Industrial Milling in the Ancient and Medieval Worlds A Survey of the Evidence for an Industrial Revolution in Medieval Europe ADAM ROBERT LUCAS In 1934 and 1935, Lewis Mumford and Marc Bloch published two very dif- ferent pioneering works in the history of technology, Technics and Civiliza- tion and “Avènement et conquêtes du moulin à eau,” the first an ambitious attempt to trace the development of technology in human civilizations over several thousand years, the second a historical overview of the development of milling technology from Greco-Roman times to the end of the Middle Ages.1 What was to prove an extraordinarily influential thesis about the development of medieval technology appeared in both publications, namely, that the second half of the European Middle Ages witnessed a rapid increase not only in the number of mills powered by water and wind but also in the range of industrial processes to which waterpower and wind power were applied. These phenomena were, according to Mumford and Bloch, emblematic of a medieval revolution in the use of power technology that laid the foundations for what happened in the Industrial Revolution Dr. Lucas works as a researcher and policy analyst at the Cabinet Office for the Govern- ment of New South Wales in Sydney, Australia. His academic research interests include ancient and medieval science and technology, the historical sociology of institutions, and the genealogy of the mechanical philosophy. His book, Wind, Water, Work: Ancient and Medieval Milling Technology, will be published by Brill Academic Publishers later this year. The author is grateful to a number of scholars for their assistance with this article, including John Schuster, Richard Holt, John Langdon, Thomas Glick, Colin Rynne, O¨ rjan Wikander, and the Technology and Culture referees. ©2005 by the Society for the History of Technology. All rights reserved. 0040-165X/05/4601-0001$8.00 1. Lewis Mumford, Technics and Civilization (New York, 1934; reprint, Orlando, 1963), 112, 113–18; Marc Bloch, “Avènement et conquêtes du moulin à eau,” Annales d’histoire économique et sociale 7 (1935): 538–63, published in English as “The Advent and Triumph of the Watermill,” in Land and Work in Mediaeval Europe: Selected Papers by Marc Bloch (London, 1967), 136–68. 1 T E C H N O L O G Y A N D C U LT U R E several hundred years later and helped to explain how European society was subsequently able to transform itself in the way it did.2 In 1941, Eleanora Carus-Wilson added some empirical flesh to the bones of Mumford’s and Bloch’s thesis. Her widely read article, “An Indus- trial Revolution of the Thirteenth Century,” argued that during the thir- teenth century traditional methods of fulling by hand and foot in the urban JANUARY cloth-manufacturing centers of England were largely replaced by mecha- 2005 nized fulling in a number of rural wool-producing areas, leading to large- VOL. 46 scale social and economic changes comparable to those that occurred in the English textile industry of the eighteenth and nineteenth centuries.3 Until the postwar period, to suggest that the Middle Ages had made any meaningful contribution to later scientific or technological achievements was to invite ridicule. The prevailing academic and popular view was that scholastic religiosity and superstition suppressed natural philosophical, mechanical, and mathematical interests. Scientific and technological stag- nation reigned in the Middle Ages as it had done in the ancient world.4 Although they approached the subject in very different ways, Mumford, Bloch, and Carus-Wilson captured the imaginations of a younger genera- tion of scholars in the emergent field of the history of technology with their 2. Mumford appears to have been the first scholar to allude to the idea that there had been an industrial revolution in the Middle Ages. Anticipating Bloch, he argues in the third chapter of Technics and Civilization, titled “New Sources of Power,” that if power machinery is regarded as one of the primary manifestations of the new capitalist econ- omy, “the modern industrial revolution began in the twelfth century and was in full swing by the fifteenth.” 3. Eleanora Carus-Wilson, “An Industrial Revolution of the Thirteenth Century,” Economic History Review 11 (1941): 39–60. Carus-Wilson drew a clear analogy between the social disruption caused by the widespread mechanization of the English fulling industry in the thirteenth century and the mechanization of the English textile industry in the eighteenth and nineteenth centuries and its consequences: “[T]he [thirteenth] century was one of striking progress industrially, though of equally striking change and upheaval. . . . It witnessed, in fact, an industrial revolution due to scientific discoveries and changes in technique: a revolution which brought poverty, unemployment, and dis- content to certain old centres of industry, but wealth, opportunity and prosperity to the country as a whole, and which was destined to alter the face of medieval England.” In a later essay on the medieval woolen industry, she wrote that the mechanization of fulling “was as decisive an event as the mechanization of spinning and weaving in the eighteenth century”; “The Woollen Industry,” in Trade and Industry in the Middle Ages, ed. M. Postan and E. E. Rich, vol. 2 of The Cambridge Economic History of Europe (Cambridge, 1952), 409. As any student of the Industrial Revolution knows, the social and economic trans- formations wrought by the mechanization of the textile industry in the eighteenth and nineteenth centuries are considered paradigmatic of that revolution. 4. Some of the relevant historiographical issues are canvassed by various authors in Michael Adas, Machines as the Measure of Men: Science, Technology, and Ideologies of Western Dominance (Ithaca, N.Y., 1993); E. Smith and M. Wolfe, eds., Technology and Re- source Use in Medieval Europe: Cathedrals, Mills, and Mines (Aldershot, Hants., 1997); Örjan Wikander, “Industrial Applications of Water-Power,” in Handbook of Ancient Water Technology, ed. Örjan Wikander (Leiden, 2000), 401–12. 2  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds new vision of medieval technological prowess. Samuel Lilley, Robert J. Forbes, W. H. G. Armytage, Bertrand Gille, and Lynn White Jr. were some of the more prominent of the young converts. Over the succeeding decades, these pioneering historians of technology articulated what was to become a core set of claims about an exploratory medieval culture exemplified by a number of technological innovations. These included mechanical innova- tions such as the cam, crank, and clockwork; administrative innovations such as double-entry bookkeeping, annual accounts, and audits; and agri- cultural innovations such as the horse harness, heavy plow, and three-field crop rotation. The effects of these technological changes were so profound, they argued, that a revolution in social and economic conditions took place in the second half of the Middle Ages. The most compelling evidence for a medieval technological revolution was, however, the rapid growth in the use of nonhuman sources of power from the tenth or eleventh century onward. By the late 1960s, a relatively detailed account had emerged of how the industrial revolution of the Middle Ages had unfolded. Its basic assump- tions appear to have been ultimately derived from Bloch. The first was that the Romans had failed to make widespread use of waterpower, even though they possessed the relevant technology for at least five centuries before the collapse of their empire. The second was that Christian monasteries were leaders in the reintroduction of Roman water-mill technology to Western Europe at the end of the so-called Dark Ages. The third was that monkish inventiveness had acted as a spur to a revolutionary growth in the use of waterpower and wind power in medieval agriculture and industry from the tenth or eleventh century onward.5 5. Bloch’s views on the medieval break with the ancient world can be found in “The Advent and Triumph of the Watermill,” 143–46. They are echoed in the history of tech- nology literature by, for example, Robert J. Forbes, “Power,” in A History of Technology, ed. Charles Singer et al., vol. 2, The Mediterranean Civilizations and the Middle Ages, c. 700 B.C. to c. A.D. 1500 (London, 1956), 601–6; Lynn White Jr., Medieval Religion and Technology: Collected Essays (Berkeley, 1978), 22; Jean Gimpel, The Medieval Machine: The Industrial Revolution of the Middle Ages, 2nd ed. (London, 1988), 7–10; Terry S. Reynolds, Stronger than a Hundred Men: A History of the Vertical Water Wheel (Balti- more, 1983), 32–35, 45; George Basalla, The Evolution of Technology (Cambridge, 1988), 146–47. Bloch’s comments on the role of monasticism in the spread of water-mill tech- nology can be found in “The Advent and Triumph of the Watermill,” 148, 150–52, and are echoed by Robert Forbes, “Metallurgy and Technology in the Middle Ages,” in Essays on the Social History of Science, ed. Samuel Lilley (Copenhagen, 1953), 50–51; Lewis Mumford, The Myth of the Machine: Technics and Human Development (London, 1967), 263–71; Lynn White Jr., Dynamo and Virgin Reconsidered: Essays in the Dynamism of Western Culture (Cambridge, Mass., 1968), 63–66; Basalla, 148; J. Kenneth Major, “Water, Wind and Animal Power,” in An Encyclopaedia of the History of Technology, ed. Ian McNeil (London, 1990), 232. Bloch’s views on the revolutionary growth in the use of waterpower and wind power can be found in “The Advent and Triumph of the Water- mill,” 141–42, 182, and are echoed in the work of Bertrand Gille, “Le moulin à eau: Une révolution technique médiévale,” Techniques et Civilization 3 (1954): 1–15, and “The 3 T E C H N O L O G Y A N D C U LT U R E To support the idea of revolutionary growth in the use of waterpower, proponents of the medieval industrial revolution thesis drew on Margaret Hodgen’s calculation that 5,632 water mills are recorded in Domesday Book.6 They also appealed to Carus-Wilson’s argument that thirteenth- century England saw a rapid adoption of waterpowered fulling in the tex- tile industry.7 A third supporting strategy (which is the main focus for JANUARY examination in this article) has been to elaborate lists of types of mills, not- 2005 ing where and when they are recorded in the manuscript sources. This VOL. 46 strategy, too, seems to have been inspired by Bloch. The various claims made by Bloch, Hodgen, and Carus-Wilson were thus woven into what proved to be a compelling narrative, one that served as the paradigmatic evidence for an industrial revolution of the Middle Ages. It argued that although the vertical-wheeled water mill was invented in the ancient Mediterranean, it was used exclusively for grinding grain, and then only sporadically, due to the prevalence of slaves, negative atti- tudes toward the banausic arts, and insufficient water resources.8 It was, in Problems of Power and Mechanization,” in A History of Technology and Invention: Progress Through the Ages, vol. 1, The Origins of Technological Civilization, ed. Maurice Daumas, trans. Eileen B. Hennessy (New York, 1969), 451, 458; Lynn White Jr., Medieval Technology and Social Change (Oxford, 1964), 88–89; Gimpel, 10–12. 