Article contents
A Tale of Japanese Technological Diffusion in the Meiji Period
Published online by Cambridge University Press: 11 May 2010
Extract
Fifteen or so years ago and for many decades before that, popularly speaking, it was not uncommon to think of the Japanese as slavish imitators of foreign technology. As research workers in economic history, we recognize that slavish imitation of foreign technology is no easy matter. Foreign technical practice is hardly uniform. Choice among a number of competing technologies is hardly child's play. Young Japanese students of a century ago were enjoined to go overseas, discover what was best and make it Japanese. Such injunctions by a semi-feudal oligarchy and its intellectual supporters while progressive in spirit were naive. One process rarely dominates international industry. What was useful for the world's leader might not be appropriate for the human and nonhuman resource endowment of late nineteenth-century Japan. Initially, Japan's worldwide search led it to adopt a French-style army, an American-style banking system, and a British-style cotton textile industry. In time each of these models were either discarded in favor of other national models or otherwise modified to meet the imperatives of assimilation.
- Type
- Papers Presented at the Thirty-third Annual Meeting of the Economic History Association
- Information
- Copyright
- Copyright © The Economic History Association 1974
References
Members of the following institutions gave me aid and comfort while I collected and interpreted some of the materials used in this paper: All Japan Spinners Association, Hitotsubashi University, Howard & Bulloughs, Kanegafuchi Spinning Company, Keio University, Kyoto University, Osaka University, Platt Brothers, Toyo Spinning Company, Yale University, and the University of Washington. I have also benefitted from the advice of the members of the Economic History Seminar at the University of Michigan.
1 Of course for almost as long as this well-known cliche has been common, there have been those that have sought to correct this impression by emphasizing the many adaptations made by the Japanese. See for example Indian Textile Journal, XVI (1905), 73ff.Google Scholar
2 This injunction was inspired by the fifth article of the famous Meiji Charter Oath issued by the boy-emperor at the time of his assumption of the Japanese throne in 1868—“Knowledge shall be sought throughout the world so as to strengthen the foundations of imperial rule.” See Tsunoda, Ryusaku, de Bary, William Theodore and Keene, Donald, Sources of the Japanese Tradition (New York: Columbia University Press, 1958), p. 644.Google Scholar
3 On the development of the Japanese army, see Hackett, Roger F., Yamagata Aritomo in the Rise of Modem Japan, 1838–1922 (Cambridge, Massachusetts: Harvard University Press, 1971), pp. 82–89.CrossRefGoogle Scholar The early course of Japanese banking is treated in Patrick, Hugh, “Japan, 1868–1914” in Cameron, Rondoet al., Banking in the Early Stages of Industrialization (London: Oxford University Press, 1967).Google Scholar
4 This paper is one of a series of papers I have written on the development of the Japanese cotton spinning industry. The present paper discusses technology. A second paper “Country Girls and the Japanese Cotton Spinning Industry” in Hugh Patrick (ed.) Japanese Industrialization and Its Social Consequences (Berkeley, California: University of California Press, forthcoming) stresses labor force problems and the influence of technology on the quality of the labor force. A third paper, “Productivity Change and Labor Absorption in Japanese Cotton Spinning 1891–1935,” (mimeo) studies econometrically the determinants of productivity change and the influence of such change on the quantity of labor employed.
5 The pre-Osaka Spinning Company period is treated in Kinugawa, T., Hompōmenshi bōseki shi (History of Our Country's Spinning Industry), Vol. I (Osaka: Mengyō Club, 1937).Google Scholar
6 A detailed if somewhat biased treatment of the Osaka Spinning Company's founding is available in Toyo bōseki nanajūnen shi (Seventy Years of the Toyo Spinning Company) (Osaka, 1953).Google Scholar The Toyo Spinning Company began operations in 1913 as the result of the merger of Osaka Spinning Company with another large firm. Takeo Yamabe was the first managing director of the Toyo Spinning Company.
