Epistemology,
observed particulars and providentialist assumptions: the fact in the
history of political economy
An essay-review of:
Mary Poovey
A History of the Modern Fact. Problems of Knowledge in the Sciences of Wealth and Society
(Chicago and London: The University of Chicago Press, 1998), xxv + 419 pp.
ISBN: 0-226-67525-4 Cloth US$49.00 £39.25; 0-226-67526-2 Paper US$17.00 £13.95.
For Studies in History and
Philosophy of Science
4 March 1999
Andy Denis, BA, MA, MSc
Lecturer in Economics
Department of Economics
School of Social and Human Sciences
City University
Northampton Square
LONDON EC1V 0HB
Telephone: 020-7040 0257
FAX: 020-7040 8580
e-mail: a.denis at city.ac.uk.
A History of the Modern Fact by Mary Poovey, professor of English and director of the Institute for the History of the Production of Knowledge at New York University, is an ‘epistemological history’. It is an attempt to chart the development of conventions and beliefs about how we understand the world in the history of economic thought. The questions it addresses are extraordinarily productive leading to an account so multi-layered and multi-dimensional that a review simply cannot do justice to its richness, but is necessarily flat, partial and one-sided. I will try to give a flavour of its contents before moving on to an aspect of the book that was of particular interest to me: the place of Adam Smith in the ‘Scottish Enlightenment’. This structure reflects my response to Poovey’s contribution. I have no problem with her general aims and approach, and the overall account of those aspects of the history and prehistory of economics which she touches on is enlightening, stimulating and subtle. However, the purview of the work is vast and daunting. Anyone who embarks on such a project is going to find themselves treading on the toes of specialists in the topics reviewed at every turn. Wherever there is controversy, one or more sides will find her reading unsatisfactory. In my case it is the characterisation of Smith’s general intellectual standpoint which fails to satisfy. This is not to decry the work as a whole: rather, we should - with Poovey herself - regard the History as a work in progress, as a research programme, as a provocation to further work adopting her methods, which will need to start with some correction and much filling-in of details.
1 Ancient, modern
and postmodern facts
First we need to sort out what Poovey means by the modern fact. By ‘modern’ she appears to understand roughly the three hundred years from the seventeenth to the nineteenth centuries. The ‘modern fact’, which in her opinion characterises this period, is a unit of knowledge understood in a particular way. General, or systematic, knowledge - theory - is separate from but based on observed particulars, facts. This epistemological stance is vulnerable to ‘the problem of induction’: theory requires that unobserved particulars are predictable from those already observed - but how can we know that before we observe them? Poovey’s contention is that each of the instances or stages of development of the modern fact she documents turns out to be an attempt to resolve the dilemma this raises for each of the knowledge projects she examines: how to convince one’s audience that the knowledge one claims is well-founded. The result is that the modern fact is something inherently ambiguous: is it an incontrovertible datum which merely shows what is true, or is it a product of all the theoretical preconceptions for which it is supposed to provide objective evidence? It is this ambiguity which drives the modern fact forward from one phase to the next of its existence.
Poovey’s history of the modern fact is introduced and concluded with some discussion of pre- and post-modern facts, the former in terms of the Aristotelian, Scholastic, view of facts as embodying general knowledge supported by commonplace observations, the latter in terms of the view that the model, or the literary document, generates its own meanings without depending on observed particulars (in post 1870s economics and in the romantic poets, respectively). The contrast between the modern fact and these earlier and later forms brings out what is significant in her conception of the modern fact. Ancient and postmodern facts, by contrast with the modern, do not face the problem of induction. For Aristotle, knowledge consisted in recognising the universal in things. Isolated particulars were the stuff of sensation, not of knowledge. The rôle of experience was illustrative: the ancient unit of knowledge consisted of experience shared by most people, placed in an explanatory context. So ancient facts drew their legitimacy from their status as both a product of theory and a commonplace of experience. Since there was no claim that theory depended on facts - indeed, the opposite was the case - this epistemological stance was not vulnerable to the problem of induction.
