Bibliographies and reviews
 Grading  Nationality  Population registers  Race  Minorities  Suicide

David Howell on identity in 19th-century Japan

Polity, status, and civilization as delimiters of "Japanese"

First posted 1 June 2014
Last updated 1 June 2014

General review Howell 2005  |  Contents | Author | Origin and focus | Title and terminology | Evaluation
Topics and issues Geographies | Identity | States, nations, nation-states | Ethnicity, ethnicization, ethnic negation | Population (household, family) registers | Subjects, nationals, citizens | Status | Japanese | Outcastes, eta, hinin, burakumin | Ezo, Ainu | Ryukyuans, Okinawans | Ogasawarans | Koreans and Taiwanese | Metis, ainoko, haafu

David L. Howell

David L. Howell
Geographies of Identity in Ninetheenth-Century Japan
Berkeley: University of California Press, 2005
x, 261 pages, hardcover


   List of Maps

1. Introduction
2. The Geography of Status
3. Status and the Politics of the Quotidian
4. Violence and the Abolition of Outcaste Status
5. Ainu Identity and the Early Modern State
6. The Geography of Civilization
7. Civilization and Enlightenment
8. Ainu Identity and the Meiji State
   Epilogue Modernity and Ethnicity

   Works Cited


Three simple maps represent the major demographic subjects of Howell's book.

Territory of the outcaste headman Suzuki Jin'emon


Chasing notes, grouped by chapter in the back of the book, is facilitated by page guides in the top margin.

Works Cited

Three archives in Japan are listed in ABC order under "Archives". All cited works are listed under "Published Works" in ABC order by author. Multiple works by the same author are listed in ABC order by title.

The vast majority of the cited works are in Japanese. The publishing particulars of the Japanese works are shown in romanization only, without Japanese script or English translations of titles -- a wise choice, actually, since anyone able to read Japanese would have no difficulty understanding the romanization.


The index is perhaps too simple for a book of this kind. Some topics which Howell touches upon more than once -- such as "race" and "racialization" -- are not indexed -- while a single mention of "ainoko" and two mentions of "métis" are listed.



David Luke Howell (デビッド・ルーク・ハウエル) was born in Fukuoka prefecture in Japan in 1959, and from the age of 10 he grew up in Hilo, Hawai'i. His father, who had been working for the U.S. military, became a sociolinguist and took his family to Tokyo during a sabbatical shortly Howell graduated from high school. During his years in Tokyo,m Howell attended Sophia University, then returned to Hilo to complete his undergraduate studies.

Howell received a BA (1981) in history from the University of Hawai'i at Hilo), then studied for a while at Hokkaido University before beginning his graduate studies at Princeton University, from which he received an MA (1986) and a PhD (1989), also in history. He was teaching at Princeton in 2005 when he published Geographies of Identity, but has been a professor of history at Harvard University since 2010.

Howell styles himself as a historian of early-modern Japan, meaning essentially the Tokugawa period (1603-1868), but he also crosses "the divide of the Meiji Restoriation of 1868" into the Meiji era (1868-1912) -- as he did in the book under review, which he says he wrote "as an early modernist looking across the transom into the modern world" (Howell 2005 pages 2-3).

Howell's first book -- Capitalism from Within: Economy, Society, and the State in a Japanese Fishery (Berkeley: University of California Press, 1995) -- was translated into Japanese with a much more specific title, as follows (my romanizations and [bracketed] translations).

David L. Howell (著)
河西英通、河西富美子 (翻訳)

David L. Howell (authorship)
Kawanishi Hidemichi and Kawanishi Fumiko (translation)
Nishin no kindaishi: Hokkaidō no gyogyō to Nihon shihonshugi
[ The recent-era (early-modern) history of herring: Hokkaido's fishing industry and Japanese capitalism ]
(Kindaishi kenkyū sōsho)
[ Recent-era (Early-modern) history studies library ]
Tokyo: Iwata Shoin, 2007
266 pages, hardcover

The book examines the emergence -- during the late-Tokugawa period, in what would become Hokkaidō during the Meiji period -- of a fishing industry at a proto-industrial stage of development, which enabled the rapid development of capitalism after the Meiji Restoration.