6. Margaret Hodgen, “Domesday Water Mills,” Antiquity 13 (1939): 261–79. Al- though Hodgen did not go as far as some of her colleagues in asserting the revolution- ary role of the water mill in medieval society, she did argue that its widespread diffusion throughout Anglo-Saxon England constituted “a far-reaching cultural and technological change” in the material culture of England and that “the distribution of Saxon watermills at the time of the Norman conquest forms a background against which to consider the utilization of water power in the later industrialization of England” (p. 262). Hodgen’s figure for Domesday mills has been cited by numerous scholars: see, for example, Samuel Lilley, Men, Machines and History: A Short History of Tools and Machines in Relation to Social Progress (London, 1948), 37; Forbes, “Power,” 611; W. H. G. Armytage, A Social History of Engineering (London, 1961), 47; White, Medieval Technology and Social Change, 84, and Medieval Religion and Technology, 67; Gille, “The Problems of Power and Mechanization,” 451; T. K. Derry and T. I. Williams, A Short History of Technology: From the Earliest Times to A.D. 1900 (London, 1970), 253; Reynolds, 52, 64; Gimpel, 10–12; and Major, 231. It was first challenged in the late 1950s by Reginald Lennard as being too low, and in the mid-1970s H. C. Darby and his colleagues calculated the now accepted figure of 6,082 mills: see R. Lennard, Rural England: 1086–1135 (Oxford, 1959), 278–80, and H. C. Darby, Domesday England (Cambridge, 1977), 361. See Richard Holt, The Mills of Medieval England (Oxford, 1988), 7–8, 11, for a cogent examination of the problems with Hodgen’s figure and her broader theory. Holt has argued more recently that there were probably around 6,500 water mills throughout England at Domesday if estimates for mill numbers in the poorly recorded north are taken into account. 7. For cited references to Carus-Wilson’s findings on industrial mills in England in the history of technology literature, see, for example, Forbes, “Power,” 611; Armytage, 47; White, Medieval Religion and Technology, 54, and Medieval Technology and Social Change, 89; Gimpel, 15–16. For uncited references, see White, Medieval Religion and Technology, 66, and Gille, “The Problems of Power and Mechanization,” 456. 8. The longest list of factors precluding ancient technological development is given by Forbes, “Power,” 601–6. 4  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds fact, medieval European “engineers” (trained by, or working in traditions established by, Christian monasteries) who developed the “Roman” water mill’s full potential through their ingenious incorporation into the milling apparatus of a variety of mechanical innovations, including the cam, crank, and trip-hammer.9 The incorporation of these innovations into medieval water mills allowed them to be applied to a range of industrial processes, from fulling cloth and crushing bark and hemp to forging iron and power- ing bellows, thus freeing human labor for other purposes, just as the steam engine did in the late eighteenth and nineteenth centuries. The widespread mechanization of industry that occurred in the second half of the Middle Ages led to transformations in the medieval economy and society similar to those seen in the “later” Industrial Revolution. Scholars who are familiar with contemporary debates in the rhetoric of the sciences and other disciplines may recognize this re-visioning of medieval technological history as an extension into the Middle Ages of the modernist metanarrative of Western progress. The main objective of this article is to show that exaggerated claims for the widespread use of water- power in a range of medieval industries have been critical to the persuasive power of this particular variation of the progress story.10 In determining 9. On the origins of the cam and trip-hammer, see White, Medieval Technology and Social Change, 79–83. White’s views on the origins and diffusion of the crank can be found in the same book, 104–10, 113–16, 118, 119, 167–68; see also “The Expansion of Technology, 500–1500,” in The Fontana Economic History of Europe, vol. 1, The Middle Ages, ed. Carlo M. Cipolla (London, 1972), 157, and Medieval Religion and Technology, 17–18, 49, 186. For a more recent assessment of the significance of these various me- chanical innovations and their origins, see Reynolds, 23, 26, 28, 29, 70–94, 115, 116. While I do not have the time or space to go into these issues in detail here, it now seems likely that not only the cam but also the trip-hammer and crank were first developed in the late classical period. Discussions of the latest occidental evidence can be found in Michael J. T. Lewis, Millstone and Hammer: The Origins of Water Power (Hull, 1997), and Wikander, “Industrial Applications of Water-Power,” 408–10. John Peter Oleson argues that the crank was probably not known to the Romans, although he does not make it clear to what kind of device he is referring when using this term; see “Water-Lifting,” in Wikander, Handbook of Ancient Water Technology, 263. It seems likely that something resembling a crank handle was used to turn the device known as the Antikythera mech- anism; see Derek J. de Solla Price, Gears from the Greeks: The Antikythera Mechanism—A Calendar Computer from ca. 80 BC (New York, 1965). Joseph Needham notes that the eccentrically placed handles on many rotary querns from the Roman period “constitute a crank”; Science and Civilisation in China, vol. 4, Physics and Physical Technology, pt. 2, Mechanical Engineering (Cambridge, 1965), 186. If we take the purist view of what con- stitutes a crank, it remains unclear how long before the eighth century the crank handle was commonly used in the West, although it was certainly known in China by the Han dynasty (206 B.C.E. to 220 C.E.); see Needham, 111–18, 374, 378–79. The more sophisti- cated mechanisms of the crank and connecting rod appear to be early medieval devel- opments; see Lewis, 6, 84, 112, 114. 10. An instructive case in point is the promotional material used for a recent con- ference on medieval water mills and windmills at which I was invited to speak, which stated that “the mill was central to communities in the Middle Ages as a source of flour, lumber, cloth, and, by the later Middle Ages, iron . . . the humble windmill and watermill 5 T E C H N O L O G Y A N D C U LT U R E the extent to which these claims have been exaggerated, the article draws on my own research as well as the findings of a number of other historians and archaeologists who have been working on the history of milling over the last two to three decades. In discussing the use of waterpower in medieval industries, scholars in several disciplines have for some time made a distinction between “agricul- JANUARY tural” mills, which grind grain, and “industrial” mills, which were devoted 2005 to what are now commonly thought of as industrial applications, such as VOL. 46 fulling cloth, forging iron, and sharpening tools. This article continues that convention.11 For reasons of space and clarity of focus, this article will not examine claims that an industrial revolution occurred during the Middle Ages that are based on rapid growth in the number of waterpowered and wind-pow- ered grain mills. Nor does it examine such claims based on technologies other than those related to waterpowered machinery. Its aim is, in fact, quite limited. It is, first, to examine whether the evidence for waterpowered industry cited by advocates of the thesis of a medieval industrial revolution is sufficiently robust to support the claims that have been made for it. Second, it is to determine whether this same evidence can be better under- stood in the light of more recent empirical research and in what areas fur- ther research may be required. In developing its case, the article sets some basic conditions for the acceptance of evidentiary claims, then works through that evidence to convey a sense of the strengths and weaknesses of the various sets of data, including the trends and patterns that may reason- ably be inferred from them.12 . . . exerted a profound influence on the shape of communities, organization of labor, considerations over taxation, and establishment of property rights . . . the engineering expertise developed by millwrights led to developments in water management, fine tech- nology, and industrial production, which fed the Industrial Revolution many centuries later.” 11. It would seem that Anne-Marie Bautier is the source of this convention, al- though Bradford Blaine’s doctoral dissertation, “The Application of Water Power to Industry during the Middle Ages” (University of California at Los Angeles, 1966), which was supervised by Lynn White Jr., appears to be the first major work of scholarship in English to clearly embrace the distinction. See Anne-Marie Bautier, “Les plus anciennes mentions de moulins hydrauliques industriels et de moulins à vent,” Bulletin philologique et historique 2 (1960): 567–626. The fact that around 80 percent of English mills in the early fourteenth century were located in rural areas, at least 90 percent of which were grain mills, provides further support for making such a distinction. On the proportion of medieval English mills located in rural areas, as well as those engaged in industrial activities, see John Langdon, “Lordship and Peasant Consumerism in the Milling Indus- try of Early Fourteenth-Century England,” Past and Present 145 (1994): 13–14, 31. 12. To the best of my knowledge, Richard Holt is the only scholar to have previously subjected the notion of an industrial revolution in the Middle Ages to critical scrutiny. Holt’s assessment of the state of industrial milling across Western Europe during the Middle Ages is largely shaped by his extensive knowledge of the English situation. My 6  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds The Early History of Industrial Milling In their efforts to depict the European Middle Ages as an important transitional phase on the road to modernity, proponents of the industrial revolution thesis continued to conceptualize the classical period and its non-European contemporaries in Enlightenment terms. They disparaged the technological achievements of the Greeks and Romans and argued that Chinese and Islamic civilizations were outstripped by Europe both eco- nomically and technologically in the second half of the Middle Ages. Thus was medieval European exceptionalism emphasized with respect to both its cultural progenitors and its rivals. This conception looks increasingly problematic when subjected to close scrutiny. An examination of the claim that the Romans and their contem- poraries made little use of the water mill when compared with medieval Europeans is a logical place to start. Over the past two decades, a number of classical archaeologists and his- torians have demonstrated that Roman use of waterpower was far more widespread and innovative than was previously thought.13 This research has been in the context of a more favorable assessment of Roman techno- logical achievements more generally. Archaeological evidence compiled by Örjan Wikander from over forty Roman-era sites suggests that the vertical- wheeled water mill was in widespread use throughout the Roman Empire from at least the first half of the second century C.E., and that it was a pre- ferred technology for some industrial applications.14 Ausonius’ reference to own findings support his research on medieval England, but indicate that the English sit- uation differed substantially from those in the French, Spanish, and northern Italian king- doms. See Holt, Mills of Medieval England (n. 6 above), chap. 9; “Medieval Technology and the Historians: The Evidence for the Mill,” in Technological Change: Methods and Themes in the History of Technology, ed. Robert Fox (Amsterdam, 1996), 103–21; “Mechanization and the Medieval English Economy,” in Smith and Wolfe (n. 4 above), 139–57. For a brief comparative assessment of the English and Continental evidence that is more in line with my own conclusions, see John Langdon, “Was England a Technological Backwater in the Middle Ages?” in Medieval Farming and Technology: The Impact of Agricultural Change in Northwest Europe, ed. G. Astill and J. Langdon (Leiden, 1997), 275–92. Langdon’s recently published Mills in the Medieval Economy: England 1300–1540 (Oxford, 2004) provides additional support for our collective assessment of the English evidence. 