I am endebted for criticism and suggestions to Andrew Barshay, Andrew Gordon, Y¯ Nakanishi, Kazuo Nimura, Rumi Price, Irwin Scheiner, Jeanne M. Smith, and Reginald Zelnik. I am, of course, solely responsible for the final form of the article, and for all statements of fact and opinion in it.

¹ An influential example is Makoto, Ikeda, Nihon kikaikō kumiai seiritsu shiron (Tokyo, 1970).Google Scholar

² Tsutomu, Hyōdō, Nihon ni okeru rōshi kankei no tenkai (Tokyo, 1971)Google Scholar, is a classic and impressively scholarly statement of this view. For a different view, putting emphasis on paternalistic ideology, see Hiroshi, Hazama, Nihon rōmu kanri shi kenkyū (Tokyo, 1964), 3–123Google Scholar; and Crawcour, Sidney, “The Japanese Employment System,” Journal of Japanese Studies, 4:2 (Summer 1978), 225–47CrossRefGoogle Scholar. Dore, Ronald, British Factory/Japanese Factory (Berkeley and Los Angeles, 1973), 375–404Google Scholar, also stresses management's role in creating the employment system, but emphasizes special conditions of late development as well as ideological influences on its decisions. Dore acknowledges the possibility that worker aspirations may also have influenced management decisions but does not follow up the suggestion. None of these studies, therefore, gives workers a significant role in shaping the content of management's reforms. An exception is Gordon, Andrew, The Japanese Factory in Historical Perspective: Labor Relations in Heavy Industry from 1853 to 1955 (Cambridge, Mass., forthcoming)Google Scholar. Gordon sees management reforms as the outcome of a continuous interaction between the concrete demands of workers and management's own strategies. My own view is similar to Gordon's, but differs in stressing the non- Westem ideological background of workers' demands.

³ Gordon, , Japanese Factory, chs. 9, 10.Google Scholar

⁴ From the beginning, advocates of labor unions associated unions with protecting and “expanding” the rights of workers, but said little about the basis of such rights. Katayama Sen made the most considered early effort to justify workers' rights in editorials published in Rōdō sekai. His argument was that the most basic human right was the right to life and hence the right to work; under present conditions of freedom of contract, workers had a “just right” (seitō no kenri) to strike in order to increase their bargaining power, and employers had the right to discharge or lock workers out. This “just right” Katayama called a “natural power” (shizen no ikioi). Rōdō sekai, no. 8 (10 March 1898), pp. 73–74Google Scholar; no. 11 (1 May 1898), p. 104.

⁵ Intellectuals also often found it difficult to keep the distinction between status and rights clearly in mind. The inaugural issue of Katayama Sen's journal, Rōdō sekai, no. 2 (1 December 1897), carried an essay questioning the existing class structure, and asked by what “right” (kenri) the Japanese upper class enjoyed its position. Its answer was, not by superior virtue or understanding but solely on the basis of wealth. The clear implication was that the position of the upper class would have been justified if based on superior virtue and understanding-the traditional and still potent justification of high status.

⁶ Tōyō taimusu, no. 39, (25 June 1921), p. 2.Google Scholar

⁷ Personal communication from a Japanese friend bom about 1912; also, Hiroaki, Matsuzawa, Nihon shakai shugi no shisō (Tokyo, 1973), 126.Google Scholar

⁸ Yūaishinpō, no. 20(15 January 1914), p. 5Google Scholar. For an account of the pioneer labor organization that published this newspaper, see Large, Stephen, The Rise of Labor in Japan: The Yūaikai, 1912¯19 (Tokyo, 1972).Google Scholar

⁹ The names of major status groups in the Tokugawa period were also occupational names: warrior, peasant, artisan, and merchant, in that order. This nominal linkage of status and function disappeared with the elimination of legal status in early Meiji. Its effect on status judgments might be compared to a hypothetical situation in which, in speaking of baseball players, the position played was omitted and the players were identified only by batting average. Thus a man who had been proud to be called a shortstop might find himself identified only as a.220 hitter.