In this book, Howell speaks of "Wajinchi" (和人地) or "Wajin land", referring to the settlements of Wajin (Japanese) -- as opposed to "Ezochi" (蝦夷地) or "Ezo land", the terrorites occupied by Ezo (Ainu), both on Ezo island, later Hokkaidō.


Origin and focus

Geographies of Identity had a long gestation period beginning no later than the early 1980s. One of Howell's first published articles -- "Early Shizoku Colonization of Hokkaidō" (Journal of Aisan History, Volume 17, 1983, pages 40-67) -- came out about the time he was studying at Hokkaido University. Numerous journal articles over the years, including some about Ainu, were finger exercises for chapters in the 2005 book.

Howell acknowledges that "a significant portion" of his book "focuses on relations between the Japanese and the Ainu, and in so doing reveals how even a group apparently external to the Japanese nation was transformed by the same processes that transformed the Japanese themselves." While "the Ainu" figure more than any other "non-Japanese" group(Howell 2005, page 9)

Howell most succinctly describes the purpose of his book like this (Howell 2005, page 9).

Tracing the formation of Japanese identity by examining relations with peoples on the peripheries of the archipelago is a common enough trategy in works like this one, but if fact most of this book focuses on the core of the Japanese polity and on the internal boundaries of identity that placed and replaced subjects in the social and political space of early modern and modern Japan."

He adds, though, that the book is "not a history of the Ainu, nor a history of Japanese policy toward the Ainu, nor, for that matter, a history of the discourse of Japanese ethnic and cultural identity."


<.the book is very much about


Title and terminology



Howell is generally a very readable writer. He clearly avoids the fashionable language adopted by most contemporary scholars who write on "identity" past or present. He is not, to his credit, interested in turning history into a polemic of identity politics.

My main beef with Howell's thesis is that he approaches "identity" from the "commensensical" understanding of "Japanese" -- and adopts subjective, racialistic, stereotypical definitions of "Japanese" at more historically important than objective legal definitions.

Though he claims that "status" is one of his principal concerns, in fact he says practically nothing about the most important status in Japan as a civil society -- according to which "Japanese" are defined by a purely civil -- i.e., non-racialist, non-ethnic, non-cultural -- status called "nationality".

The most fundamental status created in the earliest years of the Meiji period, for the purpose of governing the newly formed Japanese state, was the "standing (status) of being Japanese" (日本人タルノ分限 Nihonjin taru no bungen) (1873), which was later provisionally defined as "National standing (status)" or "Standing (status) as a national" (国民分限 kokumin bungen) (1890), and finally as a manner of possessing "the nationality of Japan" (日本ノ国籍 Nihon no kokuseki) (1899).

Howell, however, says nothing about the early Meiji legal status of being Japanese, and does speaks of "nationality" only once, wedging it between . He alludes to the 1899 Nationality Law as "a law of citizenship" he says Japan didn't have until 1899 (Howell 2005, page 68) -- but it was not a law of citizenship, and even today Japan does not have a law of citizenship.

Possession of Japanese nationality, or its equivalents before its legal definition, was an overarching status that embraced practically all of the people who Howell excludes from "Japanese" as he has chosen to narrowly define the term -- quite apart from its formal legal definition. This is the main reason I am grading his book a "C" rather than a "B". See more about this fatal flaw in Howell's thesis in the "Status" and "Subject, nation, and citizen" parts of the "Topics and issues" section of this review below.



Howell's book is about polity, status, and civilization, each of which contributes to the determination of someone's "identity" -- hence the title of his book, Geographies of Identity. Howell elaborates like this (Howell 2005, page 33, underscoring mine).