13. See Örjan Wikander, “The Watermill,” in Wikander, Handbook of Ancient Water Technology (n. 4 above), 371–400; Lewis. Wikander has conducted extensive and detailed work on Roman-era water mills for well over two decades. For some of the earlier work by classical historians who argued for either technological stagnation or the limited use of waterpower in the Roman era, see Moses I. Finley, “Technical Innovation and Economic Progress in the Ancient World,” Economic History Review 18 (1965): 29–45; H. W. Pleket, “Technology and Society in the Graeco-Roman World,” Acta Historiae Neer- landica 2 (1967): 1–25, and “Technology in the Greco-Roman World: A General Report,” Tallanta 5 (1973): 6–47. 14. Wikander, “The Watermill” and “Industrial Applications of Water-Power” (n. 4 7 T E C H N O L O G Y A N D C U LT U R E sawmills used for cutting marble on a tributary of the Moselle in the late fourth century C.E. is now accepted as authentic.15 Other literary references and recent archaeological evidence from Byzantine-era Ephesus indicate that waterpowered sawmills were widely used in some parts of the former empire until the seventh century, and probably later.16 Michael Lewis has recently presented some persuasive evidence for the JANUARY Roman use of waterpowered forge mills. The remains of large anvils with 2005 what can only be described as mechanically produced deformations have VOL. 46 been excavated at a number of iron mining sites in Roman Iberia, Britain, and Gaul. They strongly suggest that the Romans were using cam-oper- ated recumbent or vertical stamps in the manufacture of iron, and that the mechanisms involved were probably waterpowered. The large number of finds indicates that the technology was regularly used from as early as the first century C.E., as were waterpowered pestles, an alternative to water- powered grain mills.17 Walter Horn’s arguments for waterpowered pilae in the seventh-century plans for the monastery of St. Gall find some sup- port in this context.18 All of this suggests a very early and widespread use of the cam. If we look more generally at a list of the technological achievements of the Hellenic Greeks and Romans, it is far longer and more impressive than some scholars have suggested. These achievements include: the chain of pots and the compartmented waterwheel for raising water; the dough- kneading machine; the olive-crushing mill; the vertical- and horizontal- wheeled water mills; new olive presses; the reaping machine; horizontal looms; the barrow; riverboats; a range of wheeled vehicles; more effective hoists; better aqueducts, with water towers and lead pipes; hydrau- lic pumps; the use of brick and concrete in construction; a considerable above). Wikander’s findings on the relative chronologies of the vertical- and horizontal- wheeled water mills problematize much of the previous literature on the subject. 15. Ausonius, Mosella, 2 vols., trans. Hugh G. Evelyn White (London, 1919–21), 362–64. For recent scholarship on the subject, see K.-H. Ludwig, “Die technikgeschicht- lichen Zweifel an der ‘Mosella’ des Ausonius sind unbegründet,” Technikgeschichte 48 (1981): 131–34; D. L. Simms, “Water-Driven Saws, Ausonius, and the Authenticity of the Mosella,” Technology and Culture 24 (1983): 635–43, and “Water-Driven Saws in Late An- tiquity,” Technology and Culture 26 (1985): 275–76; Wikander, “Industrial Applications of Water-Power,” 404–5. White’s position that the passage is of dubious veracity (Medieval Technology and Social Change [n. 5 above], 82–83) is no longer credible. 16. Wikander, “Industrial Applications of Water-Power,” 405–6. 17. The case for Roman-era fulling mills is put in Lewis (n. 9 above), 89–100; for Roman grain-pounders and ore stamps, in Lewis, 101–10. The term “trip-hammer” is used in the literature as shorthand for cam-operated recumbent or vertical stamps. See Terry S. Reynolds, “Medieval Roots of the Industrial Revolution,” Scientific American, July 1984, 109–16, 111–12, and Lewis, 6–7, for clear explanations of the two types of trip- hammer. 18. Walter Horn, “Water Power and the Plan of St. Gall,” Journal of Medieval History 1 (1975): 219–58. 8  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds variety of war machines; and the mass production of tiles, molded pottery, and bread.19 It would therefore seem reasonable to conclude that the late classical period was far from technologically stagnant, and that it bequeathed much of technological value to both the Middle Ages and the modern era, even if some of those innovations appear to have been lost or forgotten after the collapse of the empire. What, then, of the Chinese? The documentary and artifactual evidence indicates that the Chinese were using vertical waterwheels without gearing to power trip-hammers for rice hulling and to automate bellows for iron furnaces from as early as the first century B.C.E.20 However, horizontal- and vertical-wheeled water mills were not adapted to industrial uses in China until the late third to early fifth centuries C.E., suggesting that they probably were an import from the Mediterranean.21 There is nevertheless clear evidence that waterpower was used widely in the Chinese metallurgical industry from at least the early third century onward, while the use of water mills for grinding grains and seeds was widespread from at least the fifth century onward. Water milling was so commonplace throughout China by the tenth century that two commis- sioners for water mills were appointed to oversee the industry in the east- ern and western regions, an administrative innovation that appears to have been unique to imperial China.22 19. See Örjan Wikander, Exploitation of Water-Power or Technological Stagnation? A Reappraisal of the Productive Forces in the Roman Empire (Lund, 1984), 38; Kevin Greene, “Perspectives on Roman Technology,” Oxford Journal of Archaeology 9 (1990): 209–19. Greene’s “Technological Innovation and Economic Progress in the Ancient World: M. I. Finley Re-Considered,” Economic History Review 53 (2000): 29–59, is an excellent reas- sessment of the work on this subject. 20. See Zhao Jizhu, “Agricultural Machines,” in Ancient China’s Technology and Sci- ence (Beijing, 1983), 425–26; Needham (n. 9 above), 370–72, 392. Needham’s remains the most thorough study in English of the Chinese material on early waterpowered and wind-powered machinery. 21. Jizhu, 426–27; Needham, 369–80, 390–412. Reynolds’s brief overview of medi- eval developments in the Chinese use of waterpower and powered mills for industry is heavily indebted to Needham; see Reynolds, Stronger than a Hundred Men (n. 5 above), 115–17. However, his claim that “Chinese applications of water power . . . [do] not com- pare to the European accomplishment, particularly in the period from the eleventh to sixteenth centuries” is not supported by the evidence compiled by Needham, although it is well known that the Mongol invasion of China in the thirteenth century retarded Chinese technological development for a few centuries. Furthermore, Reynolds’s claim that water levers could have operated the waterpowered bellows and pestles reported by Needham from the first century B.C.E. onward has been questioned by the Irish archae- ologist Colin Rynne, who has pointed out that water levers cannot move at the speed required to operate bellows (personal communication, May 2002). To the best of my knowledge, there has been no recent detailed work conducted on waterpower in ancient China. 22. Needham, 401. 9 T E C H N O L O G Y A N D C U LT U R E If we were to add to this brief summary of Chinese innovations in waterpower the many other technological achievements of the period, we would find once again that there is little or no validity to the idea of medi- eval European exceptionalism. Although it remains unclear to what extent Chinese technology influenced medieval European technology, we already know that Islamic and Byzantine trade routes acted as conduits for the JANUARY adoption of a variety of Chinese innovations in Europe between the tenth 2005 and sixteenth centuries, and this may well have included innovations in the VOL. 46 industrial uses of waterpower. While the manuscript sources pertaining to the industrial uses of waterpower in early medieval Islamic countries have not been adequately assessed, partially because detailed work remains to be conducted or made known to Western scholars, there is clear archaeological evidence in the Middle East for the use of water mills from as early as the seventh century.23 The archaeological evidence suggests that both horizontal- and vertical- wheeled water mills were in widespread use from at least the ninth cen- tury.24 For example, the remains of thirty-one mills now thought to date from between the seventh and thirteenth centuries have been located at two sites in Iraq and Iran, while the sites of twelve horizontal-wheeled water mills in Oman have been dated to between the eighth and tenth centuries.25 By the time of the Crusades, there were reputedly mills in every prov- ince of the Muslim world from Spain and North Africa to Central Asia.26 These mills were engaged in a wide variety of tasks, including grinding grain, fulling cloth, hulling rice, sawing timber, preparing pulp for paper- making, and crushing mineral ores and sugarcane.27 The use of waterpower 23. For a brief overview of some of the literature on industrial mills in medieval Islamic countries, see Donald R. Hill, A History of Engineering in Classical and Medieval Times (London, 1984), 169–72. For Islamic and Christian Spain, see Thomas F. Glick, Islamic and Christian Spain in the Early Middle Ages (Princeton, N.J., 1979), 230–35. See also Thomas F. Glick and Helena Kirchner, “Hydraulic Systems and Technologies of Islamic Spain: History and Archaeology,” in Working with Water in Medieval Europe: Technology and Resource-Use, ed. Paolo Squatriti (Leiden, 2000), 267–330, for a detailed discussion of the development of grain milling in Islamic Spain. According to Glick, the early Islamic manuscript sources are very limited (personal communication, April 2004). 