¹⁰ Kagawa Toyohiko, who knew the slum people intimately from social work, believed that most slums had developed from eta (outcaste) communities, and that few if any slums existed in which eta were not resident. Mikio, Sumiya, Nihon chinrōdō shiron (Tokyo, 1958), 105.Google Scholar

¹¹ Rōnin, Kai, ed., Kokkei taika ronshū, 2d ed. (Tokyo, 1888), 11–13, 39–48Google Scholar; Marumaru chinbun (a satirical magazine), East Asiatic Library, University of California, Berkeley.

¹² Tōyō taimusu, no. 69 (5 April 1922), p. 5.Google Scholar

¹³ The announcement of the founding of an early labor organization, Shokkō gundan (Knights of Labor), in 1892 states that “Workers get drunk, sing in the streets, push into crowds and pick quarrels.” iinkai, Nihōn rōdō undō shiryō, ed., Nihon rōdō undō shiryō (Tokyo, 1960), I, 175Google Scholar (hereafter cited as NRUS). NRUS is a basic collection of materials on the labor movement to 1945, planned for eleven volumes.

¹⁴ Yūai shinpō, no. 30 (15 June 1914), p. 5.Google Scholar

¹⁵ Ibid., no. 3 (3 January 1913), p. 4. A short story in this same issue (p. 5) describes two young metal workers leaving the factory and commiserating with one another on their lives, which offered no chance for advancement, only the prospect of endless factory work. One says to the other, “The world is a strange place. When I am dirty with sweat and oil I feel that even my heart is soiled and wonder if we workers are not different animals from human beings.”

¹⁶ Sabaku no harm (Tokyo, 1957), pt. 1, 13–15.Google Scholar

¹⁷ Seichō, Matsumoto, Hansei no ki (Tokyo, 1970), 35, 49.Google Scholar

¹⁸ The head of the Tokyo Industrial School explains that “the vast majority of factory workers today are completely different in both character (hinkō) and deportment (sokō) from school graduates, who therefore cannot bear to be classified along with workers.” Tokyo keizai zasshi (23 January 1908), p. 15.Google Scholar

¹⁹ Kangyō rōdō Sashō rōaikai kikanshi, no. 10 (1 September 1926), p. 1.Google Scholar

²⁰ Rōdō oyobi Kyushu, no. 5 (12 May 1926).Google Scholar

²¹ An article in the Yawata Steel Mill's union paper expresses the connection between factory work, social stigma, and low self-esteem with remarkable clarity. Workers are despised (keibetsu) by society, the paper says, because factory employment is insecure and workers accordingly feel shame at being called “shokkō” (factory workers). “Thus they have no wish to remain shokko and suffer the spiritual pain (seishinteki kutsū) of wounded self-esteem (jifushin).” Tōyō taimusu, no. 31 (5 April 1921), p. 2.Google Scholar

²² The dictionary, Kōjien (Tokyo, 1977)Google Scholar, gives the second meaning of “taigu” as “salary, status and other such treatment (toriatsukai) in the workplace.” Different degrees of taigū were often distinguished as high, middle, and low, warm and cold, and thick and thin. The Mitsubishi Shipyard (Nagasaki) personnel handbook of 1914, Rōdōsha toriatsukaikata ni kansuru chōsa hōkokusho, 1, states: ”Among employees at the Mitsubishi shipyard all those who are not yakuin [clerk/technicians] and do not receive yakuin treatment [taigū] are workers [rōd¯osha].”

²³ Hazama, , Nihon rōmu kanri shi kenkyū, 415–26, 162–67.Google Scholar

²⁴ Workers' perception of this imputation is clearly seen in a worker's letter saying that, at his factory, foremen had in the past entered by the same gate as workers, but since the foremen thought this a reflection on their characters (jinkaku), they had recently persuaded management to let them use a separate gate. Shibaura rōdō (June/July 1923), p. 9Google Scholar. Shōda Heigorō, a Mitsubishi executive, thought that the public widely regarded the treatment (taig¯) accorded workers as a denial of the morality of their characters (Jinkaku). Tōyō keizai shinpō (25 April 1907), p. 21.Google Scholar