Whereas her account of the ancient fact, though necessarily brief, was also a necessary piece of scene-setting required to make the modern fact comprehensible, Poovey’s remarks on the continuing and accelerating displacement of the modern by the postmodern fact have more the status of obiter dicta. Her view seems to be that an early variant of the postmodern fact appears in the early nineteenth century with the Romantic Poets, that it becomes the dominant epistemological form in economics following the marginal revolution of the 1870s, and that it spreads to the rest of society in an ‘epistemological revolution’ of the late 20th century. Postmodernism is characterised, she says, by the conviction that the systems of knowledge we create constitute the only source of meaning. Hence, as in the case of the premodern fact, general, systematic knowledge is privileged at the expense of observed particulars: a fact is just an application of such systematic knowledge to some feature of the observed world. The latter is seen as to a large extent socially constructed and hence potentially consistent with a range of theoretical interpretations. Again, since the postmodern fact gains its legitimacy from the general, systematic knowledge of which it is an application, it is not vulnerable to the problem of induction.
Poovey traces the development of the modern fact from its origins in two mercantile knowledge projects of the sixteenth and seventeenth centuries: the emergence of double entry bookkeeping as a system, and the attempt by early seventeenth century mercantile writers to present commerce as a system, that is, as something organic and self-regulating. The book concludes with an examination of the views on scientific method of John Herschel and John Stuart Mill in the mid-nineteenth century. Occupying a central position in the book is an examination of several knowledge projects particularly associated with the Scottish Enlightenment: experimental moral philosophy, conjectural history and political economy. Shaftesbury, writing at the beginning of the eighteenth century, is seen as a fore-runner of the experimental moral philosophy epitomised by Francis Hutcheson, George Turnbull, and David Hume. Dugald Stewart at the end of the eighteenth century described as ‘conjectural history’ one aspect of the writings of the Adams Smith and Ferguson, David Hume, Lord Kames and John Millar, amongst others. From this group, Poovey identifies Smith as forming the transition from conjectural history to political economy.
Clearly it is not possible here to rehearse Poovey’s account of each of these knowledge projects, but it is worth citing one example, and I will take her approach to double-entry bookkeeping and the mercantilists as that example. Like each of the disciplines Poovey examines, the system of double entry bookkeeping set out in John Mellis (1588) A Briefe Instruction and Maner How to Keepe Bookes of Accompts after the Order of Debitor and Creditor (basically an English version of Luca Pacioli, 1494, De computis et scripturis) prescribed what was to count as a fact and how it was to be related to the systematic representation of facts from which general theoretical knowledge may be obtained. Mellis’s system instructed merchants how to take observed particulars, the various commodities, including money commodities, traded by merchants, and to represent them in a formal, numerical system which generated useful general information about, amongst other things, the profitability of the enterprise. That the approach starts from the observed particular (the transaction), and works it up into systematic knowledge of the account (summarised in the balance), differentiated it sharply from the earlier, Scholastic approach of starting with the general and deducing the particulars therefrom. What this approach hides from view, however, is that the recorded transaction achieves the status of fact only because, and to the extent that, it embodies systematic criteria as to what is to count as fact in the first place. Poovey demonstrates clearly that many of these ‘facts’ were highly fictionalised. For instance, some of the reported ‘facts’ corresponded to no transactions but were created simply to endow the overall systematic presentation of the account with the desired qualities of, for example, symmetry and completeness. Other ‘facts’ did correspond to real transactions, but with a multitude of currencies in circulation, and volatile exchange rates between them, were reported using some conventional exchange rate into the money of account, rather than any actual market rate. Moreover, every transaction reported was represented in a highly stylised, abstract manner which suppressed some (for example, narrative), and privileged other (numerical) aspects of its information content. The social and ideological consequence of this system was, firstly, to present itself as an accurate representation of mercantile activity. That is, the internal precision of the system of accounts of each merchant seemed to imply accuracy: that the whole was a transparent and complete account of that merchant’s interaction with the world. Secondly, by entering everything twice, as debit and credit, in a complicated series of notional transactions between personifications of different aspects of the merchant’s business, the double-entry system obscured the merchant’s profit by making every item of revenue appear as the accounting counterpart to another item of expenditure. Thirdly, it presented the activity of merchants in general as a rule-bound, autonomous, self-regulating sphere of activity.