24. Ahmad Y. al-Hassan and Donald R. Hill, Islamic Technology: An Illustrated His- tory (Cambridge, 1986), 53. 25. Örjan Wikander, “Archaeological Evidence for Early Water Mills—An Interim Report,” History of Technology 10 (1985): 162–63, and “The Watermill” (n. 13 above), 376–77, in which he states that none of the Iranian and Iraqi examples can be firmly dated to before the Arab conquest. 26. Al-Hassan and Hill, 53. See Reynolds, Stronger than a Hundred Men, 117, for a detailed set of references indicating the widespread dissemination of medieval Islamic water mills. 27. See Arnold Pacey, Technology in World Civilization: A Thousand-Year History (Cambridge, Mass., 1990), 10–11; al-Hassan and Hill, 54; Donald R. Hill, Studies in Medieval Islamic Technology: From Philo to al-Jazari, from Alexandria to Diyar Bakr (Aldershot, Hants., 1998), XVIII-10. 10  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds in the manufacturing of paper in Islamic countries is supposed to have begun in the eighth century, and there are reputedly references to Islamic fulling mills in the tenth century and to ore-crushing mills in the eleventh.28 It would appear that by the mid-twelfth century, the industrial use of waterpower had spread from Islamic to Christian Spain, where fulling mills, paper mills, and forge mills are recorded for the first time in Cata- lonia. Archaeological evidence from the High Middle Ages suggests that industrial water mills were used in large factory complexes, and that water- powered sawmills were also widespread.29 All of this evidence suggests that, rather than being an autochthonous irruption in medieval Europe from the ninth or tenth century onward, industrial milling had clear precedents in earlier civilizations. The fact that there were such precedents makes claims for medieval European excep- tionalism look increasingly implausible. While a systematic review of the existing research and a thorough examination of the extant manuscript sources remain to be done, it seems increasingly likely that it was through Islamic Spain and the Byzantine Empire that a number of Roman, Islamic, and possibly Chinese innovations in industrial milling technology were conveyed to Western Europe from the tenth or eleventh century onward, providing a foundation for the train of developments that characterized the application of waterpower to industry in the European Middle Ages.30 Waterpower in the Middle Ages When citing evidence for the rapid development of industrial milling in medieval Europe, proponents of the revolution thesis have tended to rely upon chronological lists of industrial mills in various countries as their pri- mary persuasive tool. But even when all of these lists are put together, the resulting body of evidence is not sufficiently large or extensive to prove the case. Consequently, I compiled a much larger database of medieval Euro- pean industrial mills to try to gain a better sense of the geographical and temporal distribution of different types of industrial mill and of the factors that may have shaped their invention and diffusion. Analysis of the data in the larger sample leads to a series of conclusions about the need for further research in a handful of key areas. THE TRADITIONAL EVIDENTIARY BASE In order to test the veracity of claims that the numbers and applications of industrial mills grew rapidly in the second half of the Middle Ages, every 28. Hill, Studies in Medieval Islamic Technology, XVIII-10. Hill did not cite the doc- umentary evidence for these claims. 29. Glick, Islamic and Christian Spain, 231–34; Reynolds, Stronger than a Hundred Men (n. 5 above), 117–18. 30. Hill, A History of Engineering (n. 23 above), 172; Lewis (n. 9 above). 116–21. 11 T E C H N O L O G Y A N D C U LT U R E example of industrial mill cited in the literature advancing the industrial revolution thesis was placed into a table. Its reliability and coverage were then examined. This revealed that around a quarter of the total number of industrial mills in the table had never had any evidence cited to verify their existence, or that the evidence that had been cited was dubious. In finaliz- ing the data for analysis, only those medieval industrial mills for which JANUARY there are clear manuscript sources or firm archaeological evidence have 2005 been included. VOL. 46 Careful analysis of the evidence compiled in the table reveals that it is not sufficiently robust to support the claims that have been made for it. The quantity and coverage of the evidence are both matters for concern, as are the small number of scholars who compiled it. There are also major differ- ences in the variety of types of mills found in the two regions with the largest number of mills. For the whole of Europe between circa 770 and 1600, no more than four hundred industrial mills cited by proponents of the industrial revolu- tion thesis could be authenticated.31 As these four hundred examples are spread over a period of more than eight hundred years and more than a dozen countries, the data clearly do not constitute a very large or represen- tative sample. Furthermore, over 80 percent of the four hundred mills are from French and English sources. If one includes the German and Italian kingdoms, 94 percent of the evidence comes from these four regions. While this is not necessarily a problem, for reasons that will be outlined shortly, none of the scholars who have tried to demonstrate that a pan-European technological revolution was taking place in the Middle Ages has ever noted the biases in their data.32 Although it is historians of technology who have been the main advo- cates for the medieval industrial revolution thesis, over 70 percent of the documented examples of industrial mills cited by them have been drawn from the research of three social and economic historians.33 Only a little more than one hundred of the nearly four hundred mills were identified by historians of technology, and of these the vast majority (90 percent) were identified by a single author, Bradford Blaine (table 1). This was surprising in view of the rhetorical importance attached to this thesis in the history of technology literature. 31. I have followed the convention that the medieval period began circa 500 C.E. and ended in 1500, but for the purposes of capturing the data cited by most of the scholars who have written on the subject have extended the samples to 1600. 32. The distribution of the mills: France (186), England (145), Germany (28), Italy (15), other European countries (15). 33. These historians are Eleanora Carus-Wilson and Reginald Lennard, who identi- fied 140 English fulling mills between them (as well as a single tanning mill), and Anne- Marie Bautier, who identified 147 French industrial mills, around half of which (77) were fulling mills. 12  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds TABLE 1 MEDIEVAL (770–1600) EUROPEAN INDUSTRIAL MILLS CITED BY VARIOUS SCHOLARS Author No. of Mills Bradford Blaine 102 Eleanora Carus-Wilson 124 Reginald Lennard 16 Anne-Marie Bautier 147 Others 10 Total 399 Source—Adam Robert Lucas, Wind, Water, Work: Ancient and Medieval Milling Technology (forthcoming, 2005), app. A. Well over half of the total number of documented industrial mills cited by proponents were fulling mills. But the numbers of these relative to other types of industrial mill in the two main countries documented differ vastly. More than 95 percent of the industrial mills recorded from medieval Eng- land were used for fulling cloth, whereas only 40 percent of those recorded in medieval France were so used.34 Although these figures are significant, they have not received the attention they deserve from advocates of the rev- olution thesis. NEW EVIDENCE ON MEDIEVAL INDUSTRIAL MILLS Because the evidence drawn upon by proponents does not constitute a sufficiently large sample on which to base any firm conclusions, a number of scholars have recently extended the evidentiary base through primary research in the English sources and an extensive trawl of the social and economic history literature. From this material I have compiled a table of references to over fourteen hundred ancient and medieval industrial mills, around eleven hundred of which are documented in the manuscript sources or have been identified through archaeological fieldwork.35 The 34. Specifically, 140 of 145 mills in England, 79 of 186 in France. 35. The source table for this data is reproduced in appendix A to Adam Lucas, Wind, Water, Work: Ancient and Medieval Milling Technology (Leiden, forthcoming). The data are arranged according to the type or function of the mill in question—fulling cloth, grinding bark, pulping paper, and so on. Each entry includes the date or period to which the reference is supposed to refer, its location, the author who cited the example, and the primary source for the information provided, if available. Only those examples that are verifiable with respect to manuscript or archaeological sources have been treated as un- problematic and used as the basis for the analysis here. Where more than ten examples of a single type of mill have been described by a single author, the references provided have generally been grouped and the article concerned cited for the sources. 13 T E C H N O L O G Y A N D C U LT U R E medieval mills included in the table were applied to almost thirty different processes, although, as with the smaller sample compiled by proponents of the revolution thesis, the majority of them (about 60 percent) are fulling mills. It should be noted that this trawl of the primary and secondary sources revealed that while there has been some very systematic research on the JANUARY medieval sources for England, Wales, and Italy over the last two to three 2005 decades, and some recent regional studies that include detailed discussions VOL. 46 of milling practices in medieval Spain, France, and Ireland, most of the rest of Western Europe has not received nearly so much scholarly attention. Nevertheless, the more recent research that has been done on medieval England, Wales, Italy, Spain, and France, together with the earlier research of Carus-Wilson, Lennard, Bautier, and Blaine, arguably provides a reason- able basis for drawing some tentative conclusions about the extent to which these regions applied waterpower to industry. A cursory examination of the data contained in the larger sample tends to support one of the arguments advanced in favor of the industrial revo- lution thesis, in that it demonstrates that between the late eighth and fif- teenth centuries water mills came to be employed in Western Europe in the processing and manufacturing of seventeen different products: malted grains, woolen cloth, leather, sugar, hemp, iron, tools, mustard, timber, silk, paper, olive oil, mineral ores, wire, woad, pigment, and opium. It should be noted that mills dedicated to grinding malted grains have in the past been categorized as industrial mills, and misleadingly described by Anne-Marie Bautier, Georges Duby, Lynn White Jr., Bradford Blaine, and Jean Gimpel as “beer mills.” 