²⁵ Rōdō oyobi sangyō, no. 39 (1 November 1914)Google Scholar; Tōkyō keizai zasshi (7 September 1890)Google Scholar; “Shokkō gundan sōritsu shuisho” (September 1892)Google Scholar; all the above published also in NRUS, I, 173–75; III, 174.Google Scholar

²⁶ Yūai shinp¯o, no. 9 (3 July 1913), pp. 1, 3.Google Scholar

²⁷ See Naimushō, [Ministry of Interior], Rōdō sōgi gaikyō (Tokyo, 1917), 26–32, for examples.Google Scholar

²⁸ Rōdō sekai, no. 12 (15 May 1898), p. 8.Google Scholar

²⁹ NRUS, II, 9.Google Scholar

³⁰ Yūai shinpō, no. 29 (1 June 1914), p. 5.Google Scholar

³¹ Tōy¯o taimusu, no. 51 (25 October 1921), p. 2.Google Scholar

³² Yūai shinpō, no. 18 (15 December 1913), p. 5.Google Scholar

³³ Ibid., no. 7 (3 May 1913), p. 3.

³⁴ Osaka Mainichi (18 July 1902)Google Scholar in NRUS, II, 59–60.Google Scholar

³⁵ There were work-rule issues in the strike as well, but the worker manifesto in justifying the strike listed the clothing issue first. NRUS, II, 61.Google Scholar

³⁶ NRUS, II, 64.Google Scholar

³⁷ Leonard, Hugh, Home before Night (Middlesex, England, 1981), 32.Google Scholar

³⁸ NRUS, III, 565.Google Scholar

³⁹ From an early date, workers were sensible of the need to appeal to the public in the course of labor disputes. Dismissed workers at the Omiya factory of the Japan Railroad (Nihon tetsudō) issued a statement appealing to the “brave and upright men of the age” (yo no jinjin resshi). NRUS, I, 510.Google Scholar

⁴⁰ According to the study, the public tended to blame labor disputes on employer despotism (sen'ō)-a word with strong Confucian overtones. Rōdō sōgi gaikyō, 20–21Google Scholar. Osaka Asahi makes much the same point in analyzing strikes in 1907 when it compares employers and workers to rulers and ruled, adding that, since “rulers should guide and instruct,” it is the employer's responsibility to give such succor (kyūsai) to workers as would eliminate strikes. NRUS, II, 87–90.Google Scholar

⁴¹ Most Japanese historians do not regard preindustrial experience and culture as contributing positively to the development of worker consciousness. Ikeda Makoto, for example, speaks of an early “unmediated producer consciousness” (mubaikai na seisansha ishiki) among workers which arose from the confrontation with factory conditions. It was unmediated in the sense that it came from direct experience, presumably without the aid of ideology. This producer consciousness was then developed by intellectuals and union activists among workers into rights consciousness-that is, consciousness of “the right to live, the right to work, the right to collective contract and the right to strike based on justice.” Ikeda, , Nihon kikaikō kumiai seiritsub shiron, 130–80 (quotation, 134–35).Google Scholar

⁴² This is a general impression strongly confirmed by a systematic check of issues in strikes that were reported during the year 1926 in the union organ, Rōdō (pp. 112–15, 144–46, 173–76, 198–201, 230–31).Google Scholar

⁴³ Company financial reports published in the Tokyo keizai zasshi from 1890 first show bonuses (shōyo) for directors, and then increasingly for clerical/technical personnel; the earliest notice I have found in the magazine of a bonus for workers was in 1912, in the report for Boseki, Fuji Gasu (vol. 66, p. 1176)Google Scholar. Notices of workers' bonuses became more and more frequent after 1912 and were fairly common by the 1920s.

⁴⁴ Adoption of a term equivalent to “employee” that would include blue-collar worker might be regarded as symbolic inclusion of workers in the company. An unsystematic sampling of company semiannual reports suggests that the use of employee-equivalent terms came into common use in the 1910s. No such term appears in the reports for Nihon Kokan in 1915, 1917, and 1918, for example, but one-jūgyōin-is used in the 1920 report. For Tokyo Shibaura, there is none in reports between 1904 and 1909, but shiyōjin in appears in 1911. For Tokyo Gasu, none in 1903 and 1910, but shiyōnin in 1916. Most companies also show changes in the specific terms for white- and blue-collar workers. At Tokyo Shibaura, for example, shokkō for blue-collar in 1904 becomes rōekisha during 1909–25, and then kōjin during 1925–27. The company reports referred to here are in the collection of Kaisha eigyō hōkokusho, Hoover Institution and Library, Stanford University.

