly wrong. In fact, as Wittgenstein says, it is not so much wrong as ‘misleading’ – deeply misleading – which is indeed how and why it can ‘bewitch the intelligence’ even of thoughtful and reflective people. It is misleading rather than wrong, because it is not wrong to say that the nouns in language ‘name objects’ or that nouns, as kinds of word, ‘stand for objects’. But this observation can mislead, if generalised, in at least three different ways:

1. 1. One fails to see that there are different kinds of ‘objects’ and that (therefore) nouns can ‘stand for’ them in different ways – that is, the words ‘chair’ or ‘apple’ ‘stand for’ objects, in a very different way from the way the words ‘energy’ or ‘utility’ ‘stand for’ objects. In fact, in the latter cases, both the phrase ‘stands for’ and the noun ‘object’ obscure far more than they reveal.

2. 2. One uses the model of ‘noun + object’ as the ‘paradigm case’ of meaning in language. So, either one thinks that in some way all words ‘stand for objects’ or (and probably more commonly) that all other words simply serve to facilitate the noun–object relation. In fact, one thinks that this is the relation through which language meets the world, through which language gives the world meaning.

3. 3. One fails to see that we use nouns as names, but we also use them in a host of ‘non-naming’ ways. As Wittgenstein says, ‘think how singular it is to use a man’s name to call him’. Because, clearly, one can also use a man’s name to distinguish him from other men (and women), to praise him, to insult him, to make a joke at his expense, to enter him on a tax roll and so on.

By the time I finished reading the PI for the first time, I concluded that I had spent the first 15 years of my academic career bewitched by, and indeed imprisoned in, what Baker and Hacker (1980) subsequently taught me to call ‘the Augustinian picture of language’, and that I had been philosophically misled by it in the three ways above and probably others as well. But I also took the book as showing that Wittgenstein himself had spent the first 25 or so years of his career imprisoned in it.

Indeed, although a training in economics (at least mainstream economics) is a training in the most philosophically gauche and naïve of all the social sciences, it did have one advantage for my reading of Wittgenstein. It enabled me to see very clearly ‘where he was coming from’ in writing his great book. Because I had not the slightest doubt – again from page 1 – that the PI represented a reaction against what I would now call ‘scientism’, both scientism in philosophy and as an approach to knowledge generally.

I say this, because there is something of a tendency in contemporary Wittgensteinian scholarship to downplay, or ignore, this aspect of the book – the worldview against which it is pitted. Wittgenstein was first trained as a mechanical engineer (and in mathematics as an engineering tool) and his first interest in philosophy was in Frege and Russell’s philosophies of mathematics. And in my view (now also unfashionable), the Tractatus Logico-Philosophicus states and laments the limits of, what amounts to a logically rigorous, ‘analytical’⁴ version of ‘the Augustinian picture’.

‘Anything that can be thought at all can be thought clearly. Anything that can be put into words can be put clearly’ says the Tractatus. And the task of that book is to show what can be ‘thought at all’ and how to put it ‘clearly’. Broadly, what can be thought at all are factual possibilities, and they can be put clearly into words by tightly and rigorously defining the words (which means primarily the nouns) which we use to express them. The proposition ‘The cat is on the mat’, to be truth-tested, requires first, clear definitions of the nouns ‘cat’ and ‘mat’ and an equally clear definition of the spatial ‘relationship’ (another noun, note) in which they stand.

But there are also ‘things’, says the Tractatus, that either cannot be thought at all (although they can, perhaps, be felt) or which, at any rate, cannot be thought ‘clearly’. And the tragedy of the human condition is that these are the most important things in life.

So, we can make statements about the world which have both meaning and truth (factual statements) and statements about the world which have meaning (statements of value, whether moral, aesthetic or political) but no testable truth. And for the younger Wittgenstein, the former statements are practically essential but ultimately unimportant, and the latter may be felt as ultimately important but are pragmatically useless, because they are neither true nor false. A gloomy worldview indeed!