36 They are more properly described as malt mills (as they were during the Middle Ages) and classified as grain (or agricultural) mills, rather than industrial mills, as John Langdon has recently argued.37 Al- though it may seem somewhat contrary to ignore these clarifications after having drawn attention to them, I will nevertheless continue to treat malt mills as a form of industrial mill for the simple reason that this is how they have been categorized in the past. Continuing to include them in the dis- cussion also helps to illuminate some broader issues. Table 2 shows the earliest documented instances of twenty different types of industrial mill in medieval Europe; for three-quarters of them, the earliest example dates to the thirteenth and fourteenth centuries—nine to the thirteenth century, another seven to the fourteenth century. We now know that this period saw rapid growth in the medieval population and economy. Improvements in agriculture and the expansion of national and international trade are the reasons most commonly cited for this growth. If 36. See Holt, Mills of Medieval England (n. 6 above), 147–48, for a brief explanation of why such an appellation is misleading. 37. Langdon, Mills in the Medieval Economy (n. 12 above), 40. 14  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds TABLE 2 USE OF WATERPOWER FOR VARIOUS INDUSTRIAL PROCESSES IN MEDIEVAL EUROPE, 770–1469 Use Earliest Primary Source Malted grains 770 Woolen cloth c. 1040 Leather 1138 Sugar 1176 Hemp c. 1200 Iron c. 1200, 1214, 1384a Iron tools 1203 Mustard 1251 Timber c. 1300, 1347 b Silk 1272 Paper 1276 Olive oil late 13th century Mineral ores 1315, 1317 Wire 14th century? Woad 1348 Pigment late 14th century Opium 1469 a Dates are for waterpowered forges, bellows, and furnaces, respectively. bDates are for waterpowered saws and lathes. Source—Adam Robert Lucas, Wind, Water, Work: Ancient and Medieval Milling Technology (forthcoming, 2005), app. A. there was indeed an industrial revolution based upon waterpower during the Middle Ages, therefore, it would seem reasonable on the basis of these data to locate it in the thirteenth and fourteenth centuries. But a shadow creeps over these initially hopeful looking data when we look at the numbers of industrial mills recorded in the manuscript sources up until the end of the sixteenth century in order of their chronological ap- pearance. By far the most abundant types of industrial mill in the sample for medieval Europe are fulling mills and forge mills, which account for 80 percent of the sample (table 3), with tanning mills, sawmills, and tool- sharpening mills contributing another 12 percent of the total.38 If these data can be taken as reasonably representative of the situation across medieval Europe, we can further conclude that if an industrial revolution 38. The latest type of industrial mill appears at the bottom of table 3 (the water- powered blast furnace), while the earliest appears at the top (the malt mill). 15 T E C H N O L O G Y A N D C U LT U R E TABLE 3 INDUSTRIAL MILLS IN MEDIEVAL EUROPE BY TYPE, 770–1600 No. of Type of Mill Documented Mills Malt mill 23 JANUARY Hemp mill (human/animal powered) 19 2005 Oil mill 8 VOL. 46 Fulling mill 635 Tanning mill 58 Forge mill 236 Tool-sharpening mill 31 Hemp mill (waterpowered) 18 Sawmill 46 Bellows (using waterwheel) 6 Ore-crushing mill 4 Blast furnace (waterpowered) 8 Total 1,092 Source—Adam Robert Lucas, Wind, Water, Work: Ancient and Medieval Milling Technology (forthcoming, 2005), app. A. did take place in the thirteenth and fourteenth centuries, it was predomi- nantly restricted to the fulling and iron industries. However, one of the most significant things about the larger sample is that it clearly indicates that the region we now call France was a leader in the application of waterpower to medieval industry (table 4). This is strongly suggested by four key pieces of evidence relating to the extent of research conducted on the best-documented regions and what that re- search reveals. To be more specific: (1) the English, Welsh, and northern Italian manuscript sources have been examined very thoroughly by a num- ber of scholars, whereas the French evidence is not nearly so well studied, and yet it reveals a wealth of data compared to which the rest of Western Europe appears distinctly backward; (2) the French manuscripts cited by Bautier record almost all of the earliest examples of the various types of industrial mill; (3) the largest numbers of most types of industrial mill also come from France; and (4) some types of medieval industrial mill are only recorded in the French sources. The first point requires some background explanation. French and English sources provided 83 percent of the data drawn upon by proponents of the medieval industrial revolution thesis, with an additional 4 percent coming from Italian sources. The regional bias of the larger sample is even more heavily weighted, this time toward England and Wales (60 percent), 16  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds TABLE 4 FIRST APPEARANCE OF VARIOUS INDUSTRIAL MILLS IN MEDIEVAL EUROPE, 770–1443 Type of Mill Country Date Malt mill France 770 Hemp mill (human/animal powered) France circa 990 Fulling mill France 1080 Tanning mill France circa 1134 Forge mill England, France circa 1200 Tool-sharpening mill France 1203 Hemp mill (waterpowered) France 1209 Bellows (using waterwheel) Slovakia, France 1269, 1283 Sawmill France circa 1300 Ore-crushing mill Germany 1317 Blast furnace (waterpowered) France 1384 Cutting and slitting mill France 1443 Source—Adam Robert Lucas, Wind, Water, Work: Ancient and Medieval Milling Tech- nology (forthcoming, 2005), app. A. TABLE 5 INDUSTRIAL MILLS IN MEDIEVAL EUROPE BY COUNTRY, 770–1600 Country No. of Mills England and Wales 682 France 229 Italy 148 Germany 31 Poland 10 Others 27 Total 1,127 Source—Adam Robert Lucas, Wind, Water, Work: Ancient and Medieval Milling Technology (forthcoming, 2005), app. A. with most of the remaining industrial mills (34 percent) being recorded in France and Italy (table 5). Only 6 percent of the mills documented in the larger sample are from other regions. The bias in the larger sample is a con- sequence of the extensive and systematic research that has been done on the subject of industrial mills in England and Wales by six or seven scholars over the last six decades. The history of milling in medieval France and Italy 17 T E C H N O L O G Y A N D C U LT U R E has not drawn so much attention from researchers in recent decades, and comparatively little work has been done on the subject to date for the rest of Western Europe, including Portugal, Spain, Belgium, the Low Countries, Scandinavia, Germany, Switzerland, and Austria.39 This situation would be a severe constraint on drawing any conclusions about a pan-European industrial revolution in the Middle Ages were it not JANUARY for two key facts: (1) medievalists generally agree that what we now call 2005 England, France, and Italy were the most technologically advanced regions VOL. 46 in Europe from the eleventh or twelfth century onward; and (2) the English and Italian research has been very systematic. Consequently, if it is possible to gauge the extent to which waterpower was applied to medieval industry in these regions, it should also be possible to assess the veracity of the industrial revolution thesis, as whatever happened in the rest of Europe can hardly have been any more innovative. Although the number of scholars contributing to the research in the larger sample has only doubled, there are two significant qualitative differ- ences between the research conducted by the new scholars and that relied upon by proponents of the industrial revolution thesis.40 These differences relate to the research methods adopted and the researchers’ aims. The scholars involved in compiling the English and Welsh evidence have trawled thousands of individual manuscripts to build up statistically reliable samples. This includes all of Domesday Book (1086), the Hundred Rolls of Edward I (1279), and the Inquisitiones Post Mortem for the reign of Edward II (1307–27), as well as a large number of calendars of fine rolls, patent rolls and court rolls, ministers’ and manorial accounts, and lay and ecclesiastical charters and cartularies.41 This evidence is also comparable, if certain assumptions are made and understood.42 John Muendel has simi- 39. Of these countries, Spain and Germany have received some attention. See, for example, Glick, Islamic and Christian Spain (n. 23 above), and Reynolds, Stronger than a Hundred Men (n. 5 above), on medieval Spain. Walter Kuhn did some significant work on medieval German industrial mills in the 1960s, but I have not yet had an opportunity to study it. 40. These are, respectively, Eleanora Carus-Wilson, Reginald Lennard, R. A. Donkin, Ian Jack, Richard Holt, John Langdon, and myself on England and Wales, Anne-Marie Bautier on France, and John Muendel on Italy. 41. Domesday Book, the Hundred Rolls, and Inquisitiones Post Mortem record the location of the mill, to whom it belonged, and often to whom it was leased and on what terms, as well as its annual revenue, what type of mill it was, and the year in which the record was made. Although they may not include technical details about mills, these can be gleaned from the manorial account rolls that have survived from the thirteenth cen- tury onward. In my own exhaustive examination of more than fifty ecclesiastical cartu- laries, account books, and surveys, I found only thirty English industrial mills that had not been noted by earlier scholars, the vast majority of which were fulling mills. 42. See, for example, Holt’s comments on the reliability of the Domesday mill rec- ords in Mills of Medieval England (n. 6 above), 5–16 and 107–8, and on the Hundred Rolls, 24–25 and 54. See E. A. Kosminsky’s comments on the Hundred Rolls in Studies in 18  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds larly undertaken systematic searches of thirteenth- and fourteenth-century taxation documents in the northern Italian provinces of Pistoia and Fir- enze, which can for this reason be usefully compared with the English data. On the other hand, while it is difficult to determine just how systematic was Bautier’s research on French industrial mills, her single article on the sub- ject appears to have involved the same kind of trawl of royal and ecclesias- tical records as that conducted by Carus-Wilson on the English sources. The scholars who contributed most of the data for the larger sample thus conducted systematic studies of large quantities of source material to gain insights into how industrial milling fit into the wider medieval econ- omy, whereas the scholars whose work provided most of the data for the earlier sample (including Carus-Wilson and Bautier) were deliberately selective and discriminating in their sampling techniques in order to demonstrate that industrial milling was an innovative and widespread medieval phenomenon. Nevertheless, the fact that Bautier’s single article on medieval France cites the earliest reliably dated industrial mills in almost all of the categories, as well as a number of varieties of industrial mill that have not been found outside France, clearly indicates that, in terms of technological development in the medieval period, the region we now call France was exceptional. With respect to the earliest reliably documented industrial mills of var- ious types (table 2) and the regions in which they were located (table 4): the earliest documented malt mill in medieval Europe dates to the second half of the eighth century in France, the earliest fulling mill to the middle of the eleventh century in France,43 the tanning mill to early-twelfth-century France, the (waterpowered) hemp mill and the tool-sharpening mill to early-thirteenth-century France, the forge mill to the early thirteenth cen- tury in France, England, and Sweden,44 and the sawmill to the beginning of the Agrarian History of England in the Thirteenth Century, trans. Ruth Kisch (Oxford, 1956), 40–41. On the comparability of the mill records in Domesday and the Hundred Rolls, see Holt, Mills of Medieval England, 108–12. See Langdon, “Lordship and Peasant Consumerism” (n. 11 above), 7–9, on the reliability of the Inquisitiones Post Mortem. 