⁴⁵ Tōyō taimusu, no. 70 (5 May 1922), p. 7Google Scholar; no. 87 (11 May 1922), p. 7. It is evident from the phrasing in the company newspaper's announcement of the change in dress regulations that workers were ashamed of wearing their uniforms and insignia in the town of Yawata. Kurogane (15 July 1922).

⁴⁶ The Yawata paper calls such distinctions “class discrimination” (kaikyū sabetsu), leaving no doubt of the pejorative meaning of the term. Tōyō taimusu, no. 28 (5 March 1921), p. 7Google Scholar. This was a new usage and evidently not one accepted generally: the dictionary Dai Nihon kokugo jiten (Tokyo, 1920)Google Scholar defines sabetsu as the equivalent of kubetsu, meaning simply “distinction” with no pejorative sense. Daijiten (Tokyo, 1935)Google Scholar also defines sabetsu as simply “difference” or “distinction” (chigai, kejime). I do not know the date of the first dictionary inclusion of ”discrimination” as a meaning of sabetsu. Nihon kokugo jiten (Tokyo, 1976)Google Scholar gives the following as the second meaning of the word: “to treat a person as lower than another without just cause.”

⁴⁷ Tōyō taimusu, no. 73 (5 June 1922).Google Scholar

⁴⁸ The newspaper Kurogane, published by the Yawata steel mill for its workers, contains detailed summaries of meetings of the mill's factory committee, the Kondankai. Examples of workers' suggestions for increased distinctions among themselves, advanced at these meetings, include the following: in calculating separation pay, give greater weight to seniority of more than ten years; create retirement ceremonies for those retiring upon reaching the age limit (but not for those retiring otherwise); give full retirement pay to workers with ten year's seniority, and half to workers with less than ten years; change “nonabsentee” awards to “good behavior” awards and create a separate “nonabsentee” award; provide allowances for workers who serve as city or village council members; determine admission to company housing according to seniority; provide a bonus for continuous service (kinzoku teate); establish a seniority bonus (nenkō kahō shōyo) like that in the army. There are also egalitarian suggestions, but these seem always to be aimed at eliminating distinctions between workers and higher status employees, e.g., a request that identification badges of all employees be identical. Kurogane (June 1920–February 1930).Google Scholar

⁴⁹ Gennosuke, Yokoyama, Naichi zekkyōgo no Nihon (Tokyo, 1959), 42–43Google Scholar; idem, “Ninon no kaso shakai” in Yokoyama zenshū, Mikio, Sumiya, ed. (Tokyo, 1972), I, 158–59.Google Scholar

⁵⁰ Hiroshi, Hazama, “Japanese Labor-Management and Uno Riemon,” Journal of Japanese Studies, 5:1 (Winter 1979), 79, n. 8.Google Scholar

⁵¹ A report of the Agriculture and Commerce Ministry (Nōshomushō) in 1900 states: “Hiring and firing and the payment of wages is left to oyabun [work bosses who were often also subcontractors]. In factories with large numbers of workers … the employer has almost nothing to do with these matters, permitting the bosses to collect fees from the workers under them and to advance and demote workers without regard for factory regulations.” “K¯jō oyobi shokkō ni kansuru tsūhei ippan,” in Shokkō oyobi kōfu chōsa, Mikio, Sumiya, ed. (Tokyo, 1970), 56Google Scholar. The power of the oyabun and, later, the foreman over hiring, firing, and wages, though gradually reduced, was still great as late as the 1920s in most large factories. See note 59 below.