7 Some of the government records on which I have drawn are reproduced in Kinugawa, Hompō, Vol. III. Later confirmation can be obtained from the records available in Rengō bōseki geppō (Spinners Association Monthly Report), Nos. 1–20 (1889–1890).Google Scholar
8 See, e.g., Tōyō bōseki nanajunen shi, pp. 23ff. This theme is taken up in Smith, T. C., Political Change and Industrial Development in Japan: Government Enterprise 1868–1880 (Palo Alto, Calif.: Stanford University Press, 1954), p. 63Google Scholar and repeated in H. Rosovsky, Capital Formation in Japan, Chapter 4. The primary evidence for Smith's contention that government-aided mills played an important role in the cotton industry's development is simply that most mills in 1886 had government-sponsored origins and that four years later spindleage had increased by 200 percent while the number of spinning mills had increased by only 33 percent. This gross comparison is in error. Most of the increase in these four years came from the expansion of existing privately sponsored mills such as the Osaka Spinning Company, from the expansion of new firms which bought out failing government sponsored mills, such as the Mie Spinning Company and from new privately sponsored mills such as the Kanegafuchi Spinning Company and the Amagasaki Spinning Company. See Rengō bōseki geppō, No. 18 (November 1890).
9 The plans for the Osaka Spinning Company's first and second mills can be compared with plans for the Palm and Broadfield Lancashire Spinning Mills. It is interesting that in these early Japanese mills little opportunity was taken to replace power-driven machinery with manually operated alternatives offered by the English machinery manufacturers. For example, manually operated hank cop-reels were not ordered by Yamabe
10 See Moser, Charles K., The Cotton Textile Industry of Far Eastern Countries (Boston: Pepperell Manufacturing Company, 1930), p. 22.Google Scholar
11 The vital and continuing relation between Platt Bros, engineers, fitters and their Japanese clients can be established through examination of correspondence flowing between Platt Bros., Mitsui & Co. (Platt's exclusive Japanese agent), and the constituent firms of the Japanese cotton spinning industry. Note that this section of my paper may be viewed as another attempt to rehabilitate Victorian England's industrial leadership. The most recent work in this vein includes McCloskey, D., “Did Victorian England Fail?” Economic History Review, XXIII (September 1973)Google Scholar, and C. K. Harley, “Edwardian Industry and the Leontief Paradox,” University of British Columbia Discussion Paper, 73–14. Some skepticism has been expressed about the quality of Platt Bros, personnel in Kinugawa, Hompō, Vol. II, pp. 383ff.
12 One such photo is bound in the front of Shoji, O., Yamabe Takeo minden (Life of Takeo Yamabe) (Osaka: Boshoku Zasshi sha: 1921).Google Scholar
13 This period is described in ibid., pp. 9–13.
14 The importance of the student years abroad is a recurrent theme in the literature on the Meiji political and economic leadership. Most of this writing is based entirely on the personal reminiscences of these leaders decades after their experiences abroad. For another sentimental account see Sanji Muto's account of his own experience as a young man in the United States in Mutō Sanji zenshu (Tokyo: Shinju-sha, 1967).Google Scholar See particularly Mutō's reminiscences of Toyoji Wada in San Jose. Both Wada and Mutō were premier figures in the cotton spinning industry in the early twentieth century.
15 In the introduction to Sanada's, B.Boseki kiyō (Spinning Notes) (Osaka: Maruzen, 1895)Google Scholar, Yamabe writes that the publication of the handbook signifies the technological coming of age of the Japanese spinning industry.
16 Tōyō bōseki nanajūnen shi, pp. 54–55.
17 For a discussion of the relative virtues of the ring and the mule, see Taggart, W. Scott, Cotton Spinning, Vol. III (London: Macmillan and Co., 1911), pp. 331–333.Google Scholar See also Sandberg, Lars, “American Rings and English Mules,” Quarterly Journal of Economics, LXXXVIII (February 1969).Google Scholar
18 Lest there be any suspicion on this point, the company history assures us these mills were not covered by fire insurance. Tōyō bōseki nanajūnen shi, p. 38.