43. Paul Benoit and Joséphine Rouillard, “Medieval Hydraulics in France,” in Squat- riti (n. 23 above), 193, cite P. Malanima, I piedi di legno: Una macchina alle origini dell’in- dustria medievale (Milan, 1988), as evidence that “[t]he fulling mill appeared for the first time [in Europe?] in Italy, in the founding charter of 962 of a Benedictine monastery in the Abruzzi.” I have yet to verify the primary source for this claim, however. Apart from the example of an early forge mill, Benoit’s and Rouillard’s chronology of other indus- trial milling developments in France is basically consistent with the work of Bautier (n. 11 above), from which most of the French material cited in this article is drawn. 44. Benoit and Rouillard, 195, draw attention to a description from 1135 by Arnauld de Bonneval of the Cistercian abbey of Clairvaux, who states that “water activated a forge in the heart of the abbey, amidst other milling installations.” I have not had an opportu- nity to check this source, but if it is correct it pushes back the earliest date for the forge mill in France by sixty-eight years. 19 T E C H N O L O G Y A N D C U LT U R E the fourteenth century in France.45 Other processes, such as cutting and slitting metal and minting coins, also appear to have been first adapted to waterpower by the French.46 The earliest evidence of blast furnaces simi- larly comes from France.47 The French provinces in which Bautier found most of this activity were Normandy, Picardy, and Champagne in the north, Burgundy and Dauphiné in the east, and Gévaudan, Languedoc, and JANUARY Provence in the south. These provinces were variously under the jurisdic- 2005 tion of the kings of France and England and the Holy Roman Emperor be- VOL. 46 tween the eleventh and fifteenth centuries. With respect to the second key piece of evidence regarding the relative proportions of different kinds of industrial mill that have been identified in each of the four better-studied countries, it is significant that the meticu- lous research that has been done on medieval England and Wales has revealed a relatively small number and limited variety of mills compared to those recorded for less-well-studied France. Not only does the region which we now know as France appear to have been using waterpower for indus- trial purposes earlier than any other Western European country, with the possible exception of Spain, it was applying waterpower to a much wider variety of industrial processes: malt mills, oil mills, tanning mills, tool- sharpening mills, hemp mills, sawmills, and paper mills—52 percent of all such mills in the sample were from France, while 25 percent were from Italy, and only 10 percent from England (table 6). Whereas the Italian king- doms seem to have applied waterpower to a similar range of industries as the French, there are no examples of oil, hemp, or sawmills from medieval England. Furthermore, the only evidence for such unusual waterpowered mills as those used for extracting oil from poppy and hemp seeds, as well as for making mustard and polishing gems and armor, comes from France. It would therefore seem reasonable to conclude that these kinds of milling processes were mainly restricted to France, as only a little over half of those mills that appear in the sample were from elsewhere. Three-quarters of all these mills were from France and Italy. While a relatively large number of industrial mills are documented for the southern and eastern regions of Germany (most of which were used for metallurgical applications, though there are a significant number of saw- 45. With respect to sawmills, this involves excluding the late Roman and Byzantine examples cited. Benoit and Rouillard, 194, claim, without documentation, that the earli- est medieval reference to this technique is from a Norman document of 1204. If this is correct, this would also clearly establish France as a leader in the application of water- power to sawmilling. Interestingly, those two waterpowered technologies with precursors in the Roman Empire and/or ancient China, that is, the sawmill and the forge mill, do not appear in the European documentation until relatively late. The reasons for this re- main unclear, however. 46. In 1443 in the first instance and 1551 in the second. 47. The French examples are dated 1384, 1402, and 1412; the earliest examples from elsewhere being 1429 in Italy and 1496 from England. 20  TABLE 6 WATERPOWERED MILLS BY COUNTRY, 770–1600 Belguim/ Spain France Italy England Germany Poland Flanders Malt mill 0 13 0 4 1 0 5 Oil mill 1 3 4 0 0 0 0 Tanning mill 0 37 5 9 0 2 0 Tool-sharpening mill 0 13 8 5 4 1 0 Hemp mill 0 16 2 0 0 0 0 Sawmill 0 12 26 0 5 2 0 Paper mill 1 1 1 0 0 0 0 Source—Adam Robert Lucas, Wind, Water, Work: Ancient and Medieval Milling Technology (forthcoming, 2005), app. A. 21 LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds T E C H N O L O G Y A N D C U LT U R E mills as well), the only areas in which Germany appears to have been some- thing of an innovator relate to the “copper mill” (twelfth century?) and the boring mill (1480).48 The available evidence suggests that metallurgical processes were first widely adapted to waterpower in Spain, France, Eng- land, Germany, Sweden, Poland, and Italy during the thirteenth and four- teenth centuries. The earliest well-established examples of forge mills are at JANUARY Kirkstall Abbey in England (circa 1200), E´vreux and Evry in France (1202 2005 and 1203), and the village of Toaker in Sweden (1224). Other metallurgical VOL. 46 machines, such as waterpowered bellows and pumps, appear to have first emerged in Italy in the early thirteenth and fourteenth centuries respec- tively. There is reason to believe, however, that forge mills and waterpow- ered bellows may have been present in Spain and France a century or more earlier, although it remains unclear whether the machines concerned were independently reinvented by medieval Europeans or were adaptations of Chinese and/or Roman technology from Islamic Spain and North Africa. Monastic cartularies and royal charters are the main kinds of docu- ments that have informed the English case studies I have conducted myself, and are significant primary sources for the other scholars who have worked on medieval British mills as well. They are also the kinds of sources that Bautier drew upon, and yet no one who has worked on the English sources has found anything like the variety or quantity of industrial mills that Bautier discovered. The relative rarity of industrial mills in medieval England is clearly borne out by my own research and by Holt’s and Langdon’s. Langdon has undertaken the most extensive surveys to date. The most recent involved a trawl of the mill-related data contained in the Inquisitiones Post Mortem for the reign of Edward II (1307–27), which record the holdings of major sec- ular lords at the time of their deaths. Langdon found that only fifty-five of the 1,647 powered mills identified, or less than 3.5 percent, were industrial mills, and all were fulling mills.49 In an earlier study of water mills and 48. What Robert Forbes called a “copper mill” was either an ore-crushing mill, a cop- per-smelting mill, or a copper-polishing mill. It must, therefore, have been an adaptation to copper mining of an existing device or devices. Neither Forbes nor any of the other authors from whom this material is drawn have provided verifiable sources for medieval copper mills, silver mills, water-driven mining hoists, wire mills, or winding mills. 49. Langdon, “Lordship and Peasant Consumerism” (n. 11 above), 12–14. Langdon’s estimate is based on a systematic study of mill rentals on more than twelve hundred English manorial estates. Other regional studies of industrial mills in medieval England whose data were not included in this study are Scott E. Kilburn, “Early Cloth Fulling and Its Machinery,” Transactions of the Newcomen Society 12 (1931–32): 31–52; Rodney F. Butler, The History of Kirkstall Forge: 1200–1945 A.D. (Kirkstall, 1945); Ian S. Beckwith, “Londonthorpe Mill,” Industrial Archaeology 8 (1971): 25–28; Owen Bedwin, “The Exca- vation of Ardingley Fulling Mill and Forge, 1975–76,” Post-Medieval Archaeology 10 (1976): 34–64; J. G. Rollins, “Forge Mills, Redditch, Worcestershire: From Abbey Metal- works to Museum of the Needlemaking Industry,” Industrial Archaelogy 16 (1981): 22  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds windmills in the West Midlands between 1086 and 1500, he found that industrial mills for both secular and ecclesiastical lords “never exceeded 10 per cent of all mills in the pre-plague period.” 50 Three-quarters of the mills recorded by Langdon were ecclesiastical. Less than 1 percent of the total were neither grain mills nor fulling mills.51 My own exhaustive analysis of the manuscript sources pertaining to more than thirty medieval English religious houses tends to support the findings of Langdon’s West Midlands study. On six English Benedictine estates between the late thirteenth and mid-fourteenth centuries, 10 percent of the mills were industrial mills.52 On five English Cistercian estates, more than 14 percent of the total number of mills were industrial.53 On the other hand, ten Augustinian houses recorded no industrial mills before 1348. It thus seems clear that while the Benedictines and Cistercians may not have been as keen on applying water mills to industrial uses as historians of technology have tended to claim, they were undoubtedly more involved in industrial milling than most of their religious brethren, and certainly more so than lay lords, at least up until the early fourteenth century.54 However, 158–69; C. J. M. Beamish, “The Fulling Mill—One of the Oldest Mechanised Industries in Britain,” Industrial Archaeology 18 (1983): 78–82. 50. Langdon, “Lordship and Peasant Consumerism,” 14, n. 27, citing J. Langdon, “Water-Mills and Windmills in the West Midlands, 1086–1500,” Economic History Review 44 (1991): 434, table 2. Langdon found that it is only after the first major plague outbreak that the numbers of industrial mills in England began to increase, particularly in the period between 1450 and 1475. Even then, however, the increase was only marginal. 51. Langdon, “Water-Mills and Windmills in the West Midlands,” 436. 