⁵² Tōyō taimusu, no. 73 (5 June 1922), p. 6.Google Scholar

⁵³ Yūai shinpō, no. 29 (1 June 1914), p. 4.Google Scholar

⁵⁴ Tōyō taimusu, no. 15 (25 October 1920), p. 5.Google Scholar

⁵⁵ Tsutomu, Hyōdō, Nihon no okeru rōshi kankei no tenkai (Tokyo, 1980, 2d printing), 192–202.Google Scholar

⁵⁶ The Mitsubishi Shipyard at Nagasaki, for example, based payments for disablement or death on the worker's daily wage bracket (less than 35 sen, 35 to 60 sen, and more than 60 sen). Payments for the first bracket ranged from 150 to 500 yen; for the second, 375 to 700 yen; and for the third, 500 to 1,500 yen. Rōdōsha toriatsukaikata ni kansuru chōsa hōkokusho, 72–73.

⁵⁷ For an autobiographical account of such an incident, see Yūai shinpō, no. 26 (15 April 1914), p. 4.Google Scholar

⁵⁸ In a widely publicized incident, the Japan Railway in 1899 dismissed eight workers for protesting the “despotism” and corruption of an overseer; the company forced the eight to move out of company housing within twenty-four hours and confiscated bonuses held for them as savings. NRUS, I, 509–10.Google Scholar

⁵⁹ This was the strike at the steel mill in Muroran in March 1917. Workers rejected the company's plan for a variable increase because, according to a local newspaper, they feared on the basis of past experience that individual raises would be based on the recommendation of foremen and that partiality and corruption would result. NRUS, III, 567–72.Google Scholar

⁶⁰ This strike at the Nagasaki Shipyard resulted in the revision of the formula used for distribution of a wage increase of 30 percent granted by the company some months earlier. This formula had been 10 percent across the board and 20 percent on individual merit as judged by the company; the new formula was 10 percent across the board, 10 percent on attendance record, and 10 percent on merit as judged by the company. NRUS, III, 573.Google Scholar

⁶¹ Rōdō oyobi Kyushu (15 May 1926), p. 2.Google Scholar

⁶² After stating that welfare payments covered by worker regulations were granted “benevolently” (onkeiteki) by the company, the handbook went on to distinguish three categories of workers with differential claims to such payments. Of the third category it was said that “workers in this category … have no right to the benevolence [onkei ni yoku sum kenri naki mono nari] established by these regulations.” Rōdōsha toriatsukaikata ni kansuru chōsa hōkokusho, 1–2, 76–77.Google Scholar

⁶³ For example, Tōyō taimusu, no. 31 (5 April 1921), p. 2Google Scholar; Rōdō oyobi Kyushu (19 April 1924), (15 May 1926), (13 October 1926).Google Scholar

⁶⁴ “Taigū kisei daidōmeikai Ichinoseki shibu kiji,” an account, published in 1899, by one of the strike leaders, Rokujirō, Ishida. NRUS, II, 16–35.Google Scholar

⁶⁵ Tōyō taimusu, no. 40 (5 July 1921), p. 1.Google Scholar

⁶⁶ Tōyō taimusu, no. 40 (5 July 1921), p. 1Google Scholar; no. 70 (5 May 1922), p. 2; Rōdō oyobi Kyushu (19 April 1924), pp. 12–13Google Scholar. School training was often derided as “half-baked learning” (namahanka gakushiki).

⁶⁷ Kōjō seikatsu, 2:1 (07 1917), pp. 16–17Google Scholar; also Tōyō taimusu, no. 28 (5 March 1921), p. 1.Google Scholar

⁶⁸ Tōyō taimusu, no. 69 (25 April 1922), p. 2.Google Scholar

⁶⁹ Ibid., no. 40 (5 July 1921), p. 1.

⁷⁰ “Taigū kisei daidōmeikai Ichinoseki shibu kiji,” NRUS, II, 16.Google Scholar

⁷¹ Kōjō seikatsu, 1:1 (09 1916).Google Scholar

⁷² Tōyō taimusu, no. 69 (25 April 1922), p. 2.Google Scholar

⁷³ Yūai shinpō (15 October 1913), p. 3; (1 November 1913), p. 4.Google Scholar

⁷⁴ I have the impression that until about 1920 “self-consciousness” among workers meant consciousness of personal dignity, respectability, and character, that is, consciousness of jinkaku, and that after 1920 it came increasingly to mean consciousness of working-class identity. Both meanings suggest pride, but the first pride is despite the fact of being a worker, and the second is because of it.