  Like other recent contributions to the literature on Japanese identity, this book starts with the now commonsensical premise that ethnic and national identities are historical constructs. (Note 1) Its organizing idea is that three "geographies of identity" situated individuals within social groups and social groups within the political structure of nineteenth-century Japan. The first is the geography of polity, which in the Tokugawa period (1603-1868) delimited a core of shogunal territory and daimyo domains surrounded by clearly subordinate but nominally autonomous peripheries. The second is the geography of status, which placed social groups in Tokugawa Japan in specific relations of power and obligation vis-à-vis the state and, by extension, other groups. The third is the geography of civilization, which distinguished civilized subjects of the shogun from barbarians, both on the peripheries of the state and within the core polity itself.

Howell immediately adds that he uses the phrase "geographies of identity" to highlight his premise that "Japanese ethnic and national identity is the product of a long process of border drawing" (Howell 2005, page 3). He makes no attempt to examine his own agency in the perpetuating of the historical constructs which apparently represent the results of the "long process of border drawing".

Note 1 consists of a long list of books of various qualities and ideologies, beginning with Denoon et al 1996, Multicultural Japan: Palaeolithic to Postmodern; Weiner (editor) 1997, Japan's Minorities: The Illusion of Homogeneity; Oguma 1995 Tan'itsu minzoku shinwa no kigen: "Nihonjin" no jigazō no keifu, translated by Askew as Oguma 2002, A Genealogy of "Japanese" Self-Images; Oguma 1998, "Nihonjin" no kyōkai: Okinawa, Ainu, Taiwan, Chōsen: and Lie 2001, Multiethnic Japan. These works, Howell adds, "build on a large literature critiquing the myth of contemporary Japanese ethnic and cultural homogeneity" such as Befu 2001, Hegemony of Homogeneity: An Anthropological Analysis of "Nihonjinron"; Dale 1986, The Myth of Japanese Uniqueness: and Kosaku 1992 Cultural Nationalism in Contemporary Japan: A Sociological Inquiry.

Howell thus falls in step with other academics who declare (1) their awareness that the "Japanese" is a product of human, especially political culture, not a biological given, (2) their understanding that "Japanese" legally refers to people who belong to Japan's national population defined by possession of Japan's nationality, a purely civil status in both international law and Japanese domestic law, and (3) their choice to use "Japanese" in its emotional, fashionable, conventional, commonplace, "commonsensical" racioethnic sense -- as an "ethnonym" with implications of "race" and/or "ethnicity" in the narrower biological and/or cultural sense of "nation" -- because it suits their purpose to do so. In other words, Howell -- like many others who address issues of "identity" in Japan -- defends what amounts to an academic "validation" of the passively received racialist label -- without which they could not expound their favorite theories of "identity" construction.


Howell also observes that the "geographies" of his analysis "are not solely spaces of the mind." Polity, for example, "was clearly special insofar as physical borders separated the shogun's realm from its peripheries to the north and south." Status, and civilization and barbarism, also "sometimes subscribed to different actual maps of the same space" and were otherwise not strictly independent. "The tangibility suggested by a word like geography suits my purposes well", he concludes (Howell 2005, page 4).

Perhaps he considered other "tangible" metaphors -- like "topography" -- which, to my mathematical mind, is much more suitable term, for it allows the dynamic mapping of the position and movement of any object, in relation to any other object, in terms of any number of variables -- an object, such as a person or a group of persons treated collectively, in terms of, say, polity, status, and civilization.

Three multivariable and interdependent dimensions

Howell's three "geographies of identity" are better understood as the three dimensions which define the general topography on which individuals and/or groups are situated in some sort of dynamic relationship. Think of "polity" and "status" as values on the "x" and "y" axises, and "civilization" as a value on "z" axis, of a conventional 3-dimensional space. All human populations can be mapped within this space, which will also accommodate locations of all manner of human boundaries and clines.

None of the three dimensions are defined by independent variables. All are defined by a number of variables, and all share some variables, such that all to some extent dependent on each other.

I would charcterize the multivariable and interdependent qualities of Howell's "geographies" as follows.