52. The houses were Battle, Bec, Canterbury, Durham, Hereford, and Lancaster; see Adam Lucas, “Machinariarum Nihil Ex Deo: The Role of the Church in the Development of Powered Milling in Medieval England” (Ph.D. diss., University of New South Wales, 2003), chap. 3. This was not necessarily typical of English Benedictine houses, however. Holt found that whereas Glastonbury in the south had interests in three fulling mills out of the forty mills that it held by the early fourteenth century, or 7.5 percent of the total, the East Anglian houses of Bury, Peterborough, and Ramsey had only two fulling mills between them (Bury had none), out of well over 150 mills they held in total; see Mills of Medieval England (n. 6 above), 156–57. The 10 percent figure arrived at through my own analysis is therefore not necessarily representative of the numbers of industrial mills held by the Benedictines as a whole. 53. The houses were Beaulieu, Furness, Old Wardon, Sibton, and Kirkstall; see Lucas, “Machinariarum Nihil Ex Deo,” chap. 5. However, three of the five recorded no indus- trial mills, even though one of these, Furness Abbey, was heavily involved in iron and lead mining. While there is archaeological evidence from Kirkstall and Bordesley Abbeys that industrial mills were not always recorded in manorial accounts and charters, an extensive survey I undertook of the English archaeological literature from 1950 to 2000 suggests that such underrecording was not very significant. 54. See, for example, White, Medieval Religion and Technology (n. 5 above), 67; Rey- nolds, Stronger than a Hundred Men (n. 5 above), 110–12, and “Medieval Roots of the Industrial Revolution” (n. 17 above), 109; Basalla (n. 5 above), 148; Major (n. 5 above), 232. 23 T E C H N O L O G Y A N D C U LT U R E if secular lords in England held only one industrial mill for every twenty- eight grain mills, and ecclesiastical lords one for every ten, it is hard to see how this might constitute an industrial revolution in medieval England. While monastic innovation in industrial milling does appear to be a dis- tinct possibility, the English evidence indicates that such activity does not of necessity entail an industrial revolution. Although comparatively com- JANUARY prehensive data are not available for France, the fact that only four out of 2005 ten of the industrial mills Bautier recorded were fulling mills, as opposed to VOL. 46 eight or nine out of ten in England, suggests that the number of industrial mills as a proportion of all powered mills must have been significantly higher in France than in England. This supposition is supported by more recent work on various regions of medieval France recently summarized by Paul Benoit and Joséphine Rouillard.55 A clue as to why France led in the application of waterpower to indus- try relates to the profitability of industrial mills as compared to grain mills. All of the evidence from England indicates that fulling mills were only about a third to a quarter as profitable as grain mills, making them less attractive to lords as investments.56 The profits to be had from other English industrial mills were even lower.57 Holt’s suggestion that the lower comparative profitability of such mills meant that they were only ever likely to be sited in areas where there was a plentiful supply of running water remains the most consistent with the evidence.58 With regard to the profitability of industrial mills in Italy and France, Muendel’s work on medieval Florence demonstrates that fulling mills were taxed at the same rate as, or at a higher rate than, grain mills, suggesting that they were earning as much as or more than Florentine grain mills.59 Benoit and Rouillard have pointed to evidence from the accounts of the Seneschal of Carcassonne that several forge mills belonging to the king of 55. Benoit and Rouillard (n. 43 above), 193–97, 208–14. See also Patrice Beck, “Wood, Iron, and Water in the Othe Forest in the Late Middle Ages: New Findings and Perspectives,” in Smith and Wolfe (n. 4 above), 173–84. 56. Holt, Mills of Medieval England, 156–58; Langdon, “Water-Mills and Windmills in the West Midlands” (n. 49 above), 435, and “Lordship and Peasant Consumerism” (n. 11 above), 14; Lucas, “Machinariarum Nihil Ex Deo” (n. 51 above), chap. 2 and apps. F and G. It is clear from this research that Carus-Wilson was wrong to claim more than sixty years ago that the English fulling industry was a sector of the medieval economy from which “a considerable profit could be derived”; see Carus-Wilson (n. 3 above), 52. 57. Langdon, “Water-Mills and Windmills in the West Midlands,” 436. Holt, Lang- don, and I have found a number of examples of lay and ecclesiastical lords licensing small entrepreneurs to operate and maintain industrial mills on their land for a nominal annual fee, suggesting that these mills were not earning large sums of money and were no threat to lordly revenues. 58. Holt, Mills of Medieval England, 158. 59. John Muendel, “The Distribution of Mills in the Florentine Countryside,” in Pathways to Medieval Peasants, ed. J. A. Raftis (Toronto, 1981), 89–95. 24  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds France were “valued at much higher prices than grain mills.” 60 Why these mills should have been more profitable in the Italian and French kingdoms seems most likely to have to do with population density and levels of de- mand, although further research is needed to develop a clearer picture of why this was the case. The fact that they do appear to have been more prof- itable than English industrial mills does, however, provide an explanation as to why French and Italian lords were more favorably inclined to building and maintaining them. Further exploration of some of Holt’s ideas on this subject would appear to be potentially fruitful.61 Summary of Findings from the Larger Sample A detailed analysis of the data contained in the larger sample suggests that the most intense period of industrial milling activity in medieval Europe was between the thirteenth and early sixteenth centuries. The region that we now know as France put waterpower to industrial uses ear- lier than any other Western European country (with the possible exception of Spain) and applied waterpower to a much greater variety of industrial processes than any other country as well. The regions that saw the most intense industrial milling activity appear to have been France and Italy. The main industries to which waterpower was applied in these two regions were cloth, hemp, leather, and timber, as well as some metallurgical processes. In the later Middle Ages, this extended to forging iron and pulverizing and polishing ores. In England, the most thoroughly studied region of medieval Europe, industrial milling was restricted primarily to fulling cloth, although there were a small number of industrial mills operating for tool sharpening and for extracting tannin from bark for curing leather. It was only in the late Middle Ages that waterpower was widely applied to metallurgy. Holt, Lang- don, and I have established that around 90 percent of English mills were dedicated to grinding grain. No more than 10 percent were put to indus- trial uses, and most of those were in the fulling industry. Only 1 percent of English mills were involved in activities that did not involve milling grain or fulling cloth. Most of England’s industrial mills were built where there was a plentiful water supply and a local wool market (as most were fulling mills), that is, in the West Midlands, southwest England, south Wales, and some parts of northern England. 60. Benoit and Rouillard, 195–96. To the best of my knowledge, there have been no data published to date on the earnings of English forge mills. 61. See Holt, “Mechanization and the Medieval English Economy” (n. 12 above), 149–56. For example, Holt’s observation that most industrial milling took place in small workshops producing for local markets undoubtedly holds true for England, but it is not so clear that it does for France or Italy. 25 T E C H N O L O G Y A N D C U LT U R E This evidence does not support the thesis that there was an industrial revolution based on waterpower in medieval England. But what of those areas of seemingly intense industrial milling activity in southern and northern France and northern Italy? Can they be said to have experienced an “industrial revolution”? In France, it was only certain well-developed regions in Dauphiné, Pro- JANUARY vence, Aude, the Forez Mountains, Champagne, Normandy, and Picardy 2005 where the extensive application of waterpower to industry took place. The VOL. 46 same appears to be true of northern Italy, with Piemonte, Pistoia, and Fir- enze showing the largest concentrations. If we examine the social and eco- nomic conditions that prevailed in those regions of modern-day France, Italy, England, Germany, and the Czech Republic that did begin to exten- sively use waterpower in the Middle Ages, we find that they already pos- sessed well-developed local industries that had access to regional, national, or international markets. They also had access to plentiful supplies of run- ning water that could be harnessed for industry. While large-scale hemp production in Dauphiné, sawmilling in the Forez region, and mining and metallurgy in the Harz Mountains, Saxony, and Bohemia were industries atypical of medieval Europe, the regions in which they developed did possess many of the features that appear to be necessary for technological innovation and diffusion, features which are more characteristic of the modern period. A more detailed examination of these regions may therefore provide some deeper insights into the processes of technological innovation and industrialization. The essays included in recent collections pertaining to technological development during the ancient and medieval periods have tended to rely on regional studies to illuminate these larger processes.62 It should be fairly obvious from the evidence presented here that such intensive regional applications of waterpower to industry do not constitute an industrial revolution of the Middle Ages, as proponents of that thesis have claimed. Furthermore, the assertion that the reason for the general lack of evidence for industrial milling in medieval England is that it was a technological backwater is not defensible in the light of a number of recent studies.63 There must, therefore, be some other reason or reasons for the 62. Astill and Langdon, Medieval Farming and Technology (n. 12 above); Wikander, Handbook of Ancient Water Technology (n. 4 above); Squatriti, Working with Water in Medieval Europe (n. 23 above). 