TY - JOUR
AU - SAITO,, Osamu
AB - Abstract While early-modernists tend to believe that the period around 1600 saw a structural break in Japan’s economic history, research by medievalists since the mid-1970s has suggested that a market economy was on the rise from the 14th century onwards. This article examines several listings of occupations and commodities compiled between the 13th and the early 17th centuries to see if a proliferation of non-agricultural activities was underway before c.1600. As Adam Smith envisaged, the separation of one productive activity from another, i.e. an increasing division of labour, will lead to economic growth. Since this type of market-led change, often called Smithian growth, has been considered an ‘early modern’ phenomenon by economic historians, it is worth examining if there were signs of an increasing division of labour in late medieval Japan. The article’s findings indicate that some signs of occupational and product differentiation appeared by the 16th century. As far as Smithian growth is concerned, therefore, the article concludes that there was some degree of continuity between the medieval and the early modern periods. Division of labour, Market economy, Medieval Japan, Smithian growth, Textiles 1. The Issue: The Medieval Origins of Smithian Growth It has been widely recognised that many parts of Eurasia experienced modest but unmistakable growth before the onset of modern economic growth. According to the Maddison Project Database, gross domestic product (GDP) per capita in 12 western European countries grew at 0.14% per annum between 1600 and 1700 and at 0.18% between 1700 and 1850 (Bolt and van Zanden 2014: 637). Two oft-mentioned covariates of such growth are urbanisation and the growth of rural industry and trade; proponents of the proto-industrialisation thesis emphasise that fully fledged industrialisation was preceded by the latter, while other scholars see the former as a fitting predictor of longer term change in pre-modern growth. However, we may take a somewhat different approach by referring to what Adam Smith observed on the eve of the industrial revolution. The opening chapter of The Wealth of Nations begins with the statement that the productivity of labour increases with the division of labour, and it is suggested that in a ‘commercial society’, i.e. a social order in which the market is the dominant mode of exchange, even a simple woollen coat is a product of the ‘joint labour of a great multitude of workmen’. The number of those workmen involved is surprisingly large: The shepherd, the sorter of the wool, the wool-comber or carder, the dyer, the scribbler, the spinner, the weaver, the fuller, the dresser, with many others, must all join their different arts in order to complete even this homely production. How many merchants and carriers, besides, must have been employed in transporting the materials from some of those workmen to others who often live in a very distant part of the country!’ (Smith 1976, I: 22–23). He goes on to list more—ship-builders, sailors, sail-makers and rope-makers to bring together all the materials utilised, and also mill-wrights, forgers, smiths, producers of tools and craftsmen who built machines and furnaces, as well as miners, timber-fellers and charcoal-burners for the furnaces, to produce the tools and instruments used. Note that his list includes timber-fellers, who are classified in the primary sector, and also merchants, carriers and sailors of the tertiary sector. The ‘division of labour is limited by the extent of the market’ (Smith 1976, I: ch. I.3); but once the market size becomes large enough the tasks and occupations of workmen start proliferating, resulting in the growth of ‘intermediate’ market transactions between producers, as well as transactions between merchants and consumers. Since this increased ‘roundaboutness’ of production generates increasing returns in the economy at large (Young 1928), economic growth is generally associated with an increasing division of labour. In other words, the proliferation of occupations and marketed products, along with the emergence of new markets for intermediate goods for other producers, are useful measures of economic development for periods before modern official statistics became available. Smithian growth is defined as this kind of market-led economic growth, and there is consensus that Smithian growth took place in early modern Eurasian countries (Saito 2005, 2013). In western Europe, for example, specialisation took place in textiles between spinning, weaving and sewing and tailoring, while the making of, for example, scythes and other farming tools emerged in areas away from the sites of iron smelting. For Japan too, there is evidence that the extent of the division of labour increased in the second, if not the first, half of the Tokugawa era (1603–1868). The first half was a period of urbanisation as well as population growth and also saw the sudden rise and fall of the silk-for-silver trade, so that growth faltered if measured on a per capita basis. According to recent estimates of pre-1874 GDP (Saito and Takashima 2016), the average rate of per capita GDP growth for the 1600–1721 period was as low as 0.01% per annum. From the 18th century onwards, however, rural industry and trade grew—as Thomas Smith (1988)’s seminal work demonstrated—at the expense of their metropolitan counterparts, while in both cotton and silk textiles the geographical separation of production processes was underway (Saito 1983). As a result, GDP per capita grew at an annual average rate of 0.23% between 1721 and 1846. It was a gradual process of sectorally balanced growth and matched or marginally exceeded the growth performance of 0.22% for the three core European countries, i.e. Belgium, Britain and Holland, in the 1700–1800 period (the two growth rates are calculated from Saito and Takashima 2016: tables 2 and 3, 380, 382). How far can we trace back this Smithian growth trajectory in the Japanese case? This is the question I would like to explore in this article. Early-modernists, such as Hayami Akira, have long portrayed the Tokugawa regime as a system radically different from the pre-1600 past, suggesting that it was in the 17th century that an ‘economic society’—similar to Adam Smith’s ‘commercial society’—made inroads into the provincial regions outside the Kyoto area; the change coincided in most cases with the take-off of population growth (Hayami 2004: 7–12). According to Hayami, medieval Japan was a stagnant, subsistence economy with agriculture being the predominant form of economic activity. However, many medievalists believe that the real break with the ancient past came in the 14th century. Amino Yoshihiko (1974) once questioned the traditional view, suggesting that the 14th century was a period of considerable commercialisation. He also commented on Hayami’s thesis, noting that market transactions and the use of money and credit were changes already observable in the period after the Mongol invasions of 1274 and 1281 (Amino 1977; 2001). Now there are more pieces of evidence about those phenomena. Following lively discussions in the 1990s, consensus was reached that while the government had no interest in stabilising currency circulation by issuing its own coins, vast quantities of Chinese coins were in circulation, with the eventual use of credit becoming common. From the early 14th century onwards, various types of commercial bills came into existence and were used widely for transferring payments to dealers in distant places and for settling credit–debt accounts (Sakurai 2002, 2008; Segal 2011: ch. 4). It is, therefore, no exaggeration to say that ‘the era from 1280 to 1450 witnessed the birth of a fully-fledged commercial economy’ (Farris 2006: 142). For the history of market growth in pre-modern Japan, the year 1600 was certainly not a real break; and the medieval monetary practices may well have played at least some part in laying the groundwork for the monetisation drive in early modern times (Segal 2010). More recently, progress has also been made in quantitatively oriented economic history: we are now in a better position to discuss medieval macro-economic conditions in terms of GDP per capita and some other measures, as exemplified by chapters in the medieval volume of the Iwanami Lectures on the History of Japanese Economy (Fukao, Nakamura and Nakabayashi 2017).1 Currently available estimates for output are still a little speculative, but it is important to realise that new population estimates indicate somewhat more persuasively that population growth took off before 1600, suggesting that medieval Japan since the 14th century can no longer be portrayed as a stagnant society (see section 2 below). To put it another way, both recent quantification efforts and descriptive findings by medievalists are pointing in the same direction. In the following section, I shall make a brief survey of long-term trends so far identified over the four-century period. Section 3 turns to the examination of occupational descriptors by sector of the labour force. Images of sellers, makers and other tradespeople (shokunin) are available from four medieval illustrated scrolls: the first two coming from the 13th century, and the second two from the end of the 15th century. Section 4 examines similar tables for commodities by sector of production, compiled from publications which contain listings of regional specialities—one for c.1400 and the other for 1638. Section 5 gives a brief survey of evidence concerning signs of Smithian growth, focussing on textiles. The final section concludes by placing those findings in a longer-term perspective of market growth. 