Whereas Aristotle had identified common experience as one of the two components of the ancient fact (the other being that it was derived from theory), early seventeenth century mercantilist writers such as Mun argued, that an adequate basis for factual knowledge was not to be found in common experience, but in the experience of those who knew what to do with it, that is, in expert experience. Throughout the modern period there is a problem of extracting useful theoretical knowledge, which will lead to a desirable policy prescription, from observed particulars. Here, that problem is addressed by the tacit admission that commonplaces do not automatically aggregate themselves into theories, that something special is going on when theory is produced. ‘Reason of state’ doctrine said that the state, personified by the sovereign, needs to regulate social life to maximise its own welfare. The mercantilist claim was that the appropriate way to regulate commerce was to recognise that those with both the access to the required sensory data, and the expertise to work it up into policy-oriented theory, were the merchants themselves. Hence, like the epistemology associated with double-entry bookkeeping, by privileging mercantile experience, this way of understanding how knowledge was generated worked both to raise the status of merchants generally and to increase their leverage in policy making.
This account raises several themes - notably those of professionalisation and interest - which recur in Poovey’s consideration of subsequent disciplines. By arguing that expertise was required to relate the observed particulars of commerce to useful policy-oriented theory, the mercantilists unwittingly prepared the ground for the establishment of a separate profession of experts dedicated precisely to this task, a development leading to the professional academic political economists and economists of the nineteenth and twentieth centuries. At the same time, an argument which privileges mercantile expertise in the formation of economic policy is clearly open to charges of interest. This raises an enduring theme: how to present knowledge production projects aimed at privileging a sectional interest as the result of a disinterested and hence well-founded procedure. In the statistical accounts of overcrowded early nineteenth century English cities, which triggered the research leading to Poovey’s book, to take one example, this is achieved by separating the statistical material from the interpretative text. For Poovey this is the lapidary form of the modern fact. The numbers are presented as non-interpretative, theory-free facts, and the accompanying text as a theoretical interpretation flowing naturally from those facts. Poovey’s view is clearly that the appearance of a disinterested flow of interpretation from fact to theory actually hides the real, and interested, flow into the opposite direction. The theory is prior and both text and accompanying statistical tables are equally informed by theory. What distinguishes text and tables is a style of writing, where facts are presented as openly theoretical constructs in the text but as naked observed particulars in the tables. But the tables themselves crystallise prior theoretical notions as to what is to be observed, whether and how it is then to be counted, who is to do the observing and counting, how this data is to be subsequently processed, and how such facts can contribute to the establishment of reliable and useful knowledge. Change any of these notions, she suggests, and the results of the supposedly purely data-determined statistical activity will be very different.
2 Adam Smith and
the Scottish Enlightenment
Much of Poovey’s History is concerned with three knowledge projects linked to the Scottish Enlightenment: experimental moral philosophy, conjectural history, and political economy. Her account here is extremely interesting and thought-provoking. Perhaps the most tantalising and disappointing aspect of this fascinating book, however, is her treatment of a figure central to this topic as a whole: that of Adam Smith. All of these disciplines merge in Smith’s writing. Poovey chooses to portray him as a transitional figure, moving forwards, from moral philosophy and conjectural history to a more modern epistemology. My own reading of Smith (Denis, 1999) is as one who advocated and practised a fatalistic social philosophy underpinning a deeply conservative policy prescription. It is symptomatic that of the 52 index references to Smith in A History of the Modern Fact, many of several pages, only two (single-page) references are made to the Theory of Moral Sentiments (TMS): by contrast there are six references to The Wealth of Nations (WN), covering 13 pages. WN is explicitly a work of political economy: it is impossible for us now to understand its significance for Smith without taking into consideration his account, in TMS and The History of Astronomy, of the general project of which WN formed a single component. Smith has been consistently read, particularly by economists, through nineteenth and twentieth century spectacles. These interpretations tell us more about ourselves than about Smith: modern writers, from Marx to Hayek, have seen what they expected to see, an earlier version of their own modern standpoint.
