Merchants and profit in the age of commerce, 1680–1830

Why profit and loss didn’t matter: the historicized rationality of Early Modern merchant accounting Pierre Gervais, University Paris III Sorbonne Nouvelle, Paris, France The present paper has benefitted from funding from the French ANR project MARPROF, and From CNRS UMR 8533 IDHE. I thank Dominique Margairaz, Yannick Lemarchand, and Robert DuPlessis for their helpful guidance and advice. Introduction Account books from the eighteenth century are rarely presented as mysterious, opaque historical artefacts. They were ubiquitous; any archival repository in Europe or on the East Coast of the United States will hold dozens of more or less elaborate recording efforts, from the large volumes of an international merchant versed in the most arcane techniques of double-entry book-keeping, down to the single daybook on which a rural market retailer would scribble her daily transactions. What was recorded seems equally unproblematic. In the standard economic approach, keeping accounts was a self-evident imperative in a world of rational economic agents trying to maximize their profit. Focusing on double-entry accounting, Hans Derks recently summarized this view by concluding that such recording would be kept ‘to improve decision-making, to uncover gain/loss, to track the entity [for which the accounts were kept]’s rights and obligations, and to achieve a greater control or surveillance of transactions internal and external...’ H. Derks, ‘Religion, Capitalism and the Rise of Double-Entry Bookkeeping’, Accounting, Business and Financial History, 18:2 (2008), pp. 187-213, on p. 188. A detailed analysis of accounting textbooks from the Early Modern era by John R. Edwards, Graeme Dean and Frank Clarke Edwards, found six possible roles more or less explicitely given to private accounts: establishing one’s financial position, measuring overall profitability and the resulting changes in capitalization, tracking the main components of the stock of goods traded, valuing them, computing the profit or loss on each of them, and providing prospective information which would reduce uncertainty in decision-making. J. R. Edwards, G. Dean and F. Clark, ‘Merchants’ accounts, performance assessment and decision making in mercantilist Britain’, Accounting, Organizations and Society, 34 (2009), pp. 551-70. Some textbooks were far les specific than others, and actual use of accounts was even further from the most advanced models. Accounting historians have long noted that systematic bookkeeping was still rare even in the eighteenth century, and that the techniques used for balancing books were often crude. Still, it is usually held that this was only a transitional stage in an overall process, through which progress in efficiency was constantly made, and each new development in control and recording merely prolonged and improved earlier efforts within an overall human drive towards ever more rationalizing and optimizing. While improvements in cost calculations mostly came from large agricultural and proto-industrial entrepreneurs, whose accounts form the basis of most of the research in accounting history (they were presumably prompted to be more careful because of their high level of fixed capital), financial reporting did include profit calculations as well, especially when the capital was ‘socialized,’ that is, when it was shared by a plurality of actors, whether partners or stockholders. For the stagnation of commercial accounting, see J. R. Edwards, A History of Financial Accounting (London: Routledge, 1989); also A. C. Littleton, Accounting Evolution to 1900 (New York: American Institute Publishing Co. 1933), or S. P. Garner, Evolution of Cost Accounting to 1925 (University, Ala.: University of Alabama Press, 1954) and more recently D. L. King, K. M. Premo and C. J. Case, ‘Historical Influences on Modern Cost Accounting Practices’, Academy of Accounting and Financial Studies Journal, 13:4 (2009). Compare with discussions on costs in R. K. Fleischman and L. D. Parker, What is Past is Prologue: Cost Accounting in the British Industrial Revolution, 1760-1850 (New York: Garland Publishing Inc., 1997); also D. Oldroyd, ‘Through a Glass Clearly: Management Practice at the Bowes Family Estates c. 1700-70 as Revealed by the Accounts’, Accounting, Business and Financial History, 9:2 (1999), pp. 175-201. For other countries see E. Carmona and D. Gómez, ‘Early cost management practices, state ownership and market competition: the case of the Royal Textile Mill of Guadalajara’, Accounting, Business and Financial History, 12:2 (2002), pp. 1717-44. For a neo-Marxist view, also focused on cost accounting in agricultural and industrial firms, see R. A. Bryer, ‘The History of Accounting and the Transition to Capitalism in England. Part One: Theory’, Accounting, Organizations and Society 25:2 (2000), pp. 131-62; ‘The History of Accounting and the Transition to Capitalism in England. Part Two: Evidence’, Accounting, Organizations and Society 25:4/5 (2000), pp. 327-81; also J. S. Toms, ‘Calculating Profit: A Historical Perspective on the Development of Capitalism’, Accounting, Organizations and Society, 35:2 (2010), pp. 205-21. For the specific advances purportedly accomplished by joint-stock companies, cf. e.g. A. M. Carlos and N. Stephen, ‘Theory and history: seventeenth-century joint-stock chartered trading companies’, Journal of Economic History, 56:4 (1996), pp. 916-24. The present paper will offer a dissenting view, based on the factual observation that when it came to trade, a sphere of economic activity which was central to economic development the Early Modern era, merchant accounting, whether in textbooks or in practice, did not include the kind of profit calculations which would be expected within the framework of this standard, rather whiggish history of bookkeeping. In other words, a much more systematic effort should be made to understand these sources on their own terms when it comes to what commercial actors at the time called ‘profit;’ in this respect, there was an ‘Otherness’ in Early Modern economic attitudes and practices which tends to be understressed in the historiography. An interesting symptom of this Otherness is the extent to which historians have remained unable to translate Early Modern commercial accounts into profit rates. One often comes across ‘net profit’ figures for individual ‘adventures’ in the slave trade and in the long-distance colonial trade, but these shipping accounts were not balance sheets for a whole ‘firm,’ to use a partly anachronistic notion, whether for individual traders or for partnerships, and connecting these discrete accounts to a whole usually requires considerable approximations. Pierre Jeannin, arguably the most knowledgeable specialist on any area of European trade in the Early Modern era, tried only once to assess the profit of one of his traders, and concluded that the only valid result he could achieve was ‘a very general idea of the volume of business treated yearly,’ and that any serious calculation of profit and cost was well beyond what was possible. P. Jeannin, Marchand du Nord: espaces et trafics à l’époque moderne (Paris: Presses de l’ENS, 1996), p. 82. Another attempt at profit calculations in D. Hancock, Citizens of the World: London Merchants and the Integration of the British Atlantic Community, 1735-1785 (Cambridge: Cambridge University Press, 1995), pp. 411-424; it focuses on produce imported from a plantation, thus on the sphere of production, and uses a complex set of approximations to complement the account books. See also D. Hancock, Oceans of Wine: Madeira and the Emergence of American Trade and Taste (New Haven: Yale University Press, 2009), which gives only some capital and inventory overall values. Hancock also quotes a few other works on Great-Britain, none as detailed as his own. For France, see G. Daudin, Commerce et prospérité: La France au XVIIIe siècle (Paris: PUPS, 2005), pp. 274-333. Daudin is highly critical even of basic capital accumulation (‘profit rate’) figures previously offered, especially in the slave trade, because most such figures were not annualized, and therefore did not provide the true internal rate of return (‘Profitability of Slave and Long-Distance Trading in Context: The Case of Eighteenth-Century France’, Journal of Economic History, 64:1(2004), pp. 144-71, on 147). Balance sheets did occur from time to time, and more frequently among bigger traders, but they never seemed to provide the information needed to reach a reasonably complete profit and loss statement. Capital accumulation could be computed, in other words, as well as an implicit yearly rate of return, but the sources of profit or loss could not be accurately identified. Similarly, the Thirteen colonies gave rise to a considerable body of work on merchant practice, but again profit and loss calculations remained exceptional, indeed virtually non existent, and even isolated rates of capital acumulation are seldom quoted. More detailed narratives, though without attempts at reconstructing accounts, can be found in W. T. Baxter, The House of Hancock: business in Boston, 1724-1775 (New York: Russell & Russell, 1965), or K. W. Porter, The Jacksons and the Lees: Two generations of Massachusetts Merchants, 1765-1844, 2 vols (Cambridge: Harvard University Press, 1937). Later works contain even fewer quantitative data; see T. Doerflinger, A Vigorous Spirit of Enterprise: Merchants and Economic Development in Revolutionary Philadelphia (Chapel Hill: University of North Carolina, 1986), or C. Matson, Merchants and Empire. Trading in Colonial New York (Baltimore: Johns Hopkins University Press, 1998); also the various authors in special issues of the Business History Review, ‘Reputation and