2. Long-Term Trends Japan’s transition from an ancient regime began in the late 11th century. By that time, a state system with public ownership of land and people, introduced from China in the seventh century, had ceased functioning properly. The 11th century saw the rise of warriors as political powers. A war that broke out in 1180 between two rival warrior clans led eventually to the establishment of the first warrior government (the Kamakura shogunate) in 1185. The new era, however, did not lead to a centralisation of power in the hands of the warrior class; institutions of imperial-aristocratic rule, though weakened, remained in place. After the Mongols’ unsuccessful invasions, warriors gained more political influence. However, it meant more warfare between them. From the mid-15th century on, a prolonged period of warfare began. This 150-year-long warring-states period ended in the 1590s with a series of political as well as military settlements. After the battle of Sekigahara in 1600, the Tokugawa family came to power and formed a new shogunate in 1603. The 1180–1590 period was characterised by a lack of the centralised machinery of government. When the first warrior government was formed in the east, it functioned within the framework of Kyoto-based imperial rule, with economic management based largely on autonomous estates or manors (shōen) owned by aristocrats and religious institutions. The next government, the Muromachi shogunate, moved its administrative headquarters to Kyoto but failed to maintain full authority over the entire country. Thus, the final century of their rule saw provincial overlords (daimyo) warring with each other; in the process of war-making, many of them gained considerable, independent control over the lands they held (Hall 1990). In the middle of the twelfth century, the country was still sparsely populated. Population density in 1150 is estimated to have been around 0.2 persons per hectare (Saito and Takashima 2017: 61), and the size of population remained static until the end of the 13th century. The first column of Table 1 sets out estimates of changing population growth rates from 1150 to 1846.2 The population started to pick up sometime between 1280 and 1450 and continued to increase into the next century, with an average annual growth rate of a little higher than 0.3%. The density level rose to 0.35 in 1450 and to 0.6 in 1600. The growth rate accelerated to 0.51% in the 1600–1721 period. The population density reached 1.1 in 1721, after which demography became more or less static again, with the rate of population increase falling as low as 0.03%. In this long demographic cycle, the medieval era was a phase of start-up. It was a fitful process of growth, but the 16th century saw the growth start accelerating. Table 1. Growth Rates of Population and Output, 1150–1846 Period . Average Annual Rate of Growth (%) . . . . . Population . Primary-Sector Output . GDP . GDP per Capita . 1150–1280 0.01 −0.07 −0.05 −0.06 1280–1450 0.31 0.31 0.33 0.02 1450–1600 0.35 0.41 0.48 0.13 1600–1721 0.51 0.38 0.52 0.01 1721–1846 0.03 0.25 0.25 0.23 Period . Average Annual Rate of Growth (%) . . . . . Population . Primary-Sector Output . GDP . GDP per Capita . 1150–1280 0.01 −0.07 −0.05 −0.06 1280–1450 0.31 0.31 0.33 0.02 1450–1600 0.35 0.41 0.48 0.13 1600–1721 0.51 0.38 0.52 0.01 1721–1846 0.03 0.25 0.25 0.23 Sources: Saito and Takashima (2017: 61) for population; Bassino et al. (2019: tables 5 and 7) for primary-sector output, GDP and GDP per capita. Note: In the 1150–1450 period the estimates of population and primary-sector output, on which per capita GDP estimates are based, are given with a range. In both cases for 1280 and 1450, the difference from the average of the high (H) and low (L) estimates expressed as percentage of the average happens to be ±4%. For population in 1150 it is ±7%, but for primary-sector output in 1150 the range is as wide as ±44%. The growth rates shown above are computed with the H-L averages. For the H and L estimates, see Saito and Takashima (2017: 61) and Takashima (2017: 101). Open in new tab Table 1. Growth Rates of Population and Output, 1150–1846 Period . Average Annual Rate of Growth (%) . . . . . Population . Primary-Sector Output . GDP . GDP per Capita . 1150–1280 0.01 −0.07 −0.05 −0.06 1280–1450 0.31 0.31 0.33 0.02 1450–1600 0.35 0.41 0.48 0.13 1600–1721 0.51 0.38 0.52 0.01 1721–1846 0.03 0.25 0.25 0.23 Period . Average Annual Rate of Growth (%) . . . . . Population . Primary-Sector Output . GDP . GDP per Capita . 1150–1280 0.01 −0.07 −0.05 −0.06 1280–1450 0.31 0.31 0.33 0.02 1450–1600 0.35 0.41 0.48 0.13 1600–1721 0.51 0.38 0.52 0.01 1721–1846 0.03 0.25 0.25 0.23 Sources: Saito and Takashima (2017: 61) for population; Bassino et al. (2019: tables 5 and 7) for primary-sector output, GDP and GDP per capita. Note: In the 1150–1450 period the estimates of population and primary-sector output, on which per capita GDP estimates are based, are given with a range. In both cases for 1280 and 1450, the difference from the average of the high (H) and low (L) estimates expressed as percentage of the average happens to be ±4%. For population in 1150 it is ±7%, but for primary-sector output in 1150 the range is as wide as ±44%. The growth rates shown above are computed with the H-L averages. For the H and L estimates, see Saito and Takashima (2017: 61) and Takashima (2017: 101). Open in new tab As I have suggested elsewhere, a crucially important breakthrough is likely to have taken place only after the demographic threshold had been crossed (Saito 2016: 168). If this was the case in the Japanese archipelago, then the increase in population density between 1450 and 1600 must have had a positive effect on market growth. Also, the above-mentioned structural defect inherent in the medieval state system is likely to have left room for merchants and tradesmen to gain increasing freedom from political dominance. They enhanced their presence in commerce and trade, which led to the formation of guild-like institutions called za; and their monopolistic or monopsonistic power sometimes hindered spontaneous growth of the market. However, the long-term trend was for restrictions on trade to be removed, which was unmistakable after 1280 and became more marked under warring daimyo rule in the 16th century (Yamamura 1973; Farris 2006: 235–242). Given the demographic evidence, it is to be expected that the pace of urbanisation must have followed a pattern similar to that of population increase. In medieval times, especially before the 16th century, there existed only 10 or so cities with population size exceeding 10,000 (Saito and Takashima 2017: 74). The sum of inhabitants in those urban settlements, expressed as a percentage of the total population, stayed below 3% during the 11th through the 15th century. It was during the 16th century that the proportion exceeded the 5% mark and reached 6% in 1600. However, it is likely that if settlements with fewer than 10,000 inhabitants were included in the calculation, then the urbanisation trend may well have looked different. Decades ago, Amino (1976) tried to broaden the ‘urban’ category by including much smaller towns. The reason is that in this period a vast majority of the thriving market towns in the countryside and the bustling port towns on the coast were small in size, with populations usually less than 5,000. There is a mid-15th century freight report of ships entering Hyōgo, an important port on the Seto Inland Sea and Osaka Bay.3 The transport route linking the Kyoto-Nara area through this passage of water to the northern Kyushu coast on the Sea of Japan had been busy since ancient times. As this customs record tells us the names of ports of loading (as well as information about ship cargoes), it is possible to calculate how many ships came from those port towns.4 Table 2 shows the breakdowns. The listed ports numbered 107. Of this total, three were from unidentified locations, seven from east of Hyōgo on the Osaka Bay coast, and the rest from places west or south; of the latter, 68 were located on the Inland Sea coasts and islands, 230 km or less west of the port of discharge, and 29 were from places outside the Inland Sea region. At the western edge of the second group of ports was Mihara, and the region’s north side between Mihara and Hyōgo was particularly dense with small trading centres, implying that this 230-km-long coastline was lined with port towns at intervals of 6 km. Turning to the information about the number of cargo ships each port town sent, the table shows that about 60% of the total home towns listed (excluding the three unknown cases) sent 1–10 ships to Hyōgo, a little over than 30% sent 11–50 ships, 8% sent 51–100 ships, and only 3% sent 101-plus ships. The 51–100 category included Onomichi, with a recorded number of 62 ships. This was a medium-sized town on the north coast of the inland waters, whose population size is known to have been 5,000 in the 14th century (Saito and Takashima 2017: 74). If this 62-to-5000 ratio can be a rough guide for other cases, then a substantial proportion of the 104 places must have been in the category of 1,000 people or fewer. In the Seto Inland Sea region, therefore, it is likely that such small towns had already been mushrooming in the early 15th century. It is also likely that these small towns, not metropolitan cities, started growing first, and that the growth took off before 1445. Table 2. Home Ports of Ships Entering Hyōgo According to the 1445 Freight Report Region . Total . No. of Ports by no. of Ships Entering Hyōgo . . . . . . 1–10 . 11–50 . 51–100 . 101+ . East of Hyōgo 7 3 2 1 1 West of Hyōgo 68 39 22 5 2 Further west and south 29 19 8 2 0 Total 104 61 32 8 3 Region . Total . No. of Ports by no. of Ships Entering Hyōgo . . . . . . 1–10 . 11–50 . 51–100 . 101+ . East of Hyōgo 7 3 2 1 1 West of Hyōgo 68 39 22 5 2 Further west and south 29 19 8 2 0 Total 104 61 32 8 3 Source: Muto (1981: appendix table). Note: Three ports of loading whose location cannot be identifies are excluded. ‘East of Hyōgo’ corresponds Settsu province (including Hyōgo itself), ‘West of Hyōgo’ covers Harima, Awaji, Bizen, Bicchū, Bingo and Sanuki provinces, and ‘Further west and south’ includes Aki, Suō, Nagato, Awa, Tosa, Iyo and Buzen provinces. Open in new tab Table 2. Home Ports of Ships Entering Hyōgo According to the 1445 Freight Report Region . Total . No. of Ports by no. of Ships Entering Hyōgo . . . . . . 1–10 . 11–50 . 51–100 . 101+ . East of Hyōgo 7 3 2 1 1 West of Hyōgo 68 39 22 5 2 Further west and south 29 19 8 2 0 Total 104 61 32 8 3 Region . Total . No. of Ports by no. of Ships Entering Hyōgo . . . . . . 1–10 . 11–50 . 51–100 . 