63. At a recent medieval history conference a well-known historian of technology asserted that “England was a technological backwater during the Middle Ages” in response to comments by other scholars that the relative rarity of industrial mills in medieval England clearly problematized claims that there had been an industrial revolu- tion in the Middle Ages. Some recent studies which undermine that historian’s view are Bruce M. S. Campbell, James A. Galloway, and Margaret Murphy, “Rural Land-Use in the Metropolitan Hinterland, 1270–1339: The Evidence of Inquisitiones Post Mortem,” Agri- cultural History Review 40 (1992): 1–22; Bruce M. S. Campbell et al., A Medieval Capital 26  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds differences in the application of waterpower to industry in France, Italy, and England, which will most likely be revealed through a dialogue be- tween medieval historians who specialize in these regions—one that I hope to help stimulate with this article. While the evidence presented here does support the general claim made by supporters of the medieval industrial revolution thesis that there are definite continuities between the development of industry and automation in the Middle Ages and similar developments in the modern period, the evidence earlier presented pertaining to the Roman Empire, ancient and medieval China, and medieval Islamic societies points to a different set of continuities and discontinuities between their industrial activities and those of medieval Europe, making any attempt to posit a singular radical break between their respective technological advances appear increasingly implausible. While it would seem that a number of Roman advances in industrial milling were completely lost to subsequent civilizations, seem- ingly independent innovations by the Chinese and a number of Islamic societies appear to have provided the inspiration for subsequent industrial milling activities in Western Europe. It has been argued for some years by scholars of Islamic technology that the earliest use of water mills for ore-crushing, fulling cloth, and paper- making was in Islamic societies, and that a possible pathway for the diffu- sion of these technologies into Western Europe was through Islamic Spain, North Africa, and Byzantium. While a comparison of the earliest evidence for industrial milling in Western Europe with that claimed for contempo- raneous Islamic societies does suggest that this was the case, a systematic examination of the Islamic evidence remains to be conducted.64 For example, the miners of Samarkand were reputedly using stamping or ore-crushing mills as early as 973, whereas the earliest verifiable manu- script evidence from Europe for the same technology dates from 1317, from the Plauen region in Germany.65 However, two such mills may have been operating in northern Italy during the twelfth century.66 The Persians were reputedly using waterpower to full cloth as early as the tenth century.67 Meanwhile, not until the twelfth century in Europe did fulling workshops become a common industrial setting in which mills were deployed, even and Its Grain Supply: Agrarian Production and Distribution in the London Region c. 1300 (London, 1993); and Astill and Langdon, esp. Langdon, “Was England a Technological Backwater in the Middle Ages?” (n. 12 above), 275–92. 64. Thomas Glick, personal communication, April 2004. 65. Al-Hassan and Hill (n. 24 above), 54, 242–44; Hill, Studies in Medieval Islamic Technology (n. 27 above), V-184. 66. Bradford Blaine, “The Enigmatic Water-Mill,” in On Pre-Modern Technology and Science: A Volume of Studies in Honor of Lynn White Jr., ed. Bert S. Hall and Delno C. West (Malibu, 1976), 175, and “The Application of Water Power to Industry” (n. 11 above), 140–41. 67. Pacey, Technology in World Civilization (n. 27 above), 10–11. 27 T E C H N O L O G Y A N D C U LT U R E though the earliest reliably dated examples are from the first half of the eleventh century in France.68 The earliest use of waterpower for papermak- ing was reputedly in Samarkand in the eighth century, followed in the tenth century by Iraq, Iran, and Syria, while Spanish waterpowered paper mills date from the middle of the twelfth century.69 Italian, Spanish, and French paper mills date from the late thirteenth and early fourteenth centuries, JANUARY respectively, with the technology gradually spreading to the rest of Europe 2005 over the next 250 years or so.70 Most of the early well-documented cases VOL. 46 come from Italy, however. Patterns and Implications The research presented here clearly suggests that the thesis that an indus- trial revolution occurred during the Middle Ages rests, in its received form, on shaky foundations. The evidence that has been outlined from the Roman Empire, ancient and medieval China, and medieval Islamic countries demonstrates that there were indeed precursors to medieval European industrial uses of waterpower, although it remains unclear to what extent Europe was indebted to them. Medieval Islamic societies had undoubtedly established the requisite trade routes and technical traditions to transmit this knowledge to medieval Europe via Spain, North Africa, and Byzantium, but hard evidence that such a process actually occurred remains elusive. My own research and that of several other scholars on medieval Europe strongly indicate several patterns. Not all of Western Europe was equally innovative in its application of waterpower to new and existing industries. The systematic research that has been done on medieval England, Wales, and northern Italy demonstrates that there were marked regional differ- ences in both the uses of waterpower for industry and the chronologies of development of the relevant machinery. While there is some evidence for a revolutionary increase in the number of processes to which waterpower 68. With regard to various claims for fulling mills in Europe from the second half of the tenth century, see Forbes, “Power” (n. 5 above), 610; White, Medieval Technology and Social Change (n. 5 above), 84; Gille, “The Problems of Power and Mechanization” (n. 5 above), 456. None but White, however, provides sources. As Wikander has commented, “[t]he earliest alleged mills, Toscana 983, Dauphiné c. 990, Milano 1008, etc., are matters of doubt”; “Industrial Applications of Water-Power” (n. 4 above), 406, n. 27. As men- tioned above, I have not yet had an opportunity to verify the Benedictine example from 962 cited by Benoit and Rouillard. 69. Al-Hassan and Hill, 54; Pacey, 10–11, 42; Hill, Studies in Medieval Islamic Tech- nology, XVIII-10; Gille, “The Problems of Power and Mechanization,” 456–57. 70. Blaine, “The Application of Water Power to Industry,” 103–15. Although the Chinese invented the papermaking process, they do not seem to have applied water- power to it until around 1690, which would imply that they adopted this technology either indirectly from Western Europe or directly from the Islamic Near East. Needham (n. 9 above), 394. 28  LUCASK|KIndustrial Milling in the Ancient and Medieval Worlds was applied during the thirteenth and fourteenth centuries, that evidence is mainly restricted to France. A handful of regions in southern and northern France appear to have been centers of medieval technological innovation; these saw the development of new applications for industrial milling and were the only regions of medieval Europe to apply waterpower to some industries.71 Although other parts of medieval Europe had some involvement in industrial milling, including the more advanced regions of what we now call Italy, England, and Germany, that involvement appears to have been quite limited until the late thirteenth and fourteenth centuries. Even then it was largely restricted to a small number of industries, most notably the production of broad-fiber woolen cloths and metallurgy. The profitability of industrial milling compared with grain milling appears to have varied considerably between England, France, and Italy. Whereas English fulling mills generally earned only a half to a quarter as much as grain mills, Italian fulling mills appear to have been just as prof- itable, or even more profitable, than grain mills. In France, forge mills held by the king appear to have earned handsome profits, whereas there are no such indications from any of the equivalent mills in medieval England. In addition to its negative implications for the industrial revolution thesis, this research points to a pattern of industrial development in medieval Europe that may prove far more significant in its implications. Rather than illustrating a pan-European industrial revolution in which there were a few pockets of “technological retardation,” the evidence for the development of waterpowered industry in medieval Europe suggests that the very opposite was the case.72 Not only do most of the important inno- vations in industrial milling appear to have originated in earlier Islamic societies, ancient China, or the Roman Empire, those regions of medieval Europe that were engaged in industrial milling appear to have been geo- graphical pockets of technological innovation within a broader environ- ment of technological incrementalism.73 It would seem that these centers of innovation were able to develop because of characteristics they shared with centers of trade and commerce in the early modern period and later—that is, they had relatively dense populations, were resource rich, already had well-developed industries in key areas of high demand, and were geo- graphically well-placed to exploit local, regional, or international markets. 71. Considering that medieval France is considered in some quarters to have been technologically backward, the finding that with respect to all seven of the most com- monly appearing types of industrial water mill France was a clear leader should give some pause to those who have held this view in the past. 72. White argues for such a pan-European industrial revolution in Medieval Tech- nology and Social Change, 88. 73. This is not to say that these were the only geographical areas of technological in- novation in the later medieval period. 29 T E C H N O L O G Y A N D C U LT U R E Of equal importance, they all had plentiful supplies of running water, which they could exploit for purposes other than meeting the local demand for grain milling. Further research might profitably concentrate on these exceptional areas and, aided by the insight that they were, in fact, centers of innovation, might help answer a number of key questions about the transition from JANUARY medieval to modern industry. For example, which social groups were the 2005 main investors in different medieval industries? Did the cost of labor and VOL. 46 the availability of waterpower for purposes other than grain milling play major roles? What other social, economic, and environmental factors shaped the development of medieval industries in particular regions? How profitable were different medieval industries, and what levels of ongoing investment and innovation were required? To what extent did profitability vary from region to region and country to country? Did those industrial regions that were successful in the medieval period survive into the early modern period and beyond? In addressing these questions, the most fruitful subjects for study would appear to be France, Italy, Spain, England, and Germany. As we have seen, although they present very different cases, the possibility that other parts of medieval Europe significantly surpassed these regions in the application of waterpower to industry is remote. Further interdisciplinary study along these lines will help to illuminate such important issues as the factors nec- essary for technological innovation and diffusion in the premodern period and the extent to which industrial development in the modern period was built on ancient and medieval foundations. 30