Uncertainty in America’, 78:4 (2004), pp. 595-702; ‘Networks in the Trade of Alcohol’, 79:3 (2005), pp. 467-526; ‘Trade in the Atlantic World’, 79:4 (2005), pp. 697-844, and of William and Mary Quarterly, ‘The Atlantic Economy in an Era of Revolutions’, 62:3 (2005), pp. 357-526. Why was that so? Part of the story is that profit calculations were not as useful for traders in the Early Modern era as they would be now. In an important article, Basil Yamey pointed out that while accounting textbooks advising traders to compute ‘the particular gain or loss upon each article’ had been around at least since the sixteenth century, as a practical matter the crushing majority of account books, not to say all of them, did not provide reliable cost and benefit data on specific products or ventures in a timely fashion, and consequently could not possibly be used as a basis for detailed cost-benefit calculations and strategic decision-making. As I wrote elsewhere, in the absence of standardized production and enforceable norms of quality, each transaction was largely an act of faith on the part of the buyer, who was usually not expert enough to detect hidden faults and blemishes in quality, and had to rely on the supplier’s good faith; and on the selling side, as soon as a trader was of some importance, it became necessary for him or her to use commission merchants or other agents and trust them to sell the goods at their highest possible price, on a market which was basically unknown to the principal. In a world of highly segmented markets and very imperfect (and slow-moving) information, moreover, prices could fluctuate wildly, suddenly and unexpectedly, and defeat the efforts of even the best suppliers and the most committed selling agents. Thus forecasts were at best informed guesses, and the prices asked depended both on the specific quality of the good being priced and on the state of the market at the time of the transaction, two variables which never remained stable. Even granting that some textbooks developed the theoretical view that profit could be calculated in order to help decision-making and reduce uncertainty in trade (and we will see below that this is not as obvious a proposition as one would expect), the concrete conditions of commerce actually prevented such a use of accounts in most real-life cases. An analysis developed in B. Yamey, ‘The “particular gain or loss upon each article we deal in”: an aspect of mercantile accounting, 1300-1800’, Accounting, Business & Finacial History 10 (2000), pp. 1-12; Hans Derks partly concurs ‘Religion, Capitalism and the Rise of Double-Entry Bookkeeping’. I developed the same points in P. Gervais, ‘Neither Imperial, nor Atlantic: A Merchant’s Eye View of International Trade in the 18th Century’, History of European Ideas 34 (2008), pp. 465-73, and ‘Mercantile Credit and Trading Rings in the Eighteenth Century’, Annales. Histoire, Sciences Sociales [English version], 67:4 (2012), pp. 1011-48. See above all the insights in P. Jeannin, Marchand du Nord, and also Marchands d’Europe. Pratiques et savoir à l’époque moderne (Paris: Presses de l’ENS, 2002). Good records could not really help forecasting; they were also not as necessary as they are today, since Early Modern traders did not have to deal either with demanding stockholders or with highly developed regulatory environments. Derk’s third point, ‘track[ing] the entity’s rights and obligations,’ is based on the fact that both in British Lex Mercatoria as it had developed alongside Common Law, and in Statute law on the French side, well-kept books were supposed to be a key element in any evidence admissible in a Court of law besides formal written contracts. However, the link sometimes made between this purported legal status and bokkeeping is misleading; lawful evidence could be derived from any single-entry chronological record, and the complex interweaving of accounts found in double-entry accounting was entirely unnecessary for this purpose. Moreover, in British Common Law at least, while any kinds of books could serve as evidence, there is no indication that they played an especially important role. In fact, a Jacobean statute from 1609 limited the use of such books for all ‘men of Trades and Handicraftsmen’ to within a year of the transaction in the case of debt proceedings, underlining that these crafstmen ‘do demand Debts of their Customers’ after they ‘have inserted unto their said Shop-Books other Wares supposed to be delivered to the same Parties, or to their Use, which in Truth never were delivered, and this of purpose to increase by such undue Means the said Debt.’ 7 Jac. 1 c. 12, copied from The statutes at large, from Magna Charta, to the end of the last Parliament, 1761. In eight volumes. By Owen Ruffhead, Esq., vol. 3 (London, 1768-70), p. 84. The Act is quoted in W. Wills, Theory and Practice of the Law of Evidence (London: Stevens & Sons, 1894), p. 194. I could not find in secondary sources cases of account books being contested, and do not know whether this Act was significant at all (and also whether it actually concerned traders, or merely craftsmen). In France, the Ordonnance du Commerce of 1673 did state that a compulsory balancing of accounts had to take place at least every year between parties to a contract (Title 1, Articles VII and VIII); and a whole, though rather short, chapter was devoted to the issue of books (Title 3, ‘Des livres et registres des négocians, marchands et banquiers’). As in Great-Britain, however, there was no mention of double-entry. What was required was a ‘journal’ or ‘daybook,’ stamped on the first and last page by the proper authority, containing ‘all their trade, their bills of exchange, the debts they owe and that are owed to them, and the money they used for their house expences,’ and written ‘in continuity, ordered by date, with no white space left’ [between two transactions, to prevent fraud]. J.-A. Sallé, L’Esprit des ordonnances de Louis XIV, vol. 2 (Paris: Chez Samson, 1758), pp. 335-448, on pp. 355, 356. Sallé was a lawyer in the Paris Parliament. No calculation above and beyond this simple act of recording was needed. I will not go into issues of internal control and surveillance here, since except for a handful of large international ventures such as the various East India Companies, such controls were irrelevant to Early Modern merchants, who operated through trusted partners and agents with at most a handful of salaried employees. What remains is the gain/loss calculation, which is indeed what Edwards and his co-authors deem central to the role of double-entry accounting as an engine of progress, with inventory tracking and valuation as a side benefit. Such calculations did take place, either over the whole invested capital, profit then being a ‘ratio obtained by dividing some measure of profit (as a flow of income) by some measure of capital (as a stock of wealth),’ Definition from J. S. Toms, ‘Calculating profit’, p. 212. R. A. Bryer, ‘A Marxist accounting history of the British industrial revolution: a review of evidence and suggestions for research’, Accounting, Organizations and Society, 30:1 (2005), pp. 25-65, makes p. 29 a distinction between what the ‘semi-capitalist mentality’ comparing merely the consumable surplus to the initial capital, and what to him is the real capitalist computation, assessing profit as the net return on employed capital. I have taken Toms’ wider definition in order to guard against any underestimation of profit calculations among the actors I study. or for a specific merchandize or activity, for instance by balancing (closing) the account for a shipping venture or for a particular product. But what I want to argue here is that these calculations entailed economic and social assumptions very different from the ones which govern the notion of profit as we understand it, and that therefore the apparent semantic identity and continuity of the word itself should be questioned. Accounting historians today tend to argue that accounting methods and profit calculations in the past were designed to fulfill the needs of the people who used them, which is true as far as it goes. I have argued elsewhere that the core need of an eighteenth-century trader was to track closely the credit flows which underpinned every commercial activity at the time. Pierre Gervais, ‘Mercantile Credit and Trading Rings in the Eighteenth Century’. Thus the question becomes what consequences this focus on credit would have for profit calculations. In standard economic theory, credit transactions are not different from any other monetarized transactions, but I hope to show in the following pages that this rather ahistorical approach misses important components of the historical record on profit in the Early Modern age. I will rely on the two main sources for profit calculations among eighteenth-century merchants, one theoretical, the other more empirical. On one hand, there are textbooks on accounting, a source often used by accounting historians, as we have seen; on the other hand, there are the quantified traces left by daily merchant activity in various archives, which consist mostly in account books. We will use in this paper a sampling of both, analyzing several textbooks from Great-Britain and France, as well as the account books of two large eighteenth century traders, one in Bordeaux, the other in Philadelphia. Of course, any one of these sources could be unrepresentative, but the combination of all of them, pointing in the same specific direction, does constitute a good indication that we are dealing indeed with what was the standard attitude towards profit. 