But it is a worldview (and a gloom) that only gets its purchase from a highly restricted conception of what is required to ‘think clearly’ with words and to make ‘clear’ statements with them – the conception that words name objects. If there are no clear objects that words name (no rigorous definition of them to be given), then they don’t mean anything (or anything ‘clear’), and the statements they compose certainly cannot be truth-tested. So, escape this conception – this ‘scientistic’ conception of meaning – and you escape the worldview. Whether you also escape the gloom, by seeing statements of value as something more than expressions of feeling, probably depends as much on personality as anything else, and Wittgenstein was a very gloomy man. But, at the very least, when language is seen far more expansively than it is in the Tractatus, the range of possibilities for both linguistic and non-linguistic action also expands enormously. The world conjured by the PI certainly has more room for optimism (because for more types of knowledge and truth) than the tight, scientistic ‘fact-grinding’ prison that is the Tractatus.

But I digress – I mean autobiographically. We left me, you may remember, reading the PI for the first time in 1980 or so, and realising that the book was a sustained and deep critique of a worldview which I instantly recognised because I had shared it for the previous 15 years. And this reading of the PI – of its central raison d’etre – has not been altered by any of the commentary literature on Wittgenstein I have read in the subsequent 35+ years. If you do not fully grasp the Investigation’s scientistic enemy, you do not grasp the Investigations in my view. Certainly, you will not fully appreciate, or enjoy, Wittgenstein’s beautifully crafted escape tunnel if you have never been in his prison. Even more importantly, it will be difficult for you to understand how hard it is to entirely escape this prison, even when you think you have.

Here is an astounding example of what I mean. At the beginning of remark 242 of the PI, Wittgenstein says, ‘If language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgments’. Like virtually everything else in the text, 242 has been the subject of considerable secondary exegesis and discussion. Their main focus has usually been on what Wittgenstein means by ‘agreement in judgments’ and whether ‘agreement’ is the right word here, or in the preceding remark, 241 (in which he speaks, famously, about human beings ‘agreeing’ in ‘the language they use’ and in ‘form of life’).

But never mind that. Just consider the primary assertion in this first sentence itself; ‘if language is to be a means of communication there must be agreement . . . in definitions . . .’ Surely not! And certainly not if he means by this the entirety of any natural language. Does everybody who communicates in Spanish (or English or German or Mandarin) ‘agree’ in the definitions of the words they use – all of them? The simplest sample survey of ‘fluent’ ‘native’ speakers of any natural language shows that they cannot so much as proffer lexicographical definitions of many of the words they use, let alone ‘agree’ on those definitions. And the PI itself is a brilliant demonstration that they could not acquire their first language through so-called ‘ostensive definition’ either, at least not in abstraction from a large number of other practices.

So, Wittgenstein has perpetuated here, simply in passing and probably because his attention was elsewhere (on ‘agreement in judgments’, etc.), the fundamental mistake of the prisoner of the ‘Augustinian picture’. He has generalised across ‘language’ an observation which is accurate for some specialised, technical ‘languages’ or ‘regions’ of language (some parts of Physics, some parts of Mathematics, some parts of Economics), but massively inaccurate for any natural language as a whole. He has manifested a ‘residual scientism’ here, and indeed in the rest of 242 (with its assertions about ‘seeming’ to ‘abolish logic’ and the immediately following sentences re ‘measurement’). He has come out of the escape tunnel with scraps of his prison clothes still hanging about him.

But however hard a scientistic picture of language may be to escape when one has acquired it through some types of technical and scientific training, Wittgenstein did the best job that anybody has ever done in escaping it himself, showing others how to escape it, and demonstrating how fundamentally we misunderstand the human ‘form of life’ so long as we remain ‘in thrall’ to it.