101+ . East of Hyōgo 7 3 2 1 1 West of Hyōgo 68 39 22 5 2 Further west and south 29 19 8 2 0 Total 104 61 32 8 3 Source: Muto (1981: appendix table). Note: Three ports of loading whose location cannot be identifies are excluded. ‘East of Hyōgo’ corresponds Settsu province (including Hyōgo itself), ‘West of Hyōgo’ covers Harima, Awaji, Bizen, Bicchū, Bingo and Sanuki provinces, and ‘Further west and south’ includes Aki, Suō, Nagato, Awa, Tosa, Iyo and Buzen provinces. Open in new tab Such changes may have been accompanied by an increase in non-agricultural output. The recent article by Jean-Pascal Bassino et al. (2019) gives sector-specific estimates of real output over the 1100-year period, from which three columns of Table 1 are taken. The average annual rate of GDP per capita was 0.02% between 1280 and 1450 and 0.13% between 1450 and 1600, suggesting that there was non-negligible growth acceleration in the final century-and-a-half period. However, a close look at the ways in which these estimates were made reveals that the proportions of the secondary and tertiary sectors in the total output were derived by applying an estimation formula with two population-related predictors, one being the rate of urbanisation (calculated for settlements of over 10,000 people), and the other population density (Bassino et al. 2019: section 4.2).5 This means that an independent check cannot be made as far as non-agricultural growth is concerned. The column in Table 1 for primary-sector output growth, on the other hand, deserves scrutiny since it was estimated from the demand side (Bassino et al. 2019: sections 3.2–3.3; Takashima 2017: ch. 2). According to those estimates, while the sector exhibited a poor performance in the pre-1280 period and managed just to keep up with population increase in the 1280–1450 period, it appears as if agriculture’s productive capacity suddenly improved after 1450 and kept performing better during the warring-states era. In fact, if compared with the achievement of the early Tokugawa period, it is a little puzzling to see agricultural production growing faster than population in this particular period. This was an age of reclamation, owing partly to warring daimyo’s civil engineering efforts and partly to the increasing acceptance by farmers of Champa rice, an indica variety which could withstand marshy, lowland delta conditions. It should also be noted that during the 16th century there occurred a sharp drop in the frequency of famines, seemingly consistent with the finding that the agrarian growth in the late medieval period was sustainable (Takashima 2017: 115–116, quoting Saito 2015 for famine frequencies). However, while what is said about land reclamation and the indica rice is correct, the inference from the famine frequency series is not. The index of famine frequency is intended to measure changes in the frequency of mass starvation and crisis mortality. Thus, reforms introduced by warring daimyo may well have been ‘important in preventing a poor harvest from developing into a regional or cross-regional famine’, but not necessarily in reducing the incidence of harvest failure (Saito 2015: 223). In other words, the 16th-century agrarian economy may have been still subject to periodic harvest failures. It is therefore likely that the estimated growth rates of the primary sector and hence of GDP per capita in the 1450–1600 period are a little too high. That said, however, one statement remains intact: the market economy was on the rise from the 14th century onwards. The question is whether a structural break came in at the end of the 16th century. The period around 1600 saw a couple of new developments. First, there occurred a sudden rise in the output of silver from domestic mines, followed by a similarly strong increase in silver exports.6 The export boom continued until the 1630s, when the Tokugawa shogunate decided to put overseas trade under stiflingly tight control, which resulted in a substantial decline in overseas trade. Second, the century-long process of unification of the realm was completed by the Tokugawa family, and the construction of new infrastructure followed. Soon after the 1600 battle, the building of castle towns started, and this country-wide construction boom went on until the second half of the 17th century, as did population growth. According to the traditional view, therefore, the turn of the 17th century was a transition period from the medieval to the early modern. To what extent this characterisation is tenable is an issue I would like to return to in the final section. 3. The Proliferation of Occupations For medieval historians, the number of surviving documents is deplorably sparse. During the four-century period, no surveys of people were taken by government offices, nor by any religious institutions. No trial papers or inheritance documents exist in any archives or libraries from which information about ordinary people’s occupations might be collected in a comprehensive manner. In this respect, shokunin uta-awase scrolls—illustrated scrolls of an imaginary poetry competition (uta-awase) in which people in various trades are presented in pairs, accompanied by poems featuring their work, life and environment—can be an important source for showing how proliferated occupations were in medieval times. There are four extant tradespeople scrolls: two compiled in the 13th (1214 and 1261), and the other two at the end of the 15th century (1494 and 1500). In modern Japanese usage the word shokunin means craftsman, but in medieval times, as Amino (1983) emphasised, the term could cover those in services, or anyone outside agriculture. In fact, singers, dancers, blind reciters, peddlers, fortune-tellers and physicians, together with metal-workers, weavers, dyers and sake (rice wine) brewers, are all found on the illustrated scrolls. Some believe that they are probably the ‘best sources available’ for the study of non-agricultural occupations of the day (Toyoda and Sugiyama 1977: n.17 on 141), as service occupations are well represented and females better represented in those scrolls than in any other documents of the pre-modern period. The drawback, on the other hand, is that each poetry competition was written as a parody or satire by Kyoto nobles, in which the choice of paired occupations was merely the artists’ exploration of a genre theme. These were products of Kyoto nobles’ imaginations, influenced heavily by their own interests and experiences of city life, so that it is impossible to compile a list of medieval trade titles in a representative manner. Moreover, it is not unlikely that the later-day writer-artist was motivated to create something to surpass the earlier poetry competitions by choosing paired occupations which had not appeared previously. If this were the case, then the occupational distribution for the mid-13th century would probably be better represented by the sum of the 1214 and 1261 numbers than by the 1261 list alone, and that for the late-15th century by the 1494 and 1500 numbers put together. This will certainly allow us to trace how the artists’ imagined distribution of occupations changed from the mid-13th to the late 15th century. Table 3 shows the thus calculated sectoral allocations of trade titles for 1261 and 1500. In order to read the table, a few points must be made. First, given the nature of the source material, there is no occupational descriptor for farming, except in the 1494 source. All primary-sector entries in the other three scrolls are for those in mountains or in fishing and salt making (before the emergence of specialised salt farms, the making of salt had always been a part-time activity of the fishing family). According to the 1500 scroll,7 two descriptors are used here: yamabito and urabito, meaning literally mountain people and beach people, respectively. The verses composed for the trades suggest that they were engaged in not just one but multiple occupational activities. The former did hunting, gathering, felling trees, collecting firewood, and in most cases cultivating millet for themselves (also charcoal burning, but in the 1500 scroll this was listed separately), while the latter is supposed to have been engaged in fishing and making salt from seawater. However, it seems likely that the medieval writer-artists’ image of them was very much stereotyped. There is evidence that the medieval maritime economy was no longer so archaic, with emerging wealthy shipping agents among those populations and some settlements becoming bustling towns (Amino 2001; see also Iwasaki et al. 1993: 496). For the non-primary sectors the listing of trades is far more specific; nevertheless, those in dealing and wholesaling, as distinguished from selling, are hardly mentioned in the scrolls. Table 3. Sectoral Allocations of Occupational Descriptors, 1261 and 1500 Sectors and Ratios . 1261 . 1500 . A. Number of descriptors  Primary sector 3 7  Secondary sector 21 66   Makers of processed and assembled goods 19 58   Rest of secondary 2 8  Tertiary sector 25 102   Sellers of processed and assembled goods 1 30   Entertainers and other service-providers 21 52   Rest of tertiary 3 20 Total 49 175 B. Ratios  1. Secondary/ non-primary total 0.5 0.4  2. Tertiary/ non-primary total 0.5 0.6  3. (Makers+sellers)/ non-primary total 0.4 0.5  4. Sellers/ makers 0.1 0.5  5. Entertainers and other services/ non-primary total 0.5 0.3 Sectors and Ratios . 1261 . 1500 . A. Number of descriptors  Primary sector 3 7  Secondary sector 21 66   Makers of processed and assembled goods 19 58   Rest of secondary 2 8  Tertiary sector 25 102   Sellers of processed and assembled goods 1 30   Entertainers and other service-providers 21 52   Rest of tertiary 3 20 Total 49 175 B. Ratios  1. Secondary/ non-primary total 0.5 0.4  2. Tertiary/ non-primary total 0.5 0.6  3. (Makers+sellers)/ non-primary total 0.4 0.5  4. Sellers/ makers 0.1 0.5  5. Entertainers and other services/ non-primary total 0.5 0.3 Sources: Mori (1979) for the 1214, 1261 and 1494 scrolls, and Iwasaki et al. (1993) for the 1500 scroll. Note: Panel A is a substantially revised version of table 1.6 in Saito and Takashima (2017: 80). Column ‘1261’ shows the 1261 numbers added on the 1214 scroll and column ‘1500’ shows the 1500 numbers added on the 1494 scroll. The total number of occupational descriptors in each scroll was 24, 25, 33 and 142, respectively. Cases in which the descriptor and other indicators of the trade do not agree with each other are allocated to sectors according to the suffix attached to the occupational descriptor (‘-tsukuri’ or ‘-uri’). ‘Non-primary total’ denotes the sum of the secondary- and tertiary-sector totals. Open in new tab Table 3. Sectoral Allocations of Occupational Descriptors, 1261 and 1500 Sectors and Ratios . 