The primary variables of "polity" are "territory" and "government" -- i.e., the land and associated waterways and other geographical assests, all subject to political considerations, including sovereignty, occupation, jurisdiction, and other kinds of administrative control. Territories have borders which may vary in quality according to political conditions. The borders of an state or empire, for example, can expand or shrink, or be open to some people but closed to others. And there may be borders within borders -- neighborhoods, within towns and villages and cities, within provinces, within territories, within states, within regions, say. The "government" of a "territory" will be as complex as the territory. Most entities called "states" or "nations" have multiple polities -- some nested, some clustered, some sprinkled around a larger region, some even "extraterritorial" in the sense that they located within the territory of another government.

The the case of Japan in the early 19th century, you had a Japan that had, over the centuries, evolved into a collection of local domains that were tethered to the Tokugawa shogunate in Edo, and some peripheral territories that were partly under the suzerainty of a nearby domain with some direction from the shogunate.

By 1869 -- after a series of confrontations with other territorial powers in the 1850s, and the implementation of extraterritorial treaties with these powers in the early 1860s -- a revolutionary government had ousted the Tokugawa shogunate and established a monarchy with a central national government in Tokyo, as Edo was renamed. The new government transformed the domains into the prefectures that defined Japan's national territory at the time, and it nationalized the prefectural populations through the use of uniform local village and town household registers that replaced earlier registration practices which had served only the interests of the local polities.

By the turn of the century, Japan considerably expanded its sovereign dominion and negotiated an end to the unequal treaties it had signed with several Euro-American states. Japan extended its sovereignty to, and nationalized and prefecturized, two loosely controlled peripheral territories that had not been formally part of the shogun's dominion -- Ezo and related islands as Hokkaidō and the Ryūkyū islands as Okinawa -- claimed the Ogasawara islands and incorporated them and their few inhabitants into Tokyo prefecture -- swapped territorial interests with Russia, trading its claim to southern Karafuto (Sakhalin) for Russia's claim to the northern Kuriles (Chishima islands), which were incorporated into Hokkaidō -- acquired Taiwan from China as part of the settlement to the Sino-Japanese War of 1894-1895 -- and solidified its political foothold in what, by 1897, had become the Empire of Korea.

By the 1920s, Japan would acquire Karafuto and its Manchurian interests from Russia Japan, annex Korea as Chōsen, and occupy and gain legal rights to administer the former German South Pacific Islands. Various forms of expansion -- territorial and political would continue into the early months of the Pacific War, after which the Empire of Japan began to collapse, ultimately in defeat. By 1945, Japan was reduced to an occupation country consisting of only 46 of the 48 prefectures that defined its Interior territory at its largest in 1943, when Karafuto was integrated into the Interior. The USSR regained sovereignty over Karauto, and while Japan retained sovereignty over Okinawa, the islands were governed by the United States until 1972, when they reverted to Japan's state control and jurisdiction.


The primary variables of "status" are actually "society" and "government". A "society" is demographic territory consisting of a population of individuals and collectivities of individuals who embrace visions of who they and others are -- visions expressed both culturally and linguistically. A "government" is the component of a society which politically administers the territory its members inhabit, and in the process of administration the territory and its inhabitants, may formally or informally determine individual or collective statuses and the "rights" and "duties" associated with such statuses.


The primary variables of "civilization" as the content of human activity are "culture" and "language" -- which, though independent, are often mediated through each other. Civilization as a process of change involves mostly cultural change, though cultural change may be inspired or even catalyzed through contact with or acquision of a different language associated with a different culture.

Both language and culture are subject to territorial, governmental, and social (demographic) conditions, in that both are mediated by the people who inhabit politically defined territories. In fact, politically drawn boundaries are often congruent with linguistic and cultural clines -- hence, in effect, racioethnic national clines. And the extent to which civil national (国民 kokumin) clines are congruent with racioethnic national (民族 minzoku) clines will determine the degree to which a polity's inhabitant population may be said to be "homogeneous" with respect to language and culture if not also racioethnicity and even, though least likely, historical experience.



Howell's most elaborates on the meaning of "identity" -- his most important word -- like this (Howell 2005, page 12, underscoring mine).