Three liberating insights

My first reading of the PI (the only philosophy book I had ever read at that time) ended my commitment to realism. Or rather it showed me that what I had understood as ‘realism’ depended on a very particular and narrow conception of how words obtain meaning. And actually, it showed me that my ‘realism’ had been hopelessly naïve in other ways. Not only had I lacked any grasp on the underlying epistemological assumptions of my ‘realist’ economics, I had also not understood the variety of different epistemologies which march under the banner of ‘realism’ or their disputed relations to other philosophical ‘isms’ (notably ‘empiricism’ and ‘rationalism’.) As a result, in moving from economics to Althusserian Marxism, I had swapped a partly empiricist and partly rationalistic realism for a rationalistic but militantly anti-empiricist realism without a proper philosophical grasp either on what I had abandoned or what I had embraced. No wonder I was confused!

But the PI, and the broader reading on Wittgenstein and in Wittgensteinian philosophy that followed it, not only showed me what a mess I had been in, but how to extricate myself. They gave me positive, liberating, alternatives to the scientism to which I had been prisoner, and it is with these that I end this lecture. There were three of them: (a) an appreciation of the historicity of concepts and its importance, (b) an understanding that language use is an action or rather a myriad of actions and (c) the moral significance of that latter fact. I will take them in order.

The historicity of concepts

Having read the PI, I immediately grasped the fundamental problem at the heart of Class and Economic Change in Kenya. In the first three-quarters of the book (devoted to economic history), I had described and explained how the economic practices of the African population of Kenya had changed over some 65 years of colonial and post-colonial history. In this ‘historical’ section I had, quite unselfconsciously, allowed the meaning of the words I used to change as the practices they described changed. But when I came to write the concluding ‘theoretical’ quarter of the book, I reverted, very self-consciously, in a definition-heavy prose, to a historically static, ‘essentialist’ conception of word-meaning. My whole theoretical analysis was constructed on the premise that while peasant agricultural production may have changed over time, the essential meaning of the concept ‘production’ had not; while ‘the details’ of capital accumulation in Kenya may have changed over time, the essential meaning of the concept ‘capital accumulation’ had not; while the Kenyan state may have changed over time, the essential meaning of the concept ‘state’ had not. Then, in best Althusserian style, I construed these essentialist definitions of concepts as ‘real’, ‘deep’ ‘structures’ which somehow drove the ‘apparent’ historical changes, while remaining entirely unaffected by them (a kind of Marxicised version of Aquinas’s ‘first cause’ theology!).

But this, I now realised, was not merely an arbitrary division, resulting from my essentialist refusal to allow ‘theoretical’ concepts to change. It embodied an approach to meaning which, if generally or universally true, would not merely make writing history impossible (which it had almost done in my book), it would make history itself impossible.

One major thrust of the PI is to show how all concepts are open to what Wittgenstein calls ‘extension’ or ‘projection’ of their meanings as human activities, human practices, change (which means of course change in time). Wittgenstein’s famous notion of the ‘family resemblance’ of concepts, and his analogy of the extension of concepts to the spinning of fibres into threads, are used implicitly or explicitly to elucidate the praxis-related historical mutability of concepts, and the necessity of this to be so if, in Marx’s famous words, ‘men’ (people) are to ‘make their own history’. One of the ways people make history is to change their descriptions and explanations of what they are doing, as they change their doings and indeed as part of changing those doings. And they alter or ‘project’ their concepts accordingly as they make these changes.⁵ Thus, there is no ‘essential’ ‘structural’ or ‘analytical’ meaning of ‘production’, or ‘consumption’ or ‘investment’ or ‘supply’, or ‘demand’, or ‘class’ or ‘mode of production’. If there were, people could not change what and how they produce, what and how they consume, in what and how they invest, what they demand and how they demand it, what they supply and how they supply it, how they oppress and how they resist oppression. Of course, one can stipulate ‘exact’ or ‘fixed’ meanings for all these terms if one wishes. But then that is precisely what one is doing, one is stipulating meanings (another action), and one can be asked why one is doing that, what purpose is served by those stipulations.⁶