1261 . 1500 . A. Number of descriptors  Primary sector 3 7  Secondary sector 21 66   Makers of processed and assembled goods 19 58   Rest of secondary 2 8  Tertiary sector 25 102   Sellers of processed and assembled goods 1 30   Entertainers and other service-providers 21 52   Rest of tertiary 3 20 Total 49 175 B. Ratios  1. Secondary/ non-primary total 0.5 0.4  2. Tertiary/ non-primary total 0.5 0.6  3. (Makers+sellers)/ non-primary total 0.4 0.5  4. Sellers/ makers 0.1 0.5  5. Entertainers and other services/ non-primary total 0.5 0.3 Sectors and Ratios . 1261 . 1500 . A. Number of descriptors  Primary sector 3 7  Secondary sector 21 66   Makers of processed and assembled goods 19 58   Rest of secondary 2 8  Tertiary sector 25 102   Sellers of processed and assembled goods 1 30   Entertainers and other service-providers 21 52   Rest of tertiary 3 20 Total 49 175 B. Ratios  1. Secondary/ non-primary total 0.5 0.4  2. Tertiary/ non-primary total 0.5 0.6  3. (Makers+sellers)/ non-primary total 0.4 0.5  4. Sellers/ makers 0.1 0.5  5. Entertainers and other services/ non-primary total 0.5 0.3 Sources: Mori (1979) for the 1214, 1261 and 1494 scrolls, and Iwasaki et al. (1993) for the 1500 scroll. Note: Panel A is a substantially revised version of table 1.6 in Saito and Takashima (2017: 80). Column ‘1261’ shows the 1261 numbers added on the 1214 scroll and column ‘1500’ shows the 1500 numbers added on the 1494 scroll. The total number of occupational descriptors in each scroll was 24, 25, 33 and 142, respectively. Cases in which the descriptor and other indicators of the trade do not agree with each other are allocated to sectors according to the suffix attached to the occupational descriptor (‘-tsukuri’ or ‘-uri’). ‘Non-primary total’ denotes the sum of the secondary- and tertiary-sector totals. Open in new tab Second, given the omission of most primary-sector occupational descriptors, it is erroneous to discuss time trends by referring to changing proportional shares computed in relation to the total. Accordingly, all ratios in section B of Table 3, except for the sellers/makers ratio, are defined with respect to the total of the non-primary sectors. Third, in the tradespeople scrolls, there are three clues for how and by whom the trade was carried out: the trade title, an illustration of the work being conducted, and words spoken by the character in the picture. By looking at these together, we learn that not all the trades listed are specialised makers or sellers; many of them were maker-seller family economies. For example, the sōmen (noodle) trade in the same scroll is accompanied by the image of a woman tending sōmen being dried in the sun (Figure 1), but the entry bears a trade title that ends with ‘-uri’, meaning ‘seller’. In the case of kawarake, the title’s suffix is ‘-tsukuri’, meaning ‘maker’, and the accompanied text indicates that the trade was the making of coarse, unglazed earthenware. However, as Figure 2 shows, it is a male peddler selling pots and bowls. Similarly, Figure 3 depicts the sake trade, showing a woman trying to attract the attention of passers-by, but the trade name bears the suffix ‘maker’. Since what was traded is unfiltered or half-filtered sake (nigori or usu-nigori), it was probably home-made with the woman herself involved in the brewing process. These three cases have come to my attention because of these disagreements, but there is no reason to assume that the households embracing both makers and sellers were confined to those three trades. It is important to understand that the suffixes for maker and seller were two sides of the same coin in the family mode of production. Figure 1. Open in new tabDownload slide Sōmen-Seller From the 1500 Tradespeople Scroll. Source: Iwasaki et al. (1993: 77). Note: Sōmen is much thinner than ordinary noodles, but this woman says: ‘I’m going to make this thick’. Figure 1. Open in new tabDownload slide Sōmen-Seller From the 1500 Tradespeople Scroll. Source: Iwasaki et al. (1993: 77). Note: Sōmen is much thinner than ordinary noodles, but this woman says: ‘I’m going to make this thick’. Figure 2. Open in new tabDownload slide Earthenware-Maker From the 1500 Tradespeople Scroll. Source: Iwasaki et al. (1993: 37). Note: This male peddler solicits passers-by, saying ‘Buy our red earthenware! On the way back home—I’ll give you a discount!’ Red-clay pottery was used for serving pots and plates. Figure 2. Open in new tabDownload slide Earthenware-Maker From the 1500 Tradespeople Scroll. Source: Iwasaki et al. (1993: 37). Note: This male peddler solicits passers-by, saying ‘Buy our red earthenware! On the way back home—I’ll give you a discount!’ Red-clay pottery was used for serving pots and plates. Figure 3. Open in new tabDownload slide Sake-Maker From the 1500 Tradespeople Scroll. Source: Iwasaki et al. (1993: 15). Note: The woman selling sake (unfiltered, presumably) calls out: ‘Come, come! Taste our sake! We also have a now popular half-filtered one’ (paraphrased from Tabata 1999: 104). The half-filtered (usu-nigori) sake was considered better in quality than the unfiltered (nigori). Figure 3. Open in new tabDownload slide Sake-Maker From the 1500 Tradespeople Scroll. Source: Iwasaki et al. (1993: 15). Note: The woman selling sake (unfiltered, presumably) calls out: ‘Come, come! Taste our sake! We also have a now popular half-filtered one’ (paraphrased from Tabata 1999: 104). The half-filtered (usu-nigori) sake was considered better in quality than the unfiltered (nigori). These considerations have two implications. One is that multi-employment family economies were numerous. There were not only maker-seller families, which made contributions to both secondary- and tertiary-sector activities, but also a large number of farm households, which were likely to have had a family member or two with another employment in manufacturing, commerce or transport. A vast majority of them did not feature in the illustrated poetry competition. But the urabito family, which appeared in the tradespeople scrolls, is a case in point: their principal occupation could be in the primary sector with salt making, a subsidiary activity, in the secondary, or the other way around. The same can be said about the yamabito family, in which case the making of charcoal was their secondary-sector activity. Moreover, there is evidence from sources other than the tradespeople scrolls. According to Amino (1983), in the early-medieval estate economy, there were non-negligible numbers of non-agriculturalists paying their tributes in kind, i.e. goods and services they produced. In their case, while the principal occupation was in the secondary or tertiary sector, they tilled a plot or two to sustain their household. Multiple employment was probably ubiquitous throughout the period in question. The other issue concerns the role of women. Although in tabulating Table 3, only the principal occupation was taken into consideration, it is important to realise that in all those multi-employment cases, women’s involvement was crucial. From the 1500 scroll, Wakita Haruko identified 37 as trades carried out by women, or 27 if entertainers, religion-related performers and other service providers are excluded; and 21 of the 27 had trade titles ending with ‘seller’ (Wakita 1985: 26). However, this should not be taken to mean that there was a ‘men make, women sell’ kind of gender division of labour. Even in all the other trades which Wakita classified as men’s, it is difficult to assume that women’s involvement was minimal. With respect to the making of earthenware and sake, Wakita put forward a strong argument that both had long been considered women’s jobs since ancient times (Wakita 1982: 84–87). Even in the earthenware trade, however, it became a custom for men to take their products to the city for sale just before the New Year (Iwasaki et al. 1993: 502). In the medieval period, there was an unmistakable change in the way in which the trade was recognised: increasingly, the right to trade was registered in the name of the male household head. However, as Figures 1 and 3 suggest, women may well have kept working in the domestic economy. What seems certain is that the gender division of labour in manufacturing was not fixed yet. The longer-term tendency was for women’s status to decline with the emergence of specialist manufacturers producing better-quality commodities. In the family economy that started to produce filtered rather than unfiltered sake, or glazed rather than unglazed earthenware, the man assumed a leading role in the household’s productive activity while the woman took up subsidiary, more supportive roles. Accordingly, the nature of women’s work changed. That said, however, it should be remembered that they never withdrew from productive activity as long as the family economy was the principal mode of production. Women’s labour force participation must have been considerably high throughout the medieval period (Tabata 1999: 106; Wakita 1999: 90; Farris 2006: 253–254). From Table 3, in which all occupations are grouped into the primary, secondary and tertiary sectors with some additional sub-groupings, it is clear that the scroll artists paid prodigious attention to entertainers and other service providers. Their ratio to the non-primary total was very high in the 13th century, close to 50%, which declined towards the end of the 15th century (it is likely, though, that they were still over-represented in 1500). Among the rest, makers of various goods and sellers of those itemised commodities were better covered. In the first 1214 scroll, there was just one generic category, ‘merchant’, with the exception of Oharame peddler-maidens who sold firewood and charcoal to Kyoto people. However, the peddlers proliferated in the subsequent centuries as ‘Occupational specialisation in all aspects of urban life was built upon these [women] peddlers’ (Wakita 1999: 88). Thus, in the 15th-century scrolls, almost all commerce-related titles ended with the suffix of ‘seller’. On the other hand, it is a little surprising to see only a few building and transport occupations mentioned in the scrolls throughout the period, and those engaged in mining appearing only in the 1500 list. These trades, especially the first two, were probably under-represented. In order to identify which of the observed trends was real, therefore, I have calculated five ratios in panel B of Table 3. The first two ratios indicate that the tertiary sector increased its share relative to the secondary. While the last line makes it clear that the share of entertainers and service providers in the non-primary sectors declined from 0.5 to 0.3, the third shows that with makers and sellers of processed and assembled goods put together, their ratio increased from 0.4 to 0.5. Most of those makers and sellers are believed to have been in the maker-seller family economy sector. However, the fourth ratio suggests that while in the 13th-century sellers of manufactured goods had hardly been mentioned in the scrolls, their numbers became so numerous in streets by the end of the 15th century that they attracted much attention from the people of Kyoto. To conclude, what was taking place in the 14th and 15th centuries is the increase in the number of those engaged in activities of the maker-seller family economy. The change must have reflected a tendency for more commodities to come to the metropolitan market from the surrounding rural provinces. This was a gradual process, however, and it is difficult at this stage of research to determine whether the rate of increase was actually higher in the sales division than in the production division of this family economy sector. 