19 Letter from Dohrenfield to Platts, 7/25/1889, bound in Hirano Spinning Mill plans.
20 Letter, Neill to Platt Brothers, 12/3/1892, part of correspondence relating to Sugano Spinning Mill. There are many instances where Japanese machinery manufacturers showed themselves technologically capable of producing looms or spinning machinery but were spurned on economic and quality grounds by the domestic cotton industry. Indeed, spinning machinery had been produced in Japan as early as the 1870's. It was not until the late 1920's, however, that a switch was made from Platts to Toyoda. See Gary R. Saxonhouse “Country Girls and the Japanese Cotton Industry.”
21 There is no evidence that this preference for manually operated presses was anything more than an interlude. The preference certainly did not survive the First World War.
22 The sources of these data are discussed in Saxonhouse, “Country Girls.” As to the data quality, Yasuba in his study of pre-war production statistics concludes that the data available for the Japanese spinning industry appear to be entirely reliable. See Y. Yasuba, “Senzen no Nihon ni okeru kōgyō tōkei no shimpyōsei ni tsuite,” (“On the Reliability of Pre-War Industrial Statistics”) Osaka daigaku keizai gaku (November 1965). In an effort to independently confirm the reliability of the published data I have used, these numbers have been compared with the figures recorded in the unpublished, private books of a number of spinning companies. This was done most comprehensively with records of the Kanegafuchi Spinning Company, for many years Japan's largest cotton spinning company. Private and public figures never differed by more than one percent.Google Scholar
23 For many types of economic analysis, the assumption of a Cobb-Douglas production function is highly restrictive. In Gary R. Saxonhouse, “Productivity Change and Labor Absorption in Japanese Cotton Spinning 1891–1935” (mimeo), a more general relationship is imposed on the Japanese cotton spinning data.
24 In a study on cotton spinning it may seem unusual that cotton is excluded from the production function. My preliminary empirical work indicated that raw cotton was so closely correlated with capital as to make it statistically indistinguishable from this variable. For a substantively similar treatment of this point see P. David, “Learning by Doing and Tariff Protection: A Reconsideration of the Case of the Ante-Bellum United States Cotton Textile Industry, Journal of Economic History, XXX (September 1970).
25 Given what has preceded, good statistical practice might require that the composite alternative to Ho:
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20151017025750548-0649:S0022050700079687_eqnU1.gif?pub-status=live)
be treated as the null hypothesis. In absence of such an approach the design of the test may make the risk of mistakenly not rejecting the uniformity of practice hypothesis unacceptably high. Unfortunately given the state of mathematical statistics, taking a composite alternative and making it the null hypothesis is most difficult in the current context. Instead, uniformity of practice will be left as the null hypothesis and the critical region will be expanded to guard against the dangers outlined above.
26 This test statistic is discussed in Swamy, P. A. V. B., “Efficient Inference in a Random Coefficient Regression Model,” Econometrics, XXXVIII (March 1970)Google Scholar, and Rao, C. R., Linear Statistical Inference and Its Applications (New York: John Wiley & Sons, 1965).Google Scholar
27 I know of no instance where a test such as I have employed has resulted in the rejection of the null hypothesis. Admittedly such tests for aggregation bias are few and far between. On the other hand the work of Balestra and Nerlove in “Pooling Cross-section and Time Series Data in the Estimation of a Dynamic Model: The Demand for Natural Gas,” Econotnetrica, XXXIV (July 1966)Google Scholar does suggest that there are occasions when the allowance for firm effects will not contribute significantly to an increase in explanatory power.