The problem begins with the issue of providentialism: the belief that everything which happens in the world is under the control of a benign and omnipotent deity and hence forms part of a larger plan. Poovey recognises clearly the enormous rôle that providentialism has played in the history of economic thought, in particular in the predominantly Scottish Enlightenment knowledge projects of experimental moral philosophy, conjectural history and political economy. Amongst Smith’s predecessors, ‘[i]n the work of Hutcheson and Turnbull, assumptions about human nature and Providence filled the gaps where evidence was unavailable’ (xxi), while the whole of Chapter 6 is concerned with the ‘vestiges of providentialism in the new sciences of wealth’ in the early nineteenth century. Dugald Stewart is singled out as a particularly extreme case (xxii-xxiii), and the significance of Malthus’s theodicy in the Essay on the Principle of Population, and its suppression in the second (1803) and subsequent editions, is brought out (283-4). But in all this, Smith (and, indeed, Hume) is treated as somehow immune from the contagion:
"With the exception of David Hume and Adam Smith ... [the conjectural historians] typically closed the gap between what they observed and what they assumed with a single belief: the conviction that providential design was working itself out in history ... This strong providentialism, which constitutes the bedrock of most conjectural histories, is most explicit in the work of Lord Kames." (227)
Again: ‘[u]nlike Hutcheson and Lord Kames, [Smith] was only intermittently willing to refer the desire for system to a providential order ("the invisible hand").’ (246)
It is true that Smith says only once in each of TMS and WN that we are guided by an invisible hand. (It is also the case that in The History of Astronomy he criticises primitive peoples for only seeing the work of ‘the invisible hand of Jupiter’ in unusual, rather than in all events). This however, is very far from implying a merely ‘intermittent’ providential atavism. A system of providential order explicitly underlay his whole system. Indeed, he went out of his way, when preparing the final revision of TMS in the last year of his life, to beef up the providentialism of his account in the new passages on universal benevolence (TMS IV.ii.3 [Part IV Section ii chapter 3]). Consider the following:
"When providence divided the earth among a few lordly masters, it neither forgot nor abandoned those who seemed to have been left out in the partition. These last too enjoy their share of all that it produces. In what constitutes the real happiness of human life, they are in no respect inferior to those who would seem so much above them." (TMS IV.1.10)
This is not just a matter of ‘filling the gaps where evidence is unavailable’ but of ignoring evidence entirely in favour of assertions derived from providentialism. Some observable particulars - the division of the earth, for example - are recognised where necessary, while others - the standard of living dependent on that division of the earth - are simply suppressed and replaced with the ‘real happiness’ deducible from the providential assumption. His method throughout is to present us with a Panglossian world where everything is for the best:
"all the inhabitants of the universe, the meanest as well as the greatest, are under the immediate care and protection of that great, benevolent and all-wise being, who directs all the movements of nature; and who is determined, by his own unalterable perfections, to maintain in it at all times, the greatest possible quantity of happiness ... [T]his benevolent and all-wise Being can admit into the system of his government, no partial evil which is not necessary for the universal good". (TMS VI.ii.3.1-3)
Poovey’s descriptions of the providentialism of, for example, Turnbull, Kames and Stewart, apply without modification to Smith:
"Like Turnbull’s, Kames’s providentialism implied two things: that a systematic order underwrote the particulars individuals could see, and that this order was both beautiful and good. Thus his belief in providential design enabled him to direct his readers’ attention away from their immediate woes ... Indeed, in Kames’s account, individuals must learn to discount their immediate fortunes, to look beyond what they experience, and with the help of sciences like experimental philosophy and conjectural history, to recognize the pattern of which individual experience is merely an insignificant part." (228)
Compare this with Smith’s view that
"The wise and virtuous man is at all times willing that his own private interest should be sacrificed to the public interest ... he must consider all the misfortunes which may befal himself ... as necessary for the prosperity of the universe, and therefore as what he ought ... to submit to ... as what he himself, if he had known all the connexions and dependencies of things, ought sincerely and devoutly to have wished for." (TMS VI.ii.3)
Again, writers coming after Smith are presented as having regressed to a stage Smith had passed. ‘Whereas Smith had cautioned that we cannot know the ontological status of the theoretical systems we generate and crave, Stewart insisted that a beneficent design exists’ (274).