4. The Proliferation of Commodities With the changes in occupational structure, it is likely that a proliferation of commodities was also underway. Sakurai Eiji (2002) notes that for earlier centuries it is possible to trace the evolution of commodification by looking at records of tributes and taxes paid in kind to estate owners and the state, and that the records did include a number of ‘non-paddy’ products such as silk, salt, paper and hoe. However, one of the earliest educational primers published around 1400 gives a somewhat more comprehensive listing of ‘famous’ products that everyone ought to have known in those days, which numbered 62 including many industrial products. Table 4 compares the sectoral allocation of goods from this 1400 primer with that of 1638, for which date we have one of the earliest Tokugawa-period listings. The 1638 data come from a handbook entitled Feather-Blown Grasses, written by a haiku master for enthusiasts of that literary genre, in which the author ‘luxuriates in the words of the material world as he builds what was for its time a phenomenal list, numbering over 1,800 items, of the “old and new famous things of all the provinces”’ (Berry 2006: 159). Table 4. Sectoral Allocations of ‘Famous Things of All the Provinces’, c.1400 and 1638 Sectors and Ratios . c.1400 . 1638 . A. Number of commodities  Primary sector 18 814   Agriculture 7 285   Forestry 4 173   Hunting and fishing 7 356  Secondary sector 44 940   Mining 2 89   Manufacturing 42 851    Food and drink 6 78    Textiles, clothing and accessories 14 167    Iron castings and edged tools 8 40    Paper products 2 90    Other manufactured goods 14 476  Total 62 1,754 B. Ratios  1. Primary total/ food and drink 3.0 10.4  2. Secondary total/ food and drink 7.3 12.1  3. Mining/ food and drink 0.3 1.1  4. Textiles, clothing and accessories/ food and drink 2.3 2.1  5. Iron castings and edged tools/ food and drink 1.3 0.5  6. Paper products/ food and drink 0.3 1.2  7. Other manufactured goods/ food and drink 2.3 6.1 Sectors and Ratios . c.1400 . 1638 . A. Number of commodities  Primary sector 18 814   Agriculture 7 285   Forestry 4 173   Hunting and fishing 7 356  Secondary sector 44 940   Mining 2 89   Manufacturing 42 851    Food and drink 6 78    Textiles, clothing and accessories 14 167    Iron castings and edged tools 8 40    Paper products 2 90    Other manufactured goods 14 476  Total 62 1,754 B. Ratios  1. Primary total/ food and drink 3.0 10.4  2. Secondary total/ food and drink 7.3 12.1  3. Mining/ food and drink 0.3 1.1  4. Textiles, clothing and accessories/ food and drink 2.3 2.1  5. Iron castings and edged tools/ food and drink 1.3 0.5  6. Paper products/ food and drink 0.3 1.2  7. Other manufactured goods/ food and drink 2.3 6.1 Sources: Anon. (1973: 117–25) for the 1400 primer, and Matsue (1943: ch. 4) for the 1638 handbook (drawing on notes and tabulations in Wakita 1963: 95–112). Note: The 1400 allocations in panel A are revised from the ones in table 1.7 of Saito and Takashima (2017: 81). The branch of ‘food and drink’ includes tobacco. In panel B, all ratios are calculated with respect to the number of commodities in the food and drink sector. Open in new tab Table 4. Sectoral Allocations of ‘Famous Things of All the Provinces’, c.1400 and 1638 Sectors and Ratios . c.1400 . 1638 . A. Number of commodities  Primary sector 18 814   Agriculture 7 285   Forestry 4 173   Hunting and fishing 7 356  Secondary sector 44 940   Mining 2 89   Manufacturing 42 851    Food and drink 6 78    Textiles, clothing and accessories 14 167    Iron castings and edged tools 8 40    Paper products 2 90    Other manufactured goods 14 476  Total 62 1,754 B. Ratios  1. Primary total/ food and drink 3.0 10.4  2. Secondary total/ food and drink 7.3 12.1  3. Mining/ food and drink 0.3 1.1  4. Textiles, clothing and accessories/ food and drink 2.3 2.1  5. Iron castings and edged tools/ food and drink 1.3 0.5  6. Paper products/ food and drink 0.3 1.2  7. Other manufactured goods/ food and drink 2.3 6.1 Sectors and Ratios . c.1400 . 1638 . A. Number of commodities  Primary sector 18 814   Agriculture 7 285   Forestry 4 173   Hunting and fishing 7 356  Secondary sector 44 940   Mining 2 89   Manufacturing 42 851    Food and drink 6 78    Textiles, clothing and accessories 14 167    Iron castings and edged tools 8 40    Paper products 2 90    Other manufactured goods 14 476  Total 62 1,754 B. Ratios  1. Primary total/ food and drink 3.0 10.4  2. Secondary total/ food and drink 7.3 12.1  3. Mining/ food and drink 0.3 1.1  4. Textiles, clothing and accessories/ food and drink 2.3 2.1  5. Iron castings and edged tools/ food and drink 1.3 0.5  6. Paper products/ food and drink 0.3 1.2  7. Other manufactured goods/ food and drink 2.3 6.1 Sources: Anon. (1973: 117–25) for the 1400 primer, and Matsue (1943: ch. 4) for the 1638 handbook (drawing on notes and tabulations in Wakita 1963: 95–112). Note: The 1400 allocations in panel A are revised from the ones in table 1.7 of Saito and Takashima (2017: 81). The branch of ‘food and drink’ includes tobacco. In panel B, all ratios are calculated with respect to the number of commodities in the food and drink sector. Open in new tab Panel A of Table 4 sets out the sectoral distributions for the two time periods. Since the 1638 list was a far more detailed than the 1400 primer, the factor of increase in the total is as large as 28. With this yardstick, what the panel reveals is, on the face of it, surprising. The factors of increase for agriculture, forestry and fishery, and also for mining, all exceed 40, whereas that for manufacturing is 20, much lower than the average. However, this table is based on the counting of ‘famous’ commodities, which would never give us a meaningful measure of the shares of sectors in total output. The observed increase in the primary-sector share has little to do with output change but is more likely to have reflected an increase in the commodification process of the period in question. For example, the 1638 haiku handbook lists a number of rice shipments labelled with the name of the province or district that produced them, suggesting that in the Kyoto market product differentiation was in progress with the producing district used as a brand name. Also, there were products for the food industry; all sorts of ingredients used in the brewing of rice wine, soy sauce and vinegar, as well as the making of buns, candies and other sweets. In the case of sake, not much was needed other than rice initially, but as demand for sake increased more equipment, such as tubs and boxes, was required to expand the business. In medieval circumstances, much of this manufacturing equipment was made of wood, and consequently, the increased urban demand for food items acted as a stimulus not just for the food and drink trade, but for agriculture and forestry as well. Finally, it should be noted that the expansion of mining and quarrying had little to do with this kind of input-output linkage within the domestic economy. While the 1400 list did not include silver, the 1638 handbook mentioned it as a product of Iwami, Tajima, Sado, Dewa and a few other provinces, reflecting the sudden rise of silver mines and an export boom which lasted until about 1630. In the table, manufacturing is divided into five sub-groups: (a) food and drink, including tobacco, (b) textiles, clothing and accessories, (c) iron casting and edge tools, (d) paper and paper products and (e) other manufactured goods. It is clear that among the major four, the textile group’s level of product proliferation was highest in 1400 and remained so until the early 17th century. Markets for silk and ramie fabrics, having developed since early medieval times, and for cotton, which was introduced from abroad in the late 15th century, must have exerted an economy-wide impact on market growth in the subsequent centuries. However, the factor of increase for this group is lower than the manufacturing average. Neither the factor of increase in the grand total nor the manufacturing total gives us a proper reference measure, however. Given the growth of the urban population from the mid-15th century and the obvious fact that what the food and drink industry offers is almost always consumer goods, we may use the food and drink sector as a reference group since its pace of product proliferation is likely to have been closely correlated with the demographic change, on the one hand, and market growth, on the other.8 With this in mind, I have calculated ratios in panel B of Table 4, all defined with respect to the number of commodities in the food and drink sector. The first, second and third lines confirm the observations made above. While the ratio for the textile group slipped a little but remained at a high level of 2, other groups exhibited contrasting trajectories: the ratio for the paper industry (representing the rest of manufacturing) increased from 0.3 to 1.2, but the ratio for iron goods declined from a little over 1 to 0.5. In other words, the pace of product proliferation in textiles was as steady as in food and drink, suggesting that their market growth take-off had started before 1400, while that of paper and other industries came after 1400. On the other hand, the demand growth for products of iron casting and forging never accelerated in medieval and early modern Japan. It was hand technology-based, textile-centred industrial growth that was on course. 