28 See Rosenberg, Nathan, “Factors Affecting the Diffusion of Technology,” Explorations in Economic History, X (Fall 1972)Google Scholar, and Harley, C. K., “On the Persistence of Old Technologies: The Case of North American Wooden Ships,” Journal of Economic History, XXXIII (June 1973).Google Scholar
29 An excellent example of the fruits of this now almost abandoned approach can be found in W. N. Parker, “The Social Process of Diffusion,” (mimeo). Much of this section of my paper draws on the spirit implicit in Parker's analysis.
30 See, for example, North, Douglass C. and Thomas, Robert P., “An Economic Theory of the Growth of the Western World,” Economic History Review, XXIII (April 1970)Google Scholar, North, Douglass C., “Sources of Productivity Change in Ocean Shipping 1600–1850,” Journal of Political Economy, LXXVI (September-October 1968)Google Scholar and Reed, Clyde, “Transactions Costs and Differential Growth in Seventeenth Century Western Europe,” Journal of Economic History, XXXIII (March 1973).Google Scholar
31 A detailed account of many of the activities of Bōren is contained in Shoji, O., Bōseki sogyō tanshuku shi (History of Operational Curtailments in the Japanese Cotton Industry), (Osaka: Nihon Mengyō Club, 1930).Google Scholar
32 Given relatively free entry, secularly Bōren was not in the position to generate greater than competitive profits for the membership of the industry.
33 This is discussed in T. Kinugawa, Hompō, Vol. III, pp. 193–210.
34 See for example the charts and discussion in Rengō bōseki geppō, II (May 1890).Google Scholar
35 With respect to interest in foreign activities in particular and technology in general, the Bōren geppō resembles the Indian Textile Journal which began publication in 1890. The Indian Textile Journal, however, could never approach the detailed disclosure of individual firm productivity and costing which appeared in its Japanese counterpart. The standards set by the Japanese and Indian journals were not universally emulated in other “follower” textile industries. The Brazilian textile industry for example, never developed such a journal. See Stein, Stanley, The Brazilian Cotton Manufacture (Cambridge, Mass.: Harvard University Press, 1957).CrossRefGoogle Scholar
36 Some of the Meiji era reports sponsored by Boren include Fukuhara, H., Beikoku nambu bōseki jijo (Conditions of the Spinning Industry in the Southern United States), (Osaka: Dāi Nihon Bōseki Rengōkāi, 1903)Google Scholar; Mutō, S., Bōseki daigōdō ton (A Discourse on Spinning Amalgamation), (Osaka: Dai Nihon Bōseki Rengōkai, 1901)Google Scholar; Shoji, O., Bōseki kōjo sogo kasai hoken, (Osaka: Dai Nihon Bōseki Rengōkai, 1911)Google Scholar; and Atsushi, Inoue, Doitsu ni ōkeru kogyō kyōiku (Technical Education in Germany), (Osaka: Dai Nihon Bōseki Rengōkai, 1903).Google Scholar
37 See Fukuhara, Beikdku.
38 The books and pamphlets include Jackson, P. T. and Shearman, E. B., Report on the Production and Manufacture of Cotton (Boston: Friends of Industry, 1832)Google Scholar; Martineau, Harriet, Society in America (New York: Saunders & Otling, 1937)Google Scholar; Cowley, Charles, Illustrated History of Lowell (Boston: J. R. Osgood, 1868)Google Scholar; Larcohn, Lucy, An Idyll of Work (Boston: G. P. Putnam & Sons, 1875)Google Scholar; City of Lowell, Proceedings in the City of Lowell at the Semi-Centennial Celebration of the Incorporation of the Town of Lowell (Lowell, 1876)Google Scholar; Robinson, Harriet H., Loom and Spindle, or Life Among the Early Mill Girls (New York: G. P. Putnam & Sons, 1898)Google Scholar. See also, Rengō bōseki geppō, No. 20 (December 1890)Google Scholar, and Dai Nihon bōseki dōgyō rengōkai hokoku, No. 110 (October 1898).Google Scholar