Language use as an action

But all that, though important, was not the most important philosophical lesson I learned from Wittgenstein. For note how the paragraphs above are expressed, I realised what I had done in the book. Or rather more accurately I realised – fully realised one might say – that I had indeed done the book – written it and written it in specific ways. In particular, I realised that I had used words in different ways, in the ‘historical’ part, in the ‘theoretical’ part. And that was another revelation, because one of the effects of the kind of scientistic training I had received was that I acted as if – behaved professionally as if – words chose me, rather than me choosing words, as if concepts chose me, rather than me choosing concepts. To put that more exactly, one of the hallmarks of a scientistic intellectual immersion is that one acts as though the world somehow dictates to one what words to use to describe and explain it. In my case, this meant that something called ‘economic reality’ determined the right words to use to describe it/explain ‘it’. In fact, I now realised, that was what I had understood ‘realism’ to be. Realism was – to be scientifically ‘realistic’ was – to accurately ‘reflect’ or ‘picture’ the world in concepts (literary and mathematical), with that picturing being both active and passive at the same time rather in the way that on one mistaken conception, photography is. One ‘takes’ a photograph with a camera (an action), but what one ‘takes’ is determined, supposedly but mistakenly, by the photographed object or objects themselves. And photography is therefore, ‘essentially’, a passive receiving from reality.

Significantly, it was only from this time that I began to think of myself, professionally, as a writer or author. Before that I had considered myself a ‘social scientist’ and ‘economist’ but not a writer. And I find this very typical, indeed it can be used as a kind of criterion or test. Ask any social scientist, of any discipline, whether they are an ‘author’ or a ‘writer’. I observe and hypothesise that the more reluctant they are to embrace this identity, the more likely they are to be of a strongly positivist, empiricist or realist persuasion, and the more addicted to what John Dewey famously called ‘the Spectator Conception of Knowledge’ – that is, to an essentially passive, ‘data-receiving’ conception which somehow obviates any ‘authorial’ role or responsibility.

The Harvard philosopher Stanley Cavell (1969), was, in my opinion, one of the greatest, if not the greatest, of Wittgenstein’s contemporary followers. The following is a quotation from his book The Claim of Reason:

Perhaps one feels: ‘What difference does it make that no one would have said, without a special reason for saying it, that you knew the green jar was on the desk? You did know it, it’s true to say that you knew it. Are you suggesting that one sometimes cannot say what is true?’ What I am suggesting is that ‘Because it is true’, is not a reason or a basis for saying anything, it does not constitute the point of your saying something, and I am suggesting that there must . . . be reasons for what you say, or be point in your saying of something, if what you say is to be comprehensible. We can understand what the words mean apart from understanding why you say them; but apart from understanding the point of your saying them we cannot understand what you mean. (pp. 205–206)

Once someone has told you that ‘because it is true’ is not ‘a reason . . . for saying anything’, it seems self-evident, something one would never dream of denying. And even at the height of my scientism, I would never have thought that stating facts or stating ‘theoretical truths’ was an end in itself. But one of the most pervasive aspects of scientism in social studies is the transfer, from natural science, of the theory-application model. Theoretical truths about ‘how an economy’ (or a society) ‘works’ are first ‘discovered’ and then (a subsequent step) ‘applied’ to the solution of economic or social problems. So, it is not that one forgets, or overlooks, that the telling of truths must have a point or purpose. It is rather that one never asks why or how one chooses the particular theoretical (or factual) truths one chooses to tell before their ‘application’. If one wants to understand ‘how an economy works’ (a dubiously mechanistic formulation in itself), why start with assertions about monetary demand and supply? If one wants to understand ‘how a society works’ (ditto), why start with its occupational structure, or gender structure, or age structure, or class structure?

There are answers one can give to all these questions. But in most social science teaching, one is never encouraged to ask them (indeed there is usually no recognition that they even could be asked), so one’s attention is never so much as turned to what the answers might be.