5. Signs of Textile-Centred Smithian Growth So far it has been suggested that market growth in the period between the 13th and the early 17th century took the form of increased numbers of consumer goods with an associated increase in the number of sellers of finished products, accompanied by growing transactions between centre and periphery as well as between regions. This interpretation is consistent with another interpretation by monetary historians, i.e. that the medieval money economy grew without any assistance from above because it provided appropriate means of transaction for all parties, both people in the metropolis and makers and sellers in the provinces (Segal 2011: 218–219). The European proto-industrialisation debates tended to focus on two branches of industry: textiles and, to a lesser extent, metal processing (see chapters in Ogilvie and Cerman 1996), because these were the industries which featured the functional and geographical separation of the making of intermediate materials, such as yarn and bar iron, from that of finished goods, such as fabrics and tools, before the age of the factory. In the Japanese case, as we have seen in section 4 above, this kind of development was very weak in the iron and metal processing trade. In medieval Japan, proto-industrial change was driven largely by the textile trade, but little is known about the extent to which textiles increased their share of the country’s total output. However, if we are to demonstrate whether or not this change led to the emergence of markets for intermediate goods in textiles, the issue may be explored by looking at individual branches of the textile industry. This is to ask when and how the market for intermediate goods emerged in each branch of the industry. The fabric trade in Japan before c.1600 had long been two-tiered: silk for the well-to-do, and ramie for the rest of the population. From the 16th century onwards, ramie was replaced by cotton while silk gradually became a more abundant luxury. Silk is a product of silkworms; on the Japanese archipelago, these worms have been raised since the third century CE. However, common people in ancient times did not reel threads off silkworm cocoons; they obtained floss directly from the cocoons and used it for warming themselves. The reeling and weaving of silk eventually started as a product used for tax payment. The ancient state and estate proprietors wanted a number of items rendered as tributes from the commoners in specific areas, and silk was one of the most specialised items produced. Gradually, however, the provinces started marketing their products, and those markets expanded from the 13th century onwards (Nagahara 2004: ch. III). For the silk industry, the crucially important intermediate product is silk thread or raw silk. According to Sakurai, silk reeling as distinct from weaving emerged in the regions north and north-east of Kyoto, such as Kaga and Tango, in the 14th century. Those provinces emerged as supply bases for Kyoto’s weaving centre, which was the medieval prototype of the ‘early modern’ pattern observed in the Tokugawa period on an advanced scale (Sakurai 2015: 306). The 1638 haiku handbook listed five provinces, i.e. Mino, Kaga, Etchū, Tango and Tajima, as producing raw silk. Of the five, Mino dropped out eventually, and the centre of gravity of production moved further east in the 18th and the 19th centuries. But the pattern derived from the 1638 list substantiates the point Sakurai made, although it should be remembered that in between there was a long period in which Kyoto’s weaving district, Nishijin, relied increasingly on raw silk imported from China. These imports started declining by the 1680s, and it took some 70 to 80 years for the domestic reeling sector to increase output and raise its quality enough to meet the demand from Nishijin. This process of import substitution was completed by the final quarter of the 18th century (Tashiro 2004: 109–111). In ancient and medieval Japan, the fabric most widely used by common people for clothing was asa. This fabric can be made either from hemp (Cannabis sativa) or ramie (Boehmeria nivea, called karamushi in Japanese), but it was the cultivation of the latter plant that became widespread by the medieval period (Nagahara 2004: 24–27). The making of the fabric involves several stages. The first is fibre extraction from the plant, which involves separating the raw, spinnable fibre by scraping the outer bark and removing the residual material. This raw ramie fibre is called aoso in Japanese, literally blue ramie, since the material has a lustrous glow. The second stage is spinning, and the final is weaving.9 Like silk, ramie fabrics were also used to pay tributes to the state, aristocrats and other estate owners, with Echigo and other provinces in central Japan widely known for their fine products. But unlike with silk, people clothed themselves with home-made ramie clothes. Until the 13th century, therefore, the market for ramie seems to have remained underdeveloped. The next century saw the commutation of payments in kind on the rise, suggesting that the peasant producers of ramie cloth sold it locally and payed the tributes in cash (Nagahara 2004: ch. II). Moreover, by the 14th century extracted fibre too came to be sold on the market. While the 1400 list of ‘famous’ products does not include ramie fibre, the 1500 illustrated poetry scroll shows a seller of this blue fibre (paired with one selling silk floss in the 59th round). Also, the 1445 freight report of Hyōgo has 26 entries of ramie fibre. Its share of the total freight was not large, but it does indicate that the market for this intermediate product was not confined to central Japan. It was produced in the western provinces as well, since most of those are supposed to have been produced in areas west of Kyoto. Moreover, the same freight report shows that a number of cargoes of indigo dye, another intermediate good for textile production, were in the late medieval interregional market.10 As pointed out by Nagahara (2004: 74–86), the emergence of the market for ramie fibre, the rise of weaving districts in Nara, Echigo and Ōmi as well as markets for secondary products such as summer kimono, nets and ropes in Kyoto, Nara, Tanba and Echizen (these names of specialised districts are also identified from the 1638 handbook), and the evidence that estate-owning Kyoto aristocrats and emerging warring daimyo paid increasing attention to the ramie trade as revenue sources, were all mutually interrelated. Cotton has a different history. It is known to have been cultivated in 15th century Korea, from which the plant was brought to Japan in the latter half of that century. Its cultivation spread rapidly in the western half of the archipelago with southern Kanto at the northern-most edge. It is believed that from the very start, the production of cotton fabrics was market-oriented (Nagahara 2004: 230–263). By the time of the 1638 listing, the extent of regional specialisation within the cotton trade was already substantial. Eight provinces were listed as cultivators, most concentrated in central regions between Owari and Tajima. Weaving provinces numbered 12 altogether, geographically more widespread from Musashi in the Kanto region to provinces in Kyushu. And there were five centres for producing secondary items such as tabi (socks) and other kimono accessories, four of which were concentrated in the cities of Kyoto, Osaka, Nara and Nagasaki. Moreover, it is worth noting that the 1638 handbook singled out Kōriyama in Yamato province out as a district producing ginned cotton (kuriwata), the equivalent of blue fibre in the ramie trade (Wakita 1963: 103). In the early 17th century, the interregional market for this intermediate product was about to take off, and the geographical division of labour within the cotton sector went further in the subsequent period. 6. The Late 16th Century and Beyond The previous sections have made it clear that towards the 16th century, as the proliferation of commodities advanced in the manufacturing sector, so did sales occupations of those manufactured goods. Behind this was the existence of a substantially large maker–seller group of population, from which emerged a group of specialised manufacturers. The level of proliferation was particularly high in textiles, in which this article has identified some signs of Smithian growth. In the silk industry, the separation of a market for silk thread from that for woven cloth began in the 14th century, while in the ramie trade a similar change surfaced by the 16th century, suggesting that Smithian growth was about to take-off in the 16th century. The finding that the market for dyestuff emerged before 1600 is another piece of supporting evidence for this statement. However, what happened in the late 16th century and the subsequent century was a different kind of growth. Driven chiefly by land reclamation projects and the great mining boom followed by town-building programmes, the growth tended to be extensive rather than intensive. Moreover, Japan’s ability to buy Chinese raw silk increased hugely due to the dramatic rise in silver export. According to Souza (1986: 52–53), much of China’s exported raw silk went to Japan; Japan’s annual imports, which included high-quality ‘Chinese white’, rose from 1,000–1,600 picols in c.1600 to a range of 2,500–4,000 picols in the 1620s and 1630s. This meant a more than three-fold increase if calculated from periods before the 1590s to 1640, i.e. from the period of the so-called Red Seal (shuinjō) system to the time when the shogunate began to tighten its grip on overseas trade. The avalanche of Chinese goods on such a scale exerted a depressive impact on the domestic market for Japanese-made raw silk. The magnitude of its impact should not be underestimated, for the depression lasted until the 1680s when imports of Chinese raw silk through Nagasaki declined. Moreover, it took another half-century or so for the