1) Evidence from textbooks Let us start with textbooks. While we should not confuse traffic laws with actual driving, still these books were sold to the public as training tools, and thus could not very well run counter to the basic assumptions held by this same public about the goals and structuration of this particular technique. There were customers out there who would test them in the process of training themselves; even though most young traders were apprenticed, either some of them felt the need for some written help, or there was a clientele of non-apprenticed merchant wanna-bees, otherwise there would not have been so many publications on the topic. Consequently, the narrative included in works on accounting was unlikely to depart very far from what young apprenticed merchants or would-be merchants would consider the common sense reasons for using accounting in the first place, and for expecting it to be a tool for success. Double-entry accounting in particular required much, much more work than a simple recording of transactions in chronological order. Special training was needed to apply this method, developed in the fourteenth century by an Italian monk, Luca Pacioli, and which required the writing down of two entries for each transaction. The first entry debited the account receiving whatever was exchanged, goods, cash, commercial paper, etc., for the value of what was received. The ‘debtor’ account thus ‘owed’ this value received, and became accountable for it, since it would eventually either have to show where it had gone, or give it back. The second entry credited the account which had provided this same value, and which therefore was not responsible for it anymore, its ‘debit’ being diminished in proportion. This may seem counter-intuitive, partly because our banks send us regularly an account drawn from its point of view, which is actually a mirror image of what our account should be. Our ‘debits,’ and ‘credits,’ at the bank are debits and credits from the point of view of the bank, since the bank holds us ‘debitor,’ when it gives us value, and ‘creditor,’ when we give it back. A proper account at the bank from the customer’s point of view should list the bank as ‘debitor,’ each time we entrust it with money (the bank has received value from us), and ‘creditor,’ each time the same bank gives back that money, either directly to us or to a third party for our account (it has given back value, for which it is not accountable any more). Each transaction thus was translated into an increase of the value given to one account, which it would have to give back or account for when balancing the books, and a decrease in the value for which another account had been hitherto responsible, and which was diminished by the same value (note that in many cases only the value was recorded, not the specific list of goods traded; double-entry was rarely used as a tool for keeping track of inventories...). Such a system required both theoretical knowledge and practical skill, for the choice of which account to debit or credit for a given transaction was not always obvious, and went far beyond the requirements of mere recording. A whole universe of textbooks purported to instruct the young trader in the niceties of this art, but in most cases these textbooks used an approach by example which was undistinguishable from a few years of apprenticeship. And yet, in spite of these arduous prerequisites, double-entry had come to be widely in use among larger traders. Y. Lemarchand, ‘Style mercantile ou mode des finances. Le choix d’un modèle comptable dans la France d’Ancien Régime,’ Annales. Histoire, Sciences Sociales, 50:1 (1995), pp. 159-82; also P. Jeannin, Marchands d’Europe, p. 331-35, 341-51. Was this tool applied to profit measurement? According to Edwards and his co-authors, out of 45 textbooks on accounting published in Great-Britain between 1547 and 1799, no less than 26 dealt explicitely with profit calculations. But what was meant by ‘profit calculation’? I chose to concentrate on the most widely reprinted British and French textbooks from the eighteenth century, John Mair, William Gordon, and William Webster in Great-Britain, and Mathieu de la Porte and François Barrême in France. These were the few authors who had met with widespread approval to the point of achieving something close to classic status, as evidenced by a sometimes impressive number of reprints. From 1701 to 1800, 387 of the 621 textbooks listed by P. Jeannin (Marchands d’Europe, p. 351) had been printed in Great-Britain, Ireland or British North America. For Great-Britain, the number of reprints was estimated for each title using the Eighteenth Century Collections Online database at http://www.gale.cengage.com/EighteenthCentury [accessed 15 september 2013]. The database is not perfect (it does not include 4 of the 32 authors listed by Edwards et al. as published in the eighteenth century) but any widely printed author should show up in it (of the 4 missing authors, 3 cannot be found either in the British Library or in the Bodleian Library catalogs, and 1 has only 1 edition listed at the Bodleian). By number of editions in the ECCO database, John Mair is most represented (19 editions), followed by William Gordon and William Webster (12 editions each). Only 4 other authors in Edwards et al.’s sample are listed as reprinted more than 3 times, Thomas Dilworth (8 editions), Edward Hatton (7 editions), Robert Hamilton (6 editions) and Daniel Dowling (4 editions). In France during the eighteenth century, the most popular textbooks were those by Mathieu de La Porte, Jacques Savary and François Barrême. See P. Jeannin, Marchands d’Europe, p. 382, and Y. Lemarchand, ‘Jacques Savary et Mathieu de La Porte: deux classiques du Grand siècle’, in B. Colasse (ed), Les grands auteurs en comptabilité (Colombelles: EMS, 2005), pp. 39-54. Jacques Savary’s extremely popular textbook, Le Parfait Negociant, only mentioned accounting in connection with inventories, and ignored double-entry entirely. Predictably, all three British authors are listed among Edwards et al’s profit calculating textbooks. But a closer look at how profit was dealt with by them gives some surprising results. Let us start with John Mair, whose Book-keeping Methodiz’d went through at least 19 editions (and was used by George Washington). As I developed elsewhere, See a shorter version of this analysis in P. Gervais, ‘Mercantile Credit’. according to Mair accounts were ‘of three kinds, viz. personal, real, and fictitious. A personal Dr. or Cr. is a Person’s Name; as David Wilson in the preceeding [sic] Post. A real Dr. or Cr. is a Thing; as Cash, Sugar, Shalloon, &c. A fictitious Dr. or Cr. is a Term made use of to supply the want or personal or real one; as Profit and Loss, Voyage, &c.’ J. Mair, Book-keeping Methodiz’d: or, A Methodical treatise of Merchant-Accompts According to the Italian Form (Edinburgh: W. Sands, A. Murray, and J. Cochran, 1749) [1st edition 1746], p. 20 [italics by Mair]. I must thank Yannick Lemarchand for referring me to these editions and teaching me how to use them. These categories are still in use today, but the ‘fictitious’ accounts are called ‘nominal’ by 21st-century accountants, and are defined as temporary accounts, closed at the end of each accounting period and making up the income statement since they contain all revenues, expenses, gains, and losses. For Mair, however, fictitious accounts received cases that ‘cannot properly be divided into a Dr. Part and Cr. part, but consist of one of these parts only,’ the primary fictitious account being ‘Profit and Loss.’ He gave a series of examples of such cases, including an inheritance received by a friend, any gift of merchandise or money, gambling wins, as well as ‘Shop-rent, Warehouse-rent, or other Things of the like nature.’ Ibid., 17. The common element in all these operations was simply that their outcome increased or decreased holdings without creating a transactional relationship with a third party. When a gift was received, no one would be credited since it was a gift; the owner decided to add its value to the capital, and the Profit and Loss account functioned as the tool to do so. Conversely, when a warehouse was rented or custom duties paid, there would not be any debitor, and the corresponding value was lost to the owner, who would substract it from the capital, again through Profit and Loss. Thus the main account recording profit was presented as a dustbin for ‘incomplete’ operations which had to be withdrawn from the normal flow of credits and debits. It could still theoretically be used as a primitive income statement at year’s end, and in Mair’s careful reconstruction of a Ledger it did play this role; the gains or losses on individual accounts were to be recorded one by one in the Profit and Loss account, he explained, and since expenses clearly associated with a given merchandize were recorded in the corresponding account rather than in Profit and Loss, the latter did provide a stylized picture of the profit on each account. Ibid., p. 140-141, 76; and p. 116 for expenses recorded in a merchandise account. But the whole point is that Mair was primarily interested in ascertaining the ‘true State of every part, and of the whole,’ of one’s business. Once achieved, this true image could be used either to know one’s own value, or to ‘give a satisfactory Account ... to Persons concerned.’ Ibid., p. 1. The essentially descriptive status of accounting explains why neither prospective analysis not cost considerations played an important role in it. When he wrote, in a passage often quoted since, that a merchant ought to know ‘what Goods he has purchased; what he has disposed of, with the Gain or Loss upon the Sale, and what he has yet on hand; what Goods or Money he has in the Hands of Factors; what ready Money he has by him; what his Stock was at first; what Alterations and Changes it has suffered since,’ Ibid., p. 6. his primary goal in calculating profits either on a specific merchandise or overall was to make sure the assets of a merchant were correctly listed and valued. What counted was the end result, not the process whereby this result was achieved. This explains why merchant profit was not individualized, but ‘embedded’ – in the sense of Polanyi – K. Polanyi, The Great Transformation (New York: Farrar & Rinehart, 1944). Also M. Granovetter, ‘Economic Action and Social Structure: The Problem of Embeddedness’, American Journal of Sociology 91:3 (1985), pp. 481-510. within a larger set of activities, up to and including bets and card games, which seem to us very far from being part of a business. All these activities were lumped together in the Profit and Loss account, and even in some cases within particular accounts; such as in Mair’s description of a merchant who ‘takes a Piece of cloth, or any thing else from the Shop, to compliment his Friend, and omit to enter it in his Books; nothing is more certain, than that the Cloth-accompt in the Ledger would not shew how much of the Cloth were yet disposed of.’ Ibid. p. 6. The profit on the cloth was not at issue here, and Mair did not even think necessary to explain to which account this gift should be debited, since the problem was to determine what exact amount of cloth which remained on hand. Similarly, there was no reflexion on costs, because they appeared not in conjunction with profits, but only either as part of a specific merchandize or venture outside of which they didn’t make sense, or as inconvenient operations to be vacuumed out through the Profit and Loss account. The creation of subordinate ‘Fictitious’ accounts was explicitely left by Mair to the discretion of his readers, with a few possibilities offered (household expenses, insurance, interest charges, penalties incurred when a contract was not fulfilled, and ‘Charges of Merchandize’), but without any guiding principle of the kind Mair provided for real and personal accounts. Ibid. p. 26 for goods, pp. 29-30, 31-4 for debts. This was in complete contrat to Mair’s treatment of shipping ventures; according to him, ‘when a Merchant sends Goods to Sea, ... there is no Dr.; for neither is any thing received in their stead, nor is the Factor to whom they are consigned, as yet chargeable.’ Ibid. p. 17. Cf. also p. 36. In other words, the shipped goods did not belong to inventory any more (they were outside of the physical control of the merchant), but were not yet under anybody else’s responsibility. The seemingly obvious (for us) nature of these goods, as assets belonging to a real account, was thus not obvious at all for an eighteenth-century author, who gave theoretical priority to the kind of credit relationship embodied in an accounting transaction, over the tracking of profit by families of transaction. Since this was, again, a regal act of the owner, with no other party debited for the goods, it became logical to see such accounts as branches of the Profit and Loss account – from which they were extracted when they reached the factor. In practice, most merchants would sensibly take the view that adventures were real accounts; but Mair’s more complex theory is worth noting. The disconnect between profit calculations and accounting procedures is beautifully illustrated by another classic author of the period, William Gordon, in the second volume of his textbook on merchant arithmetic. Besides adopting the exact same classification and explanation of the ‘Fictitious’ account of Profit and Loss as a depository for ‘defective’ operations lacking a creditor or debtor, he adds a general discussion of profit which is particularly illuminating for our purpose. According to Gordon, profit calculations must positively not be used for prospective analysis: ‘Commerce, like the course of exchange betwixt nations, is by nature variable and fluctuating: the branch which may have afforded considerable profit at one time, may scarce be worth embarking in at another, as the markets may be overstocked by a multitude of traders.’ Consequently, what is needed above all is ‘proper intelligence of the best markets, and a thorough judgment of the quality of goods’ combined with ‘a real certainty of an advantageous and ready sale ... The profit by trading does not arise so much from charging high, as [from] ready sales and quick returns.’ W. Gordon, The Universal accountant and complete merchant, vol. 2 (Edinburgh: A. Donaldson and J. Reid for Alexander Donaldson, 1735), pp. 1, 2-3. Within such a universe, profit could be measured ex post, but its quantification had no bearing on future profit. Information, networks, quality scales, credit flows, and anticipations on the evolution of highly segmented markets were the key determinants of profit, and none of these elements were reflected in the accounts, which were thus largely useless in this respect. Mair mostly left out a reflexion on profit, but his formulations were ambiguous enough to imply that such a reflexion could take place. Gordon dismisses it explicitely. Indeed, ‘Loss and Gain’ as a chapter head does appear in his first volume, within an explanation of the way percentages could be used to compare profits on two operations of different value. Our author concludes his section with this fascinating paragraph: ‘In selling on credit, merchants generally propose a certain profit, which they calculate upon the prime cost of the goods, added to the real and imaginary charges. – The real charges on goods are freight, insurance, lighterage, porterage, ware-room rent, &c. and the imaginary charges are risk of bad debts, dilatory payments, short insurance, possible accidents in the carriage, risk of having them long on hand, &c.; and where this calculation is made according to the profit they propose, the goods are marked on the cover with something characteristical of the price, which is known only to those concerned in the shop or ware-room. After all, merchants are frequently obliged to conform themselves to the market, selling under the rate they proposed, when there are few bidders, and as demands rise, taking the best price they can get.’ Ibid., vol. 1, p. 202. Again, Gordon is explicit: most of the components of the price-setting process were entirely disconnected from any cost accounting, with the exception of the cost of acquiring the goods, and of some charges – and even then since at least warehouse charges were explicitely listed in both Mair and Gordon as belonging to the ‘Fictitious’ account, a part of the price component could well end up coming from outside the account of the merchandise under consideration. Together, Mair and Gordon present a very clear view of what accounting was supposed to achieve in relation to profit quantification. The latter was useful only within the framework of a general balancing of the account, carried out in order to determine precisely ‘the true State of one’s affairs,’ to use an eighteenth-century formula. Even partial balances on a specific venture of for a specific merchandize were drawn with this overarching objective in mind, and certainly not to contribute to any kind of strategic reflexion on profit. The end result of accounting was indeed to determine how much had been gained and lost, but the goal was precision and exhaustivity rather than comprehension of the underlying sources of whatever profit was gained. This, then, was the view of the most widespread textbooks in Great-Britain, including Webster, our third author, who did not add any significant element to the overall picture. He barely mentioned overall profitability, introduced the Profit and Loss account only in the framework of the closing of the accounts, and explained it very succinctly as follows: ‘All such Accompts, as House-Expences, Charges on Merchandizes, Refusals of Bargains, &c. as they are only Particulars of irrecoverable Expences, or Disbursements which tunr to no Account, so they are all ballanced by Profit and Loss.’ W. Webster, An Essay on Book-Keeping, According to the True Italian Method of Debtor and Creditor, by Double Entry (London: D. Browne, C. Hitch and L. Hawes, 1755) [12th edition], pp. 11, 12 [Italics by Webster]. The reasoning is similar to Mair’s and Webster,’ but this older author (the first edition dates back to 1719) stuck to the traditional teaching method by examples, and therefore did not feel obliged to try and outline general principles – his ‘general rules’ cover all of twelve pages in a pocket format. The practical examples he provided, though, are fully consistent with the general view I just outlined. The French authors do provide us with an even more direct expression of this view, however, thanks to what was already then (as now?) a French specialty: the grandiose methodological statement. One of the most edited French works in the eighteenth century was Mathieu De La Porte’s Science des négocians, originally published in 1704. M. de La Porte, La science des négocians et teneurs de livre, ou instruction générale pour tout ce qui se pratique dans les Comptoirs des Négocians, tant pour les affaires de banque, que pour les Marchandises, & chez les Financiers pour les Comptes (Rouen: P. Machuel et J. Racine, 1782). The book is a rewriting of an earlier version from 1685, the Guide des négocians, already a smashing success. In his opening sentences,’ De La Porte already drastically limited the role of accounting: ‘The Science of Merchants is made up of two points: 1° To know all the qualities & circumstances of the goods they trade: 2° To know how to make the necessary entries to manage their trade in an exact order, which will give them a perfect knowledge of it at all times. The knowledge pertaining to the first point is acquired more through the practice of it among Merchants, than through the precepts one could provide about it. The Science pertaining to the second point, or to the Entries practiced in the counting-houses of Merchants, can be reduced with certainty to a set of principles or rules...’ Ibid., p. vii. ‘La Science des Négocians consiste en deux points: I° A connoître toutes les qualités & les circonstances des choses dont ils font commerce: 2° A savoir faire les écritures nécessaires pour conduire ce commerce dans un ordre exact, qui en donne une parfaite connoissance en tout tems. La connoissance renfermée dans le premier point, s’acquiert plus par l’usage que l’on en fait chez les Négocians, que par les préceptes que l’on en pourroit donner. La Science du second point, ou des Ecritures qui se pratiquent dans les Comptoirs des Négocians, se peut réduire à des principes, ou règles certaines...’ What is striking here is what was excluded from accounting, namely everything pertaining to the quality of the goods traded, and generally everything belonging to the material side of the transaction. Since quality was a key issue in any transaction, disconnecting book-keeping from its study turned into into a tool of limited usefulness for profit analysis. But De La Porte was even more radical a few paragraph later, writing that: ‘It must nevertheless be agreed that a merchant who uses cash for all his purchases, who borrows neither merchandise nor money for his Commerce, & who lends nothing to anyone, could dispense with maintaining & keeping books, because he cannot fall into any of the cases described by the Ordonnance. He has neither active nor passive debts; as a consequence he fears neither failures nor bankruptcies, & he is in no danger of failing himself, nor of causing losses to his Creditors, because he has none. This case is not without example, and I have seen a Merchant (a retailer, in truth) who, during the more than sixty years that his shop was open, although he even had a fair amount of business, never borrowed nor loaned anything, & as a consequence he had no Book: However he conducted his business with much honor and probity, & without encountering any financial difficulties. But this is a very rare thing, & could not be the case of a Merchant with a somewhat considerable trade. It is therefore necessary for he who borrows and lends to keep his books very exactly, in order to see at all times the state of his Affairs. His Books will teach him which affairs and which negotiations have profited him or caused him losses, and he will know who are his debtors and creditors, in order to satisfy the latter and get himself paid by the former, & furthermore he will be in a condition to be able to account for his conduct, in the unfortunate event that his affairs began to suffer and he not have what he needs to satisfy his Creditors.’ Ibid., p. viii-ix. ‘Il faut néanmoins convenir qu’un marchand qui achète tout comptant, qui n’emprunte ni marchandises, ni .argent pour son Commerce, & qui ne prête rien a personne , se pourrait dispenser d’avoir &. de tenir aucun Livre, parce qu’il ne peut tomber dans les cas prévus par l’Ordonnance. Il n’a ni dettes actives, ni dettes passives; ainsi il ne craint point les faillites &: banqueroutes, & n’est point dans le cas de manquer lui-même, ni de faire perdre à ses Créanciers puisqu’il n’en a point. La chose n’est pas sans exemple, & j’ai vu un Marchand (en détail à la vérité) qui pendant plus de soixante ans de boutique ouverte, quoique mêmeil ait fait des affaires assez fortes, n’a rien emprunté ni rien prêté, & qui par conséquent n’a eu aucun Livre: cependant il a conduit son négoce avec beaucoup d’honneur & de probité, & sans aucun embarras. Mais la chose est très-rare, & ne pourroit pas être dans un Marchand qui feroit un négoce un peu considérable. Il est donc nécessaire que celui qui emprunte & qui prête tienne les Livres exactement, afin de voir en tout tems l’état de ses Affaires. Ses Livres lui apprendront quelles affaires & quelles négociations lui ont été à profit ou à perte, il saura quels sont ses Débiteurs & ses Créanciers, pour satisfaire aux uns & se faire payer des autres, & outre cela il sera en état de rendre compte de sa conduite, en cas que par malheur ses affaires venant à manquer, il n’ait pas de quoi satisfaire à ses Créanciers.’ Yes, accounts could tell a trader ‘which negotiations have been profitable, and which have led to losses’ and the ‘state of one’s affairs,’ but why were they useless in the case of cash transactions? The only possible explanation is that in a cash transaction the amount gained (or lost) was immediately clear, and that de La Porte, just as Mair and Gordon, was exclusively interested in measuring assets, and then again only once credit flows entered the picture. Accounting was useful only for tracking credit, and did not include issues of costs, supply and demand, or quality, which existed in a cash transaction, but did not deserve to be recorded since the change in assets was easy to track and its amount easy to calculate. Profit and loss became items to be recorded only insofar as the use of credit made it more difficult to compute them and keep track of them. The same order of priorities is shown even more clearly in the work of another major French authors, François Barrême, who proposed to create two accounts which, combined, corresponded exactly to Mair’s Profit and Losses. The ‘Capital’ account received all exceptional gains or losses, such as ‘32000 Livres received as an inheritance from one of my Uncles,’ a sum which Barrême told his reader to credit to Capital since ‘all these assets come to my profit [and] I will never be obliged to render an account of them to any of my correspondants.’ This is exactly the same approach we found in Mair’s work, but Barrême went further by crediting and debiting systematically all ordinary revenues and expenses to a second account, ‘Profits and Losses,’ completely bypassing real and personal accounts. His example was the following entry: ‘Cash owes to profits and losses 800 Liv[res] that I received as a payment for 6 months interest on 40000 £ State Securities.’ Barrême argued that this way, the account containing these securities would contain only the value they represented, not an addition of this value and the income derived from them. F. Barrême, Traité des Parties Doubles ou Methode aisée pour apprendre tenir en Parties Doubles les Livres du Commerce & des Finances (Paris: Chez Jean-Geofroy Nyon Libraire, 1721), pp. 26-7, 59, 83. Barrême also opened a ‘Household Expenses,’ account, which he closed into Profits and Losses when balancing his books. But the result of this was that neither charges not gains could be related to the original activity which generated them, since they were jumbled into the Profits and Losses account. For Barrême made no mention of ordering these gains and expenses according to their source, which confirms that his Profits and Losses account should not be read as a forerunner of the modern profit and loss statement. As with Mair, the point was not to analyze the sources of profits and of losses (an issue which again is not even raised in the passage), but to sift away revenues and expenses, in order to have a detailed pictures of existing assets and liabilities. And even this last formulation does not do full justice to the peculiarities of the Early Modern merchant mind. In a chapter entitled ‘Reflexions on various accounts established in the Ledger,’ Barrême noted that ‘Capital is the head account to which all the other accounts are subordinated, to which all the other accounts are forced to give accounts of their income & their expense. Cash is a Cashier to whom Capital entrusted the use of his money. The State Securities account is a Clerk to whom Capital has entrusted his State Notes. The Brandy account is a Clerk to whom Capital has entrusted his Brandies...’ Ibid., p. 238; capitals and italics by Barrême. ‘Capital est le compte chef auquel tous les autres comptes sont subordonnés, auquel tous les autres comptes sont obligez de compter de leur recette & de leur dépense. Caisse est un Caissier auquel Capital a confié le maniement de ses deniers. Le compte des Billets de l’Etat est un Commis auquel Capital confie les Billets de l’Etat. Le compte des Eaux-de-vie est un Commis auquel Capital confie ses Eaux-de-vie...’ The vocabulary here was not one of assets and liabilities, but of borrowing and loaning; the balance sheet was a list of loans. To build this general account, which described all that was owed to Capital (the primary ‘Compte du Chef’ in French accounting parlance), and all that Capital owed, one had to extract from all other accounts all net revenues and expenses, as shown by the example of the revenue from State notes. Recording this revenue was thus not perceived as a possible tool to assess the profit from State notes in particular, and eventually to compare with each other the profits gained in various activities seen as competing parts of an interconnected whole, but as a necessary step in cleaning up each of these activities and reducing them to what they owed, or were owed to, from a credit point of view. The French ‘Comptes du Chef’ and the Birish ‘fictional accounts’ were thus versions of the same underlying reality, that of credit flows, the tracking of which trumped any concern of profit measurement. Again, I owe thanks to Yannick Lemarchand, who pointed out to me this identity between Comptes du Chef and Mair’s ‘incomplete’ accounts. Neither De La Porte nor Barrême, Gordon or Mair saw accounting as a tool for tracking credit flows, not a tool for analyzing profits. What defined a transaction was not the activity which generated it, but its place in the credit structure the trader had built. The key differenciation was between transactions generated from outside this credit structure, entailing income or expense straight from or towards the trader’s own capital, and transactions resulting from transfers of value between accounts belonging to this structure of credit. While profit calculations on discrete activities remained possible (discounting the issue of overhead), all authors considered such calculations as marginal at best. The key goal was the careful tracking of the value transferred between creditors and debitors, including inventory value. To reach this goal, ‘fictional accounts’ or ‘Comptes du Chef’ were set up in order to sift out incomes and expenses: in the minds of seventeenth-century accountants, they functioned as tools to charge and discharge capital stock, not as centers of cost and profit accounting. 