39 Tōyō bōseki nanajūnen shi, pp. 54–55.
40 Kinugawa Hompō, VI, p. 60. Other instances of firm to firm technological assistance include Osaka Spinning Co. helping Ichikawa Spinning Co. (Kinugawa, Hompō, III, p. 76); Tamashima Spinning Co. helping Snodoshima Spinning Co. (Kinugawa, V, p. 232); Kuwana Spinning Co. receiving help from Owari Spinning Co. (Kinugawa, VII, p. 68); Ichinomiya Spinning Co. received help from Nippon Spinning Co., Meiji Spinning Co. and Mie Spinning Co. (Kinugawa, I, p. 121); Kuwahara Spinning Co. from Settsu Spinning Co. (Kinugawa, II, p. 235); Okayama Spinning Co. from Himiji Spinning Co. and Shibutani Spinning Co. (Kinugawa, II, p. 292); Ōsaka Spinning Co. from Kuwahara Spinning Co. (Kinugawa, II, p. 390); Kishiwada Spinning Co. from Senshu Spinning Co. (Kinugawa, VI, p. 62); Shimomura Spinning Co. from Kuwahara Spinning Co. and Tamashima Spinning Co. (Kinugawa, III, p. 108); Mie Spinning Co. from Taihei Spinning Co. (Kinugawa, III, p. 469); Kurashiki Spinning Co. from Shimomura Spinning Co. (Kinugawa, V, p. 36); Kofu Spinning Co. from Ichikawa Spinning Co. (Kinugawa, V, p. 167); and Settsu Spinning Co. helping Takaoka Spinning Co. (Kinugawa, V, p. 20).
41 Ainlie to Platt Bros., 3/15/96, letter found in plans for an addition to Osaka Spinning Company.
42 Clark, W. A. G., Cotton Goods in Japan (Washington: United States Department of Commerce, 1914), p. 213.Google Scholar
43 For example, in a memorandum dated 10/30/1900 to Tokyo Gassed Yarn Spinning Company, Platt Brothers discussed in detail the experience of the Nippon Spinning Company.
44 Letter from Hempstock of Platt Brothers to Dormon of Mitsui and Co., 11/26/1925.
45 Acting collectively through Bōren, the Japanese spinning industry was able to break the P & O Shipping Line's monopoly on the Bombay-Kobe routes. This important episode is described in Kakakō, Sampei, Nihon mengyō hattatsu shi (Tōkyō: Keio Shōbu, 1941), pp. 311–316.Google Scholar
46 The documentation for the assertation that the Japanese cotton spinning persisted in using what was by international standards a young female and very in experienced labor force throughout the fifty years prior to 1940 is presented in Gary R. Saxonhouse, “Country Girls and the Japanese Cotton Spinning Industry.” This paper also developed in greater detail the hypothesis that technological uniformity inhibited human capital formation in this industry.
47 The Japanese experience, of course, is best appreciated in comparative perspective. The Japanese spinners deserve to be compared with such foreign spinners as say Moses Brown of Almy and Brown. See Ware, Caroline F., Early New. England Cotton Manufacture (New York: Russell and Russell, 1966).Google Scholar That co-operation was the mode in the Japanese industry should not obscure the many bitter controversies which were part of the industry's experience. See some of the accounts given in T. Kinugawa, Hompō, IV, pp. 454–476.
48 It is possible that equity in the Japanese cotton spinning industry was more concentrated than the number of firms would indicate. Thus some of the technological co-operation cited above was not a matter of externalities at all. Nonetheless it should be understood that unlike other nascent Japanese industries, cotton spinning was not dominated by a handful of banks. Interlocking control of a number of separate firms was no more characteristic of the Japanese industries than it was of the Indian, New England, or Lancashire industries.
- 59
- Cited by