domestic silk sector to reorganise itself and to complete the import substitution. It was a period of adjustment in response to changes in overseas trade as well as to changes in the political economy of institution building. However, this cannot be taken to imply that there was a structural break in the evolution of market economy. While the silk industry suffered a setback, market-orientedness in the non-silk textile sector is likely to have kept growing. It is true that the ramie trade was outperformed by the domestic cotton trade. Since both cotton and ramie were used in clothing for common people, and since the level of labour productivity was higher in cotton than in ramie (Nagahara 2004: 315–322),11 much of the latter’s market was taken over by cotton products during the late 16th to the 17th century. As we have seen, the structure of the division of labour within the cotton sector was similar to what had developed in the ramie trade. The emerging market for ginned cotton enabled the geographical dispersion of weaving districts, which went further in the subsequent periods. By the end of the 17th century, cotton weaving districts emerged in areas north of Musashi (the northern-most province mentioned in the 1638 listing), such as Mōka of northern Kanto and Akita in the Tohoku region (Abe 1988: 79–81). Those weaving centres in the north were far away from cotton-growing regions, so that they must have bought ginned cotton from the market. Later, some cotton-producing districts in the provinces around Osaka and Kyoto began to put cotton yarn (kaseito, literally reeled thread) on the market, although it is difficult at this stage to determine whether the separation of reeling from ginning of cotton started in the 18th or the 19th century (Nagahara 2004: 313–314, 328). Whichever the chronology, it is evident that all these early modern developments were on a trajectory whose starting point was in medieval times. Finally, it should be reiterated that in the silk trade too, reeling emerged as a rural industry in both medieval and early modern settings of development. In both cotton and silk, therefore, it worked as a force to further the urban–rural division of labour, thus providing a conducive atmosphere for market growth and increased roundaboutness of the economy, thereby leading to the eventual take-off of Smithian growth. References Abe , Takeshi . 1988 . ‘Kinsei ni okeru orimono seisandaka’ (Weaving Output in the Early Modern Period) . In Bakumatsu-Meiji no nihon keizai (Japanese Economy in the Late Tokugawa-Meiji Transition) , ed. Odaka Konosuke and Yamamoto Yuzo. Nihon Keizai Shinbunsha: 69–83 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Amino , Yoshihiko . 1974 . Mōko shūrai (The Mongol Invasions) . Shogakukan . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 1976 . ‘Chūsei toshi-ron’ (On Medieval Cities). In Iwanami Kōza Nihon Rekishi (Iwanami Lectures on Japanese History), vol. 7: Chūsei (Middle Ages) 3 . Iwanami Shoten: 253–303 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 1977 . ‘Chūseishi no tachiba kara: Hayami hōkoku o megutte (From the Vantage-point of Medieval History: Comments on Hayami’s Presentation)’ . In Atarashii -Edo-jidaishizō o motomete (Towards a New Perspective on Edo-era History) , ed. Socio-economic History Society . Toyo Keizai Shimposha: 255–259 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 1983 . ‘Some Problems Concerning the History of Popular Life in Medieval Japan’ . Acta Asiatica 44 : 77 – 97 . OpenURL Placeholder Text WorldCat ——. 2001 . ‘Commerce and Finance in the Middle Ages: The Beginnings of “Capitalism”’ . Acta Asiatica 81 : 1 – 19 . OpenURL Placeholder Text WorldCat Anon . 1973 . Teikin ōrai (Home Education Primer) , ed. Ishikawa Matsutaro. Heibonsha . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bassino , Jean-Pascal , Stephen Broadberry, Kyoji Fukao, Bishnupriya Gupta, and Masanori Takashima. 2019 . ‘Japan and the Great Divergence, 730–1874’ . Exploration in Economic History 72 ( 1 ): 1 – 22 . Google Scholar Crossref Search ADS WorldCat Berry , Mary Elizabeth . 2006 . Japan in Print: Information and Nation in the Early Modern Period . University of California Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bolt , Jutta , and Jan Luiten van Zanden. 2014 . ‘The Maddison Project: Collaborative Research on Historical National Accounts’ . Economic History Review 67 ( 3 ): 627 – 651 . OpenURL Placeholder Text WorldCat Damian , Michelle M . 2015 . ‘Late Medieval Japan’s Seto Inland Seascape: Shipping, Sailors, and Seafaring’ . Ph.D. dissertation, University of Southern California . Available at http://digitallibrary.usc.edu/cdm/ref/collection/p15799coll3/id/558542 (accessed 9 July 2019 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Farris , W. Wayne . 2006 . Japan’s Medieval Population: Famine, Fertility, and Warfare in a Transformative Age . University of Hawai‘i Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Fukao , Kyoji , Nakamura Naofumi, and Nakabayashi Masaki, eds. 2017 . Iwanami kōza Nihon keizai no rekishi (Iwanami Lectures on the History of Japanese Economy), vol. 1: Chūsei (Middle Ages) . Iwanami Shoten: 289–294 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Fukao , Kyoji , Osamu Saito, Masanori Takashima, and Kiyoshi Kida. 2017 . ‘Seisan, bukka, shotoku no suitei’ (Estimates of Output, Prices and Incomes) . In Iwanami kōza Nihon keizai no rekishi (Iwanami Lectures on the History of Japanese Economy), vol. 1: Chūsei (Middle Ages) , ed. Fukao Kyoji, Nakamura Naofumi and Nakabayashi Masaki. Iwanami Shoten . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hall , John W . 1990 . ‘The Muromachi Bakufu’ . In The Cambridge History of Japan, vol. 3: Medieval Japan , ed. Kozo Yamamura. Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hanley , Susan B . 1997 . Everyday Things in Premodern Japan: The Hidden Legacy of Material Culture . University of California Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hayami , Akira . 2004 . ‘Introduction: The Emergence of “Economic Society”’. In The Economic History of Japan: 1600–1990, vol. 1: Emergence of Economic Society in Japan, 1600–1859 , ed. Akira Hayami, Osamu Saito and Ronald P. Toby. Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hayashiya , Tatsusaburo , ed. 1981 . Hyōgo Kitaseki nyūsen nōchō (Customs Roll of Hyōgo’s North Port) . Chuo-koron Bijutsu Shuppan . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Imatani , Akira . 1981 . ‘Setouchi seikaiken no suii to nyūsen nōchō (Maritime Supremacy in the Seto Inland Sea and the Customs Roll) . In Hyōgo Kitaseki nyūsen nōchō (Customs Roll of Hyōgo’s North Port) , ed. Hayashiya Tatsusaburo. Chuo-koron Bijutsu Shuppan: 272–288 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Iwasaki , Kae . 1987 . Shokunin uta-awase: Chūsei no shokunin gunzō (Tradespeople’s Poetry Competition: Portraits of Tradespeople) . Heibonsha . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Iwasaki , Kae , Yoshihiko Amino, Kiichi Takahashi, and Ko Shiomura, eds. 1993 . Shichijūichi-ban shokunin uta-awase emaki, Shinsen kōka-shū, Kokon ikyoku-shū (Illustrated Scroll of Tradespeople’s Poetry Competition in Seventy-one Rounds, New Collection of Satirical Poems, Collection of Provincial Tunes in All Ages). Shin Nihon koten bungaku taikei (New Compendium of Japanese Classical Literature) , vol. 21 . Iwanami Shoten . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Matsue , Shigenori . 1943 . Kefukigusa (Feather-blown Grasses) , ed. Takenouchi Waka. Iwanami Shoten . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Mori , Toru , ed. 1979 . Ise shin meisho-e uta-awase, Tōhokuin shokunin uta-awase emaki, Tsurugaoka hōjō-e shokunin uta-awase emaki, Sanjūni-ban shokunin uta-awase emaki (Illustrated Pictures of Poetry Competition with Ise’s New Sightseeing Spots, Illustrated Scroll of Tradespeople’s Poetry Competition at Tōhokuin, Illustrated Scroll of Poetry Competition at Tsurugaoka’s Hōjō-e Ceremony, Illustrated Scroll of Tradespeople’s Poetry Competition in Thirty-two Rounds). Shinshū -Nihon emakimono zenshū (New Collection of Japanese Illustrated Scrolls), vol. 28 . Kadokawa Shoten . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Muto , Tadashi . 1981 . ‘Chūsei no Hyōgo-tsu to Setonaikai suiun: Nyūsen nōchō no sensekichi hitei ni kanshite’ (Port of Medieval Hyōgo and Water Transport of the Inland Sea: Identification of Ships’ Home Ports) . In Hyōgo Kitaseki nyūsen nōchō (Customs Roll of Hyōgo’s North Port) , ed. Hayashiya Tatsusaburo. Chuo-koron Bijutsu Shuppan: 232–271 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Nagahara , Keiji . 1990a . Shin momen izen no koto: Choma kara momen e (Days before Cotton: A New Version. From Ramie to Cotton) . Chuo-koronsha . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 1990b . ‘The Medieval Peasant’ . In The Cambridge History of Japan, vol. 3: Medieval Japan , ed. Kozo Yamamura. Cambridge University Press: 301–343 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 2004 . Choma, kinu, momen no shakaishi (Social History of Ramie, Silk and Cotton) . Yoshikawa Kobunkan. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ogilvie , Sheilagh C. , and Markus Cerman, eds. 1996 . European Proto-industrialization . Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Saito , Osamu . 1983 . ‘Population and the Peasant Family Economy in Proto-industrial Japan’ . Journal of Family History 8 ( 1 ): 30 – 54 . Google Scholar Crossref Search ADS WorldCat ——. 1991 . ‘Shohyō: Nagahara Keiji, Shin momen izen no koto’ (Book Review: Keiji Nagahara, Days before Cotton: A New Version) . Rekishigaku kenkyū (Historical Studies) 615 : 44 – 48 . OpenURL Placeholder Text WorldCat ——. 2005 . ‘Pre-modern Economic Growth Revisited: Japan and the West’, Global Economic History Network (GEHN) Working Papers 16/05 . London School of Economics . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 2013 . ‘Proto-industrialization and Labour-intensive Industrialization: Reflections on Smithian Growth and the Role of Skill Intensity’. In Labour-intensive Industrialization in Global History , ed. Gareth Austin and Kaoru Sugihara. Routledge: 85–106 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 2015 . ‘Climate, Famine, and Population in Japanese History: A Long-term Perspective’ . In Environment and Society in the Japanese Islands: From Prehistory to the Present , ed. Bruce L. Baten and Philip C. Brown. Oregon State University Press: 213–229 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 2016 . ‘Japan’. In A History of the Global Economy: 1500 to the Present , ed. Joerg Baten. Cambridge University Press: 167–184 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 2018 . ‘1600-nen no zenkoku jinkō: 17-seiki jinkō-keizaishi saikōchiku no kokoromi’ (Japan’s Population in 1600: An Attempt to Reconstruct the History of Population and Economy in the Seventeenth Century) . Shakai keizaishigaku (Socio-economic History) 84 ( 1 ): 3 – 23 . OpenURL Placeholder Text WorldCat Saito , Osamu , and Masanori Takashima. 2015 . ‘Population, Urbanisation and Farm Output in Early Modern Japan, 1600–1874: A Review of Data and Benchmark Estimates’. Hitotsubashi University RCESR Discussion Paper 15-3. Available at http://hdl.handle.net/10086/27295 (accessed 7 January 2019 ). ——. 2016 . ‘Estimating the Shares of Secondary- and Tertiary-sector Outputs in the Age of Early Modern Growth: The Case of Japan, 1600–1874’ . European Review of Economic History 20 ( 3 ): 368 – 386 . Crossref Search ADS WorldCat ——. 2017 . ‘Jinkō to toshika to shūgyō kōzō’ (Population, Urbanisation and Employment Structure). In Iwanami kōza Nihon keizai no rekishi (Iwanami Lectures on the History of Japanese Economy), vol. 1: Chūsei (Middle Ages) , ed. Fukao Kyoji, Nakamura Naofumi, and Nakabayashi Masaki. Iwanami Shoten: 57–89 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Sakurai , Eiji . 2002 . ‘Chūsei no shōhon shijō’ (Medieval Commodity Markets) . In Ryūtsū keizaishi (Economic History of Money and Trade) , ed. Sakurai Eiji and Nakanishi Toru. Yamakawa Shuppan: 199–234 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 2008 . ‘Currency and Credit in Medieval Japan’ . International Journal of Asian Studies 5 ( 1 ): 53 – 70 . Crossref Search ADS WorldCat ——. 2015 . ‘Chūsei no gijutsu to rōdō’ (Technology and Labour in the Middle Ages) . In Iwanami kōza nihon rekishi (Iwanami Lectures on Japanese History), vol. 9: Chūsei (Middle Ages) 4 . Iwanami Shoten . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Schreurs , Geert . 2019 . ‘Government Institutions and Economic Development in Tokugawa Japan: A Tale of Systems Competition’ . Ph.D. dissertation, Hitotsubashi University . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Segal , Ethan . 2010 . ‘Money and the State: Medieval Precursors of the Early Modern Economy’ . In Economic Thought in Early Modern Japan , ed. B. Gramlich-Oka and G.J. Smits. Brill: 21–46 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC ——. 2011 . Coins, Trade, and the State: Economic Growth in Early Medieval Japan . Harvard University Asia Center . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Shimbo , Hiroshi , and Akira Hasegawa. 2004 . ‘The Dynamics of Market Economy and Production’. In The Economic History of Japan: 1600–1990, vol. 1: Emergence of Economic Society in Japan, 1600–1859 , ed. Akira Hayami, Osamu Saito and Ronald P. Toby. Oxford University Press: 159–191 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Smith , Adam . 1976 . ‘An inquiry into the nature and causes of the wealth of nations’. In The Glasgow Edition of the Works and Correspondence of Adam Smith, vol. I–II, ed. William B. Todd. Oxford: Oxford University Press. Smith , Thomas C . 1988 . ‘Pre-modern Economic Growth: Japan and the West’ . In Native Sources of Japanese Industrialization, 1750–1920 , ed. Thomas C. Smith. University of California Press: 15–49 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Souza , George B . 1986 . The Survival of Empire: Portuguese Trade and Society in China and the South China Sea, 1630–1754 . Cambridge University Press . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Tabata , Yasuko , 1999 . ‘Women’s Work and Status in the Changing Medieval Economy’. In Women and Class in Japanese History , ed. Hitomi Tonomura, Ann Walthall, and Haruko Wakita. University of Michigan Center for Japanese Studies: 99–118 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tashiro , Kazui . 2004 . ‘Foreign Trade in the Tokugawa Period—Particularly with Korea’ . In The Economic History of Japan: 1600–1990, vol. 1: Emergence of Economic Society in Japan, 1600–1859 , ed. Akira Hayami, Osamu Saito, and Ronald P. Toby. Oxford University Press: 105–118 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Takashima , Masanori . 2017 . Keizai seichō no Nihonshi: Kodai kara kinsei no chō-chōki GDP suikei, 730–1874 (Economic Growth in the Japanese Past from Ancient to Early Modern Times: Estimating GDP, 730–1874) . Nagoya Daigaku Shuppankai . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Toyoda , Takeshi , Hiroshi Sugiyama, and V. Dixon Morris. 1977 . ‘The Growth of Commerce and the Trades’. In Japan in the Muromachi Age , ed. John W. Hall and Takeshi Toyoda. University of California Press: 129–144 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Wakita , Haruko . 1982 . ‘Chūsei ni okeru seibetsu yakuwari-buntan to josei-kan’ (Medieval Division of Gender Roles and Women) . In Nihon josei-shi (History of Japanese Women), vol. 2: Chūsei (Middle Ages) , ed. Josei-shi Sōgō Kenkyūkai. Tokyo Daigaku Shuppankai: 65–102 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ——. 1985 . ‘Chūsei josei no yakuwari buntan: Kōtō-no-naishi, hisagime, kanjin-bikuni’ (Division of Roles by Medieval Women: Kōtō-no-Naishi, Peddler-woman, Alms-soliciting Nun) . Rekishigaku kenkyū (Historical Studies) 542 : 16 – 30 . OpenURL Placeholder Text WorldCat ——. 1999 . ‘The Medieval Household and Gender Roles within the Imperial Family, Nobility, Merchants, and Commoners’. In Women and Class in Japanese History , ed. Hitomi Tonomura, Ann Walthall, and Haruko Wakita. University of Michigan Center for Japanese Studies: 81–97 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Wakita , Osamu . 1963 . Kinsei hōken shakai no keizai kōzō (Structure of Early Modern Feudal Society) . Ochanomizu Shobo . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Yamamura , Kozo . 1973 . ‘The Development of Za in Medieval Japan’ . Business History Review 47 ( 4 ): 438 – 465 . Google Scholar Crossref Search ADS WorldCat Young , Allyn A . 1928 . ‘Increasing Returns and Economic Progress’ . Economic Journal 38 ( December ): 527 – 542 . Google Scholar Crossref Search ADS WorldCat Footnotes * I thank two anonymous referees and the Special Issue editor of this journal for their helpful comments and suggestions. 1 Recent estimates of output, population, prices and incomes for the period before 1600 are assembled and published as an appendix to that Iwanami volume (Fukao et al. 2017). 2 Estimates for ancient and medieval benchmark years are taken from Farris (2006: 8–9, 13–26, 95–100); the 1600 estimate is from Saito (2018), and Hiroshi Kito’s estimates for the 1721–1846 period are from Saito and Takashima (2015: table 1). See also Saito and Takashima (2017). 3 Hyōgo had two ports: north and south. The freight report that survived covers the north port (Kitaseki) only for a period from January of Bun’an 2 (1445) to January of the next year. The report is transcribed in Hayashiya (1981). For this source material, see Damian (2015: 113–117). 4 For the locations of incoming ships’ home ports, see the map in Damian (2015: fig. 28 on 120). They were identified by Muto (1981), an article appended to the printed text of the freight report. 5 The estimation formula used is an application of the methodology originally devised to estimate the Tokugawa-era shares of secondary- and tertiary-sector output. It applies the changing levels of urbanisation and population density to the regression equation derived from the Meiji-period prefectural panel-database (Saito and Takashima 2016). 6 Earlier works tended to over-emphasise the level of silver production and trade figures in this period, arguing that Japan’s total supply rivalled the mines of Spanish America, with an aggregate mining output amounting to about 10% of the nation’s gross domestic output (Shimbo and Hasegawa 2004: 167). For more reasonable estimates of Japan’s output and exports, see Schreurs (2019: ch. 4). 7 In interpreting the lists of trade names, I greatly benefited from notes on all the 142 entries written by Yoshihiko Amino, Yoshiaki Fujiwara and others, appended to the transcribed text (Iwasaki et al. 1993: 485–558), as well as Iwasaki (1987)’s explorations of the contents. In the following paragraphs, five of the 142 trades are mentioned: sake brewer in the sixth round of the 1500 poetry competition, yamabito and urabito in the 11th, kawarake maker in the 17th, and sōmen seller in the 37th. 8 One may also note in this connection that ‘By the late-Muromachi years’, i.e. c.1400–1570, ‘all of the major elements of what can be considered traditional Japanese cuisine were present’, and changes that occurred since then were primarily ‘refinements’ (Hanley 1997: 85). 9 For a brief account in English of this less well-known textile trade, see Nagahara (1990b: 324). Much of his account is about ramie, but unfortunately, the term hemp is used to describe the production processes. For example, ‘aoso’ is translated as ‘blue hemp’, which is misleading. 10 This important fact was brought to light by Imatani (1981: 277). Previously, it was assumed that the emergence of the interregional indigo trade did not precede c.1600. 11 Nagahara emphasises the revolutionary nature of cotton in the history of Japanese textiles. While his comparative account concerning the successive production processes from the cultivation of the plant to the weaving of fabrics is revealing and interesting, it seems to me that he tends to underestimate the probability that there was continuity in terms of market differentiation between the ramie and cotton trades. I raised this point in my book review (Saito 1991) of his earlier work (Nagahara 1990a). © The Author(s) 2020. Published by Oxford University Press in conjunction with the University of Tokyo. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - The Medieval Origins of Smithian Growth: The Proliferation of Occupations and Commodities in Japan, 1261–1638
JO - Social Science Japan Journal
DO - 10.1093/ssjj/jyaa003
DA - 0008-01-10
UR - https://www.deepdyve.com/lp/oxford-university-press/the-medieval-origins-of-smithian-growth-the-proliferation-of-jorm9F6BQ0
DP - DeepDyve
ER -