2) Daily merchant usage: two case studies Theory is one thing; what of practice? The following section offers a short overview of the way two transatlantic traders of the second half of the eighteenth century managed their accounts. First up is Abraham Gradis, whom we know thanks to one of the best kept merchant archives in France, that of the David Gradis & Sons house in Bordeaux, well-known and abundantly used by French Early Modernists such as Paul Butel. P. Butel, ‘La croissance commerciale bordelaise dans la seconde moitié du XVIIIe siècle’, (Ph.D. diss., University Paris I, 1973). On the Gradis family, see R. Menkis, ‘The Gradis Family of Eighteenth Century Bordeaux: A Social and Economic Study’, (Ph.D. diss., Brandeis University, 1988); M. Martin, ‘Correspondance et réseaux marchands: la maison Gradis au dix-huitième siècle’, (M.A. diss., Université Paris I, 2008); Silvia Marzagalli, ‘Opportunités et contraintes du commerce colonial dans l’Atlantique français au XVIIIe siècle: le cas de la maison Gradis de Bordeaux’, Outre-mers, 362-363 (2009), pp. 87-111; J. de Maupassant, ‘Un grand armateur de Bordeaux. Abraham Gradis (1699 ?-1780)’, Revue historique de Bordeaux et du département de la Gironde 6 (1913), pp. 175-96, 276-97, 344-67, 423-48 (on Gradis’s career until 1760). According to Butel, the firm had outfitted 220 ships from 1718 to 1789, twice the average outfitting numbers of the most active firms in Bordeaux, making Gradis by far the biggest outfitter in Bordeaux. Most of this shipping activity was directed toward the Sugar Islands and Canada (Gradis was in business with François Bigot, the infamous intendant du Canada who ended up in the Bastille after 1760); a few slaving expeditions also took place. Concurrently, Gradis was dealing in Bordeaux wine, and in various agricultural produce, mostly within his region and in the West of France. Gradis’ journals have reached us for mot of the eighteenth century, and I have analyzed the year 1755. Fonds Gradis, Centre d'Archives du Monde du Travail, Roubaix, ‘Journal, 20 août 1751-14 mai 1755’, 181 AQ 6* (hereafter referred to as ‘AN 181 AQ 6*’; ‘Journal, 1 June 1755-26 October 1759,’ 181 AQ 7* (hereafter referred to as ‘AN 181 AQ 7*’) I would like to thank the Gradis family, who granted me access to their archives. I analyze in details Gradis’ account structure in P. Gervais, ‘Mercantile Credit and Trading Rings’; see also P. Gervais, ‘A merchant ot a French Atlantic? Eighteenth-century account books as narratives of a transnational merchant political economy’, French History, 25:1 (2011), pp. 28-47, over a slightly different timespan (October 1754-September 1755 instead of January-December 1755). Like most of his peers, Gradis completely ignored French law on regular statements of accounts; there is no trace of a general balancing of the book in Gradis’ two journals, which covered eight years from 1751 to 1759! The lack of ‘fictitious’ accounts was also remarkable: both general and personal expenses were distributed over various accounts, insuring that no clear view of them could be gained. Thus the only profit Gradis could calculate was ‘the gain or loss on a particular article,’ or on a particular venture. In practice, the account structure was overwhelmingly dominated by personal accounts (211 of them), to which should be added eleven partnership and factorage accounts, which in most cases included personal transactions along with those concerning the commissioned goods, and at the very least meant that a third party was involved. Moreover, some of the 9 shipping accounts listed should be considered partnership accounts as well: while Gradis’ accounts rarely specified whether a specific outfitting or cargo was the result of a joint investment by several traders, at least a few silent investors were almost always resorted to. Anyway, the choice of calculating returns on each shipment, rather than over specific goods or specific destinations, made profit calculations most likely useless beyond the usual, descriptive need to measure changes in assets, since each cargo mix was different. More generally, in all these cases, profit calculations concerned personal or credit relationships rather than product lines or defined markets: these accounts were individualized as possible profit and cost centers because of their position as source or destination of credit flows, with no obvious reference to the economic activity which underpinned Gradis’ relation to the person or persons holding these accounts. Thus, setting asides thirteen other accounts for Cash, commercial paper held by Gradis or his bankers, and some types of special contracts such as bottomry loans and freight fees advanced to customers (all these accounts bearing also on credit relationships anyway), unambiguously real accounts connected to merchandise rather than credit flows were limited to three accounts of cargoes entirely owned by Gradis (which Mair would still have seen as fictional accounts rather than real ones), and 10 accounts of merchandises. Besides Cash, general credit accounts were as follows: Bills Payable, Bills to broker, Bills receivable, Protested Bills, Suspense accounts (disputed transactions), Contracts of sale, Bottomry loans, Bottomry Loans in Cadiz, Freight due by His Majesty, Freight due by Sundries, and two accounts with bankers Chabbert & Banquet, and Gaulard de Journy. However, out of these 10 accounts, one, ‘General Merchandise’ was obviously not a specific product, and two others, ‘Land in Talance’ and ‘Wines from Talance,’ were producer accounts, since Gradis had his own vineyard, and traced separately the wine he grew, thus bearing out the well-known link between producer status and cost accounting. Moreover, the ‘Flour’ account included flour sent to the King, whereas whatever flour Gradis bought and sold for his own account was recorded in ‘General Merchandize,’ See for instance AN 181 AQ 6*, 2 March 1755, ‘Marchandises Genérales Dt a Arnaud Gouges & Comp.e £ 2496.17 p 100 Barrils farine qu’ils nous ont envoyé p[esan]t net 17975# a £ 12.10, & 50s par Barril.’ so that it was impossible to calculate profit on ‘flour’ in general. Two sugar accounts (‘Sugar our a/c’ and ‘Unrefined sugar our a/c’) were virtually inactive throughout the year, with total sales of 143 £ tournois – though Gradis found a profit of almost 22,000 £ tournois when he closed them in September 1755, a hefty sum which had sat unrecorded for at least eight months. The same was true for the ‘Campeche wood,’ account, which appears in 1755 only to be closed, and found to have been generating a profit of 2000 £ tournois. AN 181 AQ 7*, 3 March 1755 (sales of sugar), and 11 September 1755 (closing of the three accounts). Overall, only 3 accounts out of 259 (‘Wine,’ ‘Brandy,’ and ‘Indigo our a/c’), can be seen as offering an opportunity to calculate ‘the gain or loss on a particular article,’ rather than within the framwork of a particular venture. Unfortunately, even then significant quantities of the good which could have been theoretically tracked were transferred to factors or cargoes in ships, with no attempt at keeping these subsets of merchandise identified in the accounts Thus the 15 casks of wine sent to Quebec as part of ‘various ships,’ by having entailed unitemized ‘costs’ directly credited to Cash with all the other costs of the cargo, would eventually bring back an amount of money which would be listed within ‘Merchandize sent to Quebec on our account,’ with no way to trace back what that wine brought in compared to other wine sent by other ships, or sold back home. AN 181 AQ 6*, 29 April 1755: ‘Marchandises envoyées à Quebec p n/C dans divers navires Dt, a Divers £ 11780.7.7 p les Suivantes Chargées dans le N.e Le st Nicolas Cap Vincent Suivant Le Livre de factures a f° 135 à Marchandi. Generales p Cordage & Beure £ 4783.7.7 à Vins achetés p 15 th.x a 40# 1800 à la Ve La Roche de Girac pr du Papier 4040 à la Ve Brun p fraix au d.t Papier 340 à Primes d’assurance p £ 11000 a 3 3/4 p C.t 412.10 à Caisse pour fraix 404.10.’ This confirms that the tracking of credit relationships was more important than the tracking of specific products; Gradis could have easily created a series of ‘Wine’ accounts (‘Vins pour compte de la Société pour la cargaison N°7,’ ‘Vins sur le navire Le Benjamin,’ ‘Vins en commission’), which would have enabled him to calculate the expenses and profits associated to wine as a product line. And indeed he did precisely that for his ‘Vin de Talance;’ but this is because the wine in question was the one he produced himself on his own property in Talance. Because he was a producer of wine, he considered important to keep track of the revenues generated by that wine; because the King was a special customer, he created a special account for the King’s flour; and because each shipping expedition was unique, he created an account for each. But the structure he set up made it impossible to track one particular barrel or crate of goods from its acquisition to its sale, so that buying price and selling price could be compared, and to this day no historian is in a position to come up with any such profit calculation, even though the Journal is perfectly well-kept. Whatever work Gradis devoted to selling high and buying cheap was done in a non-quantified, non-recorded fashion, while accounts focused almost exclusively on providing a clear view of the assets and liabilities at any one time, especially the credit position, in full agreement with the developments found in textbooks. The adequation with textbook theory is especially clear if we consider that the Profit and Loss account had at most one subordinate ‘account concerning incomes or expenses (a ‘nominal’ account in today's parlance), namely an ‘Insurance Premiums’ account – actually an ambiguous item, since an insurance premium was potentially both an expense and an asset, depending on the outcome of the insured shipping expedition. It did become a cost once the ships or goods had reached their final destination, and Gradis actually transferred the value of the insurance contract to the adventure which had been insured. But if the ship or goods were lost, the premium became an asset owed by the insurer. Opening a separate account for insurance premiums was thus necessary, but it was so because of the peculiar nature of the transaction, and it would probably be incorrect to analyze the presence of such an account as the result of an analysis of the components of charges and revenues. Thus we end up with only one account, ‘Profit and Loss,’ accumulating day after day all the income received, or expenses incurred, from commissions, discounted bills, interests, closed personal or real accounts, adventures successful or not, as well as a host of hard-to-categorize operations, such as the rather obscure ‘Profit and Loss Dr. to Jacob Mendes £ 237.5 for 4 tables he ordered from Holland,’ an expense which may be personal, or a gift to a third party (4 tables for one person?), or part of a wider transaction... AN 181 AQ 7*, 18 August 1755, ‘Gains & Pertes Dt, a Jacob Mendes £ 237.5 pour 4 tables qu’il a fait venir de Hollande.’ One finds also a long list of corrections on mistakes or omissions, sometimes several months old, and almost always made when ‘settling,’ i. e. closing, a personal account, one more proof that Gradis was much more interested in keeping track of what he owed and was owed than in calculating balances and profits. In this category belongs the spectacular entry which follows: ‘Profit and Loss Dr. to Wines on 1/2 Account with Baillet £ 1298.2.6 for 7 c(sks) 3 Bbls which he delivered heretofore to us pr our 1/2 debiting Profit and Loss considering that the Wines bought Account was Balanced without having been debited for this article.’ AN 181 AQ 6*, 9 December 1754: ‘Gains & Pertes Dt a Vins de Compte a 1/2 avec Baillet £ 1298.2.6 pour 7 th(x) 3 Bq. qu’il nous a cy devant Livré p notre 1/2 debitons a Gains & Pertes atendu que Le Compte des Vins achetés a été Solde Sans y avoir debité cet article.’ Gradis sold 1298 £ worth of wine, left over from a previous venture, the proceeds and remaining inventory of which had been split with one Baillet, then, upon balancing his wine account, found that he had made a profit of 6,166 £ tournois, realized that this was too high a figure, and tracked the mistake down to the fact that he had forgotten to record the arrival of Baillet’s wine in his stock – whereupon he debited directly Profit and Loss rather than correcting the final balance of the ‘Wines bought’ account. This proves, first, that Gradis did not track his inventory through his accounts (otherwise he would have realized sooner that the casks he was recording as sold were coming out of nowhere...), and second that he was not interested in using his merchandise accounts as a basis for profit calculations, since he did not bother to calculate the correct profit for wine as he should have listed it, namely 4,868 £ tournois and not 6,166 £ tournois. Anyway, as we have seen above, the true profit on all this wine, which was most probably listed at buying prices, would only show (or rather, not show) after it was sold in Quebec or elsewhere, and would never make it back into the ‘Wines bought’ account. Gradis was one of the largest French traders, and there is no reason to think his behavior was eccentric. Moreover, it can be shown that the same behavior was found in very different settings, thanks to Levi Hollingsworth, a large Quaker trader and importer in Philadelphia, who left behind archives as rich as the Gradis fund. In some ways, Gradis and Hollingsworth represented two basic elements of international, i. e. mostly colonial, trade. Our Bordeaux businessman imported into France colonial goods in exchange for finished goods and staples from Europe, towards which he redispatched his imports as well as his high-value Bordeaux wine. On the other side of the Ocean, the Philadelphian imported finished goods and some staples from Europe in a newly independent former colony at the periphery of the British Empire, and exported what agricultural goods he could, mostly flour, to the Caribbean plantations, as well as sugar back to Europe. Remarkably, both men used exactly the same double-entry accounting language, to the point where the two journals are basically identicals, if one discounts the ‘Laus Deo’ with which Abraham Gradis, a Jew in a Catholic country, took care to start each of his pages. And above all, both men structured their accounts in much the same way. We find, again, the same crushing domination of personal accounts, the same role of credit categories as key determinants of various accounts, the same paucity of nominal accounts, whether for expenses or incomes, basically replaced by one large Profit and Loss account filtering away the results of the various other accounts. Personal accounts were even more pervasive in Hollingsworth's case; the 686 accounts I found are only part of a larger whole, since should be added to this total an unknown number of commission accounts opened in two separate books lost for the year I studied, 1788, a ‘Flour journal’ and a ‘Sales book’ – only the latter was included in the yearly balance, with no itemization, but for a total amount of over 7000 £Pa (Pennsylvania pounds). On the whole, there is good reason to think that Hollingsworth kept open over a thousand accounts at any time, which made him look a lot like a regional bank. Otherwise, the categories I found fitting for Gradis can be reapplied to Hollingsworth with virtually no modification. Contrary to what one could be led to believe, a number of accounts which look to us like expense accounts (‘Hauling,’ ‘Weighing,’ ‘Cooperage,’ ‘Inspecting,’ ‘Storage,’ etc.) were nothing of the sort; each of these account listed fees incurred by Hollingsworth as commmissioner for others, while the same fees, when paid on account of his own goods, were often (but nor always) debited to ‘Charges of Merchandize,’ or to individual merchandise accounts. Similarly, ‘Shallop’s Disbursements,’ ‘Freight,’ and ‘Outstanding Freight’ were accounts belonging to a partnership Hollingsworth had entered into with two other traders, and which owned a flotilla of coastal ships. The Philadelphian recorded the expenses incurred for these ships, as well as the income he had received from them (‘Freight,’ which was thus not an expense account at all but an account listing income from an asset held in partnership!), then closed each side to the credit and debit of the partnership. He recorded separately, in the ‘Outstanding Freight’ account, the share his partners owed him on that part of the proceeds of the partnership which had come into their hands. All in all partnership, commission, or loan agreements accounted for no less than 35 accounts. There were also 20 ‘Adventure’ accounts, some of them probably in partnership as well. Conversely, the Profit and Loss account was barely more developed than in Gradis’ books thirty years earlier. Insurance premiums were as usual accounted for separately. The listing of household expenses, or of cash and goods Hollingsworth withdrew for himself, was merely a continuation of the medieval tradition of keeping track of one’s expenses. In terms of the structure of the accounts, cost analysis, such as it was, was limited to the creation of the catch-all ‘Charges of Merchandize’ account, and to the appearance of a separate account for Hollingworth’s horses. Some differences do appear, partly as a function of time and space. Means of payment in the newly created United States were more diversified, so that ‘State Money’ and ‘Soldiers’ certificates’ appeared alongside ‘Cash.’ There were also banks, and bank-related accounts (two: one for stock, one for deposits), again leading to a diversification of the means of payment. On the other hand, the dearth of currency was such in Philadelphia in 1786 that Hollingsworth did not even bother to keep his Bills receivable account separate from his main Cash one; by all appearances, his ‘cash’ box contained only commercial paper, and virtually no metallic currency. On the lack of metallic currency, cf. W. T. Baxter, The House of Hancock. Also noticeable is the limited range of credit contracts individualized in the accounts, with no signs of bottomry loans or notarized sales contracts, for instance. Compared to Gradis, our American trader was living in something of a backwater, and while he may have entered into these more complex forms of credit relationships, his daily activity was confined to simpler forms. But, quite strikingly, there was the same relaxed treatment of mistakes, with a ‘Suspense Sales’ account which listed a set of sales so badly recorded that Hollingsworth could not determine to whom the proceeds belonged. On 31 January 1788, this account included the sale of a barrel of coffee, the owner of which was unclear, for the significant sum of 106 £Pa; the same sale was already listed in the preceding inventory of January 1787, and one must wonder how many years it took Hollingsworth to decide that the rightful beneficiary would never show up. Hollingsworth Fund, Collection 289, Historical Society of Pennsylvania, Philadelphia, Series 2a, Vol. 86 Journal L 1786-1788 (herafter ‘Journal L’), p. 295. An identical account, ‘Suspense sales of Flour,’ could be found in the 1786 Flour Book, listing no less than 17 barrels of Flour, worth 36 £Pa, described as ‘so much Flour to be accounted for when claim’d, sold as follows.’ Hollingsworth Fund, Vol. 113, Flour Journal 1784-86 p. 317. The 1787-88 Flour journal is unfortunately lost. This rather carefree recording hardly fits with the careful tracking of inventory and precise calculations of profit figures double-entry accounting was supposed to make possible. Hollingsworth’s system is more original in its use of the Profit and Loss account. A ‘Profit and Loss running account’ account was active year round, as with Gradis, and received all expenses, costs, and incomes which Hollingsworth felt convenient to write off. On top of this however, our trader made a general balance of his books each year, around the end of January or February (the year in the British colony still started on 1 March). He started by closing all personal accounts into a final balance. As a second step, he opened a new Profit and Loss account, in his Journal rather than in his Ledger. Then he closed into this final, ‘Journal Balance’ Profit and Loss account the ‘Profit and Loss’ running account he had used all year, and also all yje active accounts, mostly merchandise, partnership and commission accounts, which he had not yet closed into his other balances or into the original ‘Profit and Loss running account.’ He did so by transferring the value of what merchandise was left, and of whatever had not yet been paid or received and remained due by or to him, into a new account, in turn closed into the general balances, and crediting or debiting the remaining value, in other words the net result of that particular account, to the same general Profit and Loss account he had just opened. This is why some accounts, like ‘Bar iron,’ appeared twice; once as a closed account, with a net profit debited to Profit and Loss, and once as a real account containing Hollingsworth’s stock in Bar iron and a few unpaid debts linked to that particular activity, and listedas an asset in the final balances. Once all the accounts had been cleaned up in this fashion, Hollingsworth drew up his general balance, calculating the value of his assets and liabilities, including the balance of Profit and Loss, and producing a ‘Nett worth’ figure which translated efficiently the state of his wealth, and could be compared from year to year. It could be argued that Hollingsworth’s accounting was significantly more sophisticated than Gradis.’. He did follow the contemporary practice of listing part of his expenses in asset accounts, thus inflating the value of his landed properties, for instance: every time he paid a fee on these landholding, he added this expense to their value. But he also created a depreciation account for the two boats in owned in full, reducing their value at every settlement, and his two Profit and Loss accounts, along with a few subordinate accounts such as House expenses, could form the basis of at least the beginnings of an income statement, admittedy not a complete one, but one clearly separated from assets when the balances were drawn. But the way these Profit and Loss accounts were generated is actually a perfect illlustration of the theoretical approach used in textbooks. They were primarily, and consciously, used as a filter which would clean up, so to speak, all revenues and expenses, and leave the net value of assets and liabilities. Because very few subordinate expense accounts were created, and because the Profit and Loss account remained open throughout the year (and in 1788 at least, took up no less than 4 different pages of the Ledger), various sources of revenues and expenses were jumbled together, and no analysis of the sources of profit and loss was easily achievable. Indeed revenues and expenses recorded during the year did not reappear in any way in the final process whereby the Profit and Loss account was closed, since only the balance of the ‘running’ Profit and Loss account was transferred into the ‘final’ one, losing all details on this particular result. Moreover, numerous costs, fees and expenses were recorded within the merchandise accounts, over the whole year, which means that they did not appear in the Profit and Loss account at all. And of course Hollingsworth followed contemporary practice when shipping goods, and transferred the goods shipped to the adventure account, so that the 357£Pa 15s 4p listed in 1788 as the profit on the flour stocked and sold in his shop remained entirely disconnected from the profit realized on ‘Flour taken up in the Jerseys,’ or within the framework of any other adventure. Conclusion On the whole, what we can gather from the activity of these two large traders buttresses the vision we derived from textbooks: double-entry accounting was undoubtedly used in the context of tracking credit flows as precisely as possible, but had little to do if at all with the measurement of profit and loss. This is quite logical, since trade in the eighteenth century rested first and foremost on the forbearance of creditors, whose advances enabled one to go on with one’s business. When one adds to this element the necessity to use outside expertise for the goods one bought, it becomes obvious that one of the keys to success was the ability to rely on other traders, on a circle of allies. This was far more important than any bottom line in any discrete transaction, or even in a series of transactions. Double-entry accounting thus kept track of the quantitative, revolving part of a larger credit, that granted to a trader by other traders. There was a difficult balancing act to maintain on this score alone, making sure that enough cash was on hand to satisfy punctual demands for payment, and weeding out as much as possible potential defaulters from the ranks of those to whom a trader granted credit, with temporalities playing a key role in the whole proces. Again, this was not a simple issue of bottom-line, since the other part of the equation, the necessity to have agents, had to be taken into account. When Gradis committed his wine to Jonathan Morgan in Cork, Ireland, he relied on Morgan’s agency, both legally and materially, to fetch a good price and a profit. If Morgan delivered, then cutting him loose because of a late payment meant forsaking this particular source of income, at least for a time; another agent would have to be found and tested, and there was no telling what would come of it, nor whether the terms of payments one would obtain would not end up being worse for the change. On the whole, each personal account should be read as a partial record of a more complex story, that of a relationship built over the years between two or more people around a specific market, product or place. There was not much interest in calculating a profit, even at this basic level, since the judgement on a relationship would entail a host of other parameters, most of them qualitative. But such a calculation became utterly pointless as soon as one tried to combine several such accounts; apple and pears could not be added. To take, again, Gradis’ wine, each shipment was contextual, and Morgan’s performance in Ireland was not to be compared with the results of a shipping venture to Quebec with the backing of the local Intendant. Conversely, the relationship with a supplier was centered on whether the quality level one was promised would actually be delivered, at what point in time – and at what price but this was merely one piece of information among many. Again, the overall judgement could not have much to do with accounting, or even with prices; whether the price of a wine was right was eminently a qualitative judgement, and was much more important in assessing a supplier than whether its price was low. Which did not mean that nothing should be recorded; when one had to deal with several hundred accounts exchanging their values on a daily basis, it became interesting, and even necessary, to track these flows as precisely as possible, all the more since making too many mistakes could well dent one’s reputation, and lead to a loss of credit, a far, far worse risk than a mere loss of assets in this credit-dominated age. And it did not mean either that traders could not calculate the overall change in their assets, only that not too much should be read into their doing so. Observing an increase in assets was always gratifying, but did not have any practical impact on profit-making, since whatever lesson the past held for the future was not contained in the numbers of an account, but in the intangibles which underpinned the relationship it represented. And even when it came to overall performance, relying on accounting may not have been such a good idea. The distribution of one’s wealth, as reflected in a general balance if one chose to draw it up, was bound to be transient, since the primary imperative was flexibility and the search for opportunity. For this particular quest, the hard numbers of Early Modern commercial accounting were much too static a source to be of much help, which means other tools were used, in ways the rest of this volume explores. Notes PAGE 1