⁷⁵ The earliest use of the term jinkaku found in the collection of dictionaries in the East Asiatic Library at the University of California at Berkeley is in the Dokuwa jiten taizen of 1905 under “Person, Personlichkeit.” The earliest reference in a Japanese-Japanese dictionary is in 1908 in Jirin. The first use of the word in the Christian magazine Rikugō zasshi appears to be in an article by Morita Kumato, published in issue no. 177, in 1897. (I am indebted to Andrew Barshay for these notes.) The dictionary Gensen of 1912 defines jinkakusei (jinkaku-ness) as a quality formed by “self-consciousness (jikaku), reason (risei), and autonomy (jiko kettei).”

⁷⁶ Dai Nihon kokugo jiten (1920)Google Scholar defines hinsei as hitogara (personal character or appearance), hinkaku (grace, dignity), hin (elegance, refinement). Hinkō is defined as behavior (okonai) or hitogara (as above). A contributor to the Yūai shinpō, no. 10 (13 August 1913), p. 10Google Scholar, praises the workers' recreation room, recently opened under the paper's auspices, as a means of improving the use of leisure time by workers and hence raising their hinsei. It is significant that, when workers at Yawata established a recreation center at their own expense with facilities for “moral cultivation” (shūyō), and for “group activities and sports,” they announced that the new facility would raise worker hinsei-not jinkaku. Tōy¯ taimusu, no. 49 (5 October 1921), p. 2.Google Scholar

⁷⁷ Yūai shinpō, no. 33 (1 August 1914), p. 5.Google Scholar

⁷⁸ Tōyō taimusu, no. 57 (25 December 1921), p. 7.Google Scholar

⁷⁹ Yūai shinpō, no. 29 (1 June 1914), p. 5.Google Scholar

⁸⁰ Tōyō taimusu, no. 70 (5 May 1922), p. 2.Google Scholar

⁸¹ Early unionists, who were almost as one in advocating workers' rights, hoped that unions would serve as a basis for cooperation with management; but they foresaw that relations with management would also be marked at times by confrontation and strikes and so stressed the need for worker solidarity. Katayama Sen held that strikes were not necessarily bad since employment relations were based on “free contract,” and hence the employer had the “right” to discharge workers and workers had the “right” in tum to strike. “Domei hiko,” Rōdō sekai (5 October 1898), p. 2Google Scholar. Suzuki Bunji associated unions with the struggle for survival, which was a “law of evolution” that operated at both the individual and group levels. Rōdō oyobi sangyō, no. 56 (1 04 1916), in NRUS, III, 331–32.Google Scholar

⁸² Tōyō taimusu, no. 28 (15 June 1920), p. 1Google Scholar. The Toyo Taimusu was the organ of Doshikai, a Yawata union that survived the strike of 1920 while participating in it, although two other unions, Rōyukai and Yūaikai, which played more active roles in the strike, did not. The latter two were expelled from the mill, along with workers prominently affiliated with them. In taking the Dōshikai paper to represent worker opinion subsequent to its strike, as I believe it did broadly, one must keep the events surrounding the strike in mind, since they must have dampened the expression of opinion in some degree. From a careful reading of workers' letters to the paper, however, I do not have the impression of a subterranean current of opinion very different from the views of the paper. Perhaps such a current could not be detected in any case. Yet the workers' letters are not cautiously phrased; they are vehemently expressed and cover a great range of subjects, from complaints of the most personal kind about working condition, overseers, and promotions, to highly abstruse criticisms of management.

⁸³ Ibid., no. 23 (15 January 1920), p. 1.

⁸⁴ Matsumoto, , Hansei no ki, 49, 76–77, 81–82.Google Scholar

⁸⁵ Status in the factory was based, automatically and rigidly, on education-an achieved status. But education became the basis of ascription in a work situation in which, although status derived from it, education was not an accurate predictor of performance. Although workers were widely aware and resentful of the ascriptive function of education in the factory, so far as I know they had no word to signify a meaning equivalent to “ascription.”