But once one’s attention is turned (as Wittgenstein and Cavell turn it), it is immediately evident that the answer cannot be something manifestly question-begging like ‘you start with supply and demand because supply and demand is the most important thing in any economy’ or ‘you start with the class structure because the class structure is the most important thing in any society’. For the moment, such answers are proffered, more questions immediately tumble forth. ‘Important in what respect? Important to whom? Important for what purpose, to do what?’

Speaking or writing truths – just like asking questions, or describing events or physical things, or telling a joke or a lie, or intervening in an argument – are actions, linguistic actions. And therefore, like all actions they must have a point or purpose. (Indeed, this is analytically true. It is part of the definition of an action that it has a point or purpose.) This in turn means that it is important not merely what truth one states (its ‘content’ as it were) or what question one asks, or what joke or lie one tells, but when and to whom one tells them – the ‘context’, in a word – in which one tells them. For that context, as much as the words in the statement, question or lie, determines what it/they mean. Or to be precise, and exactly as Cavell says, the context in which you say anything makes it possible to know, not just what your words mean but what you mean when you use them (then and there, as it were).

The moral significance of language use as an action

It isn’t, then, an acceptable answer to why one should ‘start’ the study of Economics with supply and demand to say that it’s ‘the best’ way to start or that these are the ‘most important’ economic activities. It isn’t an acceptable answer to why one should ‘start’ the study of Sociology with social roles or the class structure that these are the ‘most important’ social phenomena. It isn’t an acceptable answer to why one should ‘start’ the study of Psychology with the study of emotions, that emotions are ‘the most important’ or ‘the most basic’ psychological phenomena. In fact, no answers to these questions are acceptable which focus simply on the ‘facts’ or ‘data’ – or alleged ‘facts’ or ‘data’ – to be studied. Rather, the only non-question-begging answers must refer to the objectives or purposes of students and teachers as well as to what is to be studied.

All such answers are therefore ‘objective or action conditional’. ‘Well if you want to know how goods and services get distributed amongst a population the best place to start is . . .’, ‘Well if you want to know how inequalities arise and are maintained among human beings the best place to start is . . .’ ‘Well if you want to know why people act irrationally, and even when they know they are acting irrationally, the best place to start is . . .’

And such conditionals themselves invite a further ‘clarifying’ question. ‘So, what do you want to know?’ Because, long before we become students of any social science, the mere living of our lives poses questions to us, generates puzzles, difficulties or problems, to which we want to find answers. ‘Why do my parents behave that way?’ ‘Why is Paul’s school so different from mine?’ ‘Why do company directors get paid so much?’ ‘Why is it only immigrants who do these jobs?’ So that, whether we are aware of it or not, we come to the study of any social science already wanting to know things. Any student can answer the question ‘what do you want to know?’ if a teacher puts it. But so few teachers ever do. And that often means that beginning students of all types of social science are not aware that there are things (disciplinary-relevant things) they want to know, and already have ideas about, because they are not asked the question that would make them aware.

This then is an example of the asking of a question (‘What do you want to know?’) raising awareness in the people (the students) being asked. But in other contexts, asking them questions can frighten people. In other contexts, telling them a fact can frighten, (or delight, or puzzle, or enlighten) people. In yet other contexts, expressing surprise (‘Really?’) can dismay, or surprise, or bore, or exasperate, people. And in yet other contexts, explaining something can surprise or bore (who didn’t know that?) or alarm, or terrify, or enlighten the person or people to whom you explain it.

So, in certain contexts, many contexts, asking questions, stating facts and offering explanations can have effects – emotional effects, psychological effects, political effects – beyond simply eliciting answers, providing information and solving puzzles. Therefore, in certain contexts, many contexts, what you choose to say, when you choose to say it, and to whom you choose to say it, are not merely acts but morally charged acts. (‘Should I say that? Will it frighten him? Will it depress her even more? Perhaps it will cheer him up?’) And in certain contexts, in many contexts, not merely is speaking a morally charged act, so is keeping silent (‘I should have told her what I knew then. It would have made so much difference!’). It can matter morally, not just what you say or to whom you say it, but when you say it (or don’t).