Throughout the book I use identity as a relational term, meaning a way of distinguishing among groups of people based on their place in political and social institutions. Therefore, when I speak of identity I am not referring primarily to individuals' self-perceptions, although I suppose that externally marked identities and individual subjects' sense of self frequently overlapped. Rather, my principle concern is with identity from the outside looking in. How could one tell who was Ainu and who was Japanese in the Tokugawa period? What did it mean -- as a question of political power, economic activity, and social relations -- to be a commoner rather than a samurai or an outcaste?

a way of distinguishing among groups of people based on their place in political and social institutions   In other words, people as individuals are assigned a "place" in the political and social order in the form of an ascribed affiliation with a "group" politically and/or socially defined "group".

I suppose that externally marked identities and individual subjects' sense of self frequently overlapped   I would guess so. If babies are stamped "princess" or "prince" or "samurai" or "farmer" or "craftsman" or "merchant" or "eta" or "Ainu" or "ijin" when born -- and children are raised accordingly -- then the odds are that adults will think of themsleves as such -- at least until a prince falls in love with an eta maiden, or a princess wants to marry a samurai who aspires to be a merchant -- or whatever.

Howell does not mean that "individuals and groups have no agency in expressing politically meaningful identities, or that their agency is not important." Rather he stresses that, "at a certain level, surely it matters that institutions mark people as members of this ethnic group or that, this nationality or that, this race or that, with little regard for how the people themselves feel about the act of marking or the particular form the marking takes" (Howell 2005, pages 12-13, underscoring mine).

Here Howell lists ethnic group and nationality and race as "markers" of identity -- topics which I will comment on later (below).

Howell takes a breath then continues like this (Howell 2005, page 13, underscoring mine).

"Or, to express my aim in the idiom of current discourse on nationalism, I will not say much about how individuals imagine ethnic and national communities (in their own minds or through their discourse), but I will instead focus on the ways political, social, and economic institutions revolved to demarcate and hence contain the field within which communities could be imagined. After all, integral to the process of imagining a community of Japanese was the need to reach a rough consensus on where and what Japan was. The state, through its institutions, defined the terms by which such a consensus would be reached.

For Howell, then, Japanese "identity" was -- at least during the period of his interest -- or has been, and possibly still is today -- determined mainly through "consensus" by authorities within Japan's "state" institutions.


States, nations, nation-states

Howell uses "state" and "nation" and "nation-state" throughout his book in ways that he appears to distinguish. However, he does not index -- or anywhere define -- these terms.

Howell declares that "premodern" Tokugawa Japan did not become "modern" Japan until the Meiji period, as a result of adopting "the structures and technologies of the modern nation-state" (Howell 2005, page 3).

Apparently Howell considers Japan to have been a "state" during the Tokugawa period -- that the Tokugawa government -- presumably representing the "polity" of Japan -- qualified Japan as a "state" then, as much as successive governments have supposedly continued to qualify Japan as "state" down to the present -- though Karel van Wolfren offers a number of good reasons to think that the quality of Japan's government today justify his characterization of the country as a "stateless nation" -- in which "nation" refers to the (now) sovereign "people of Japan" -- defined as those who are "nationals" (国民 kokumin) of Japan because they possess its "nationality" (国籍 kokuseki) -- a purely civil, raceless and non-ethnic status regulated by Japan's Nationality Law (国籍法 Kokusekihō).

"Early" and "early modern" Japan as a "state"

Van Wolferen argues that Japan fails to fully qualify as a state because its government is not sufficiently accountable to the Japanese nation -- the people it constitutionally represents. The nation is therefore stateless. Japan's geographic borders are clearly defined and, with the exception of territorial disputes with its neighbors over a few small islands, are internationally recognized. Throughout its history, except during the Allied Occupation following World War II, when its sovereigty was suspended under international law, Japan has been fully responsible for its own diplomatic affairs, and for its own defense. Some people have argued that Japan's sovereignty continues to be compromized by the presence of U.S. military bases in the country 70 years after its defeat 1945, but Japan has chosen to accept U.S. demands for a mutual security alliance that includes the maintenance of U.S. bases in the country and the essentially extraterritorial rights U.S. forces enjoy within the bases.

Finally, a succession of governments have defined