"In Stewart’s work we see that the providentialism so prominent in eighteenth-century moral philosophy persisted well into the nineteenth century ... So committed was Stewart to the notion that the philosopher should describe God’s plan that he was willing to wholly discount whatever particulars departed from the order he believed God superintended ... Stewart ... thought that observations should always be subordinated to belief and ... [that] particulars were irrelevant to the vision the devout philosopher could see." (xxii-xxiii)
It is true that, while Smith does believe strongly that the beneficent order exists - and says how unhappy a good and wise man must be who does not believe it - he always states it as something which we believe, rather than know by means of evidence and reason. Evidence and reason support belief, but they don’t substitute for it. Smith systematically downgrades reason (artificial, bad) in favour of belief (natural, good): a good theoretical point is ‘first impressed by nature, and afterwards confirmed by reasoning and philosophy’ (TMS II.5.3), while a bad one is ‘not the doctrine of nature, but of an artificial ... refinement of philosophy. All our natural sentiments prompt us to believe [the opposite]’ (TMS p 91 n). Poovey paints Stewart as regressing to an earlier phase: he believes in providentialism, whereas Smith was agnostic about it. The opposite is true. They both believe in providentialism, but Smith, like Hume (and Kant), is sceptical of the powers of reason to tell us anything very much about it (or anything else), while Stewart, on Poovey’s own account, is more optimistic about the potential of reason and evidence. Since the providential design exists, reason will be able to discover it.
The failure to take on board Smith’s providentialism, his Panglossianism, is linked to a misapprehension of his policy stance and, in particular, his attitude to the state. ‘Smith would never have endorsed the argument that the central government should grow strong precisely by collecting and using the kind of fact he helped naturalize’ (217). Poovey offers no evidence for this assertion, and none could be offered other than the standard view of Smith as an individualist pure and simple. What is not now understood is that Smith was a systematic conservative. The focus for everything in his writings is to reconcile the reader to the status quo. Smith was interested in property: he wanted ‘natural liberty’ for men as owners of property, not as individual human beings. He was very happy to support strong central government where it supported property, as he believed would be the case.
Smith’s account in TMS of the structure of personal psychology with which God had endowed us illustrates this point. That we bestow ‘reverence’ and ‘respectful affection’ on our rulers, even when they are ‘the most brutal and savage barbarians’ such as Attila, Gengis or Tamerlane, is evidence of ‘the wisdom of God’ and His benevolence towards us – because it teaches us ‘to submit more easily to those superiors whom the course of human affairs may assign to us’ (VI.iii.30). And however savage or brutal it might appear, government was always to be supported because its very raison d’être was the defence of property: ‘the very end of [government] is to secure wealth, and to defend the rich from the poor’ (WN V.i.b.12 n. 21); the rich man is ‘at all times surrounded by unknown enemies’ ‘from whose injustice he can be protected only by ... civil government’ (WN V.i.b.2).
A further example of this concerns Smith’s defence of high wages, which Poovey considers at some length in order to illustrate what is to count as evidence, and what rôle numbers might have, in Smith’s epistemology. But a significant point about Smith’s defence of high wages which she misses is this. He is not saying ‘wages should be higher’: he is saying, firstly, that wages are already high; that is, I, a philosopher and political economist, can define them to be high. This means that what is, is good - from the workers’ point of view. But, secondly, that the high wages he has identified do not constitute a problem for society. Again, what is, is good - this time from the point of view of society. This would have been controversial: the mercantilists such as Petty and Mandeville had identified the source of the wealth of nations in the poverty of the working class. Smith’s approach is not to do anything, not to change anything, but to interpret everything as being far nicer then it was previously thought to be. Far from a radical call for an improvement in the condition of the working class, Smith’s account is an exercise in reconciliation to what is.
Nothing in these remarks detracts from Poovey’s general approach. Indeed, she would have made things much easier for herself by recognising the continuity between Smith and his intellectual environment. The awkward and, I think, erroneous passage on Smith’s alleged two theories of nature, for example, would have been unnecessary. In conclusion, Poovey’s History represents a new and exciting way of looking at the history of economic thought but there remains much work to be done in applying it to the detail of the sequence of knowledge projects which constitutes our theoretical heritage.
References
Denis, Andy (1999) ‘Was Adam Smith an individualist?’ History of the Human Sciences 12.3 (August) pp 71-86
Smith, Adam (1976) The Glasgow Edition of the Works and Correspondence of Adam Smith Oxford: Clarendon Press/OUP. Reprinted (1981-82) Indianapolis: Liberty Fund). Volume I The Theory of Moral Sentiments (TMS) ed AL Macfie and DD Raphael. Volume II An Inquiry into the Nature and Causes of the Wealth of Nations (WN) ed RH Campbell, AS Skinner and WB Todd
Back to my Research page.
Send me an e-mail at: a.denis at city.ac.uk.
Revised: Wednesday, 1 September 1999