So an exposure to Wittgenstein’s extraordinary later philosophy has taught me that studying social science itself is an action and is thus driven as much by the purposes of the student as by the ‘subject matter’ being studied. But more than that, it has taught me that studying social science, as a practice or set of practices, is not ontologically different from living in society. In living among our fellow human beings, we ask questions, supply information, offer explanations, provide comfort, allocate blame, attempt to persuade or convince, express affection and demonstrate technical expertise. And we do all that – or rather we should do all that – in social science too. In doing social science, we are not acquiring knowledge in order to ‘apply’ it to the solution of problems (although that may be what natural scientists do, it is only an illusory scientistic description of social scientific practice). Rather, we are learning to address, to speak to, our fellow citizens in order to elicit certain responses in them (responses which we think will be right or good in some way). And what does that mean? It means what it ‘always’ means; asking them questions, providing them information, suggesting explanations, being persuasive, expressing surprise and so on. And in social science too, when, where and for whom you write, and when, where and to whom you speak, determines what others will take you to mean in that speech and writing. What a social science text means is never simply a matter of what is ‘in’ it. And in social science too, it can be a moral act to keep silent – to not speak or write.

Conclusion

I am not the only person to have arrived at these sorts of conclusions. Apart from Wittgenstein himself, a variety of scholars in the English-speaking world – Peter Winch, Hanna Pitkin, Rupert Read, Joel Backström, Michael Billig, for example – have argued in very similar ways from Wittgensteinian premises. On the whole though, they – we – have had little or no success in changing the way the vast majority of social scientists go about their business.

The reason for our failure is not far to seek. In arguing that natural science is not an appropriate model on which to base the study of human activity, we are arguing against a deeply institutionally entrenched set of practices and self-interests, not simply against a ‘philosophy’. In scientistic societies, the jobs, incomes, professional status and promotion of academics (and of many other people) depend on their pursuing social science research in the approved ‘scientistic’ manner. Indeed, as Michael Billig (2013) has brilliantly demonstrated, even research in applied linguistics critical of academic language and its shortcomings, uses precisely the same ‘nominalistic’ (i.e. abstract-noun-heavy) prose it is criticising!⁷

Certainly, any argument to the effect that social scientific research produces descriptions and explanations of human social life less precise and less insightful than the ‘ordinary’ descriptions and explanations employed as part of that life is met with incredulity by the vast majority of social scientists.⁸ And that includes those who are ‘critical’ or ‘radical’ economists, political scientists, sociologists or social psychologists, as well as their more orthodox colleagues. It is not difficult to understand that incredulity. For while the scientific methods which have achieved such success in the study of the natural world may have to be modified somewhat to be applied to the social world, they surely cannot be so out of place there that they tell people nothing more than they knew about each other’s motives and behaviour before social scientists began researching them!

That is probably right. The social science which has been carried out in the last 150 or so years has taught human beings some things about themselves they did not know before. It has not produced absolutely nothing. But given the amount of time, money and fine human brains it has cost, its genuine discoveries have surely been minimal in comparison with the achievements of natural science in the same period.

And that leads me to the tragic paradox with which I end this article. All forms of social science have been a failure in comparison with natural science. But it is precisely the blinding light of natural scientific achievement – and the resulting emergence of ‘scientism’ as the defining ideology of knowledge in all modern societies – that prevents social scientists seeing their failure clearly and investigating its causes properly. Fascinated and transfixed by the success of natural science, social scientists cannot see that their failure is at bottom philosophical – a failure to properly understand how human beings, uniquely, live both as part of and apart from, the natural world and how that affects what it means to study them. But if the failure of social scientists is philosophical – a failure to see natural science, human social life and their own activity clearly – its rectification must also be philosophical. For it is, in the end, a failure of will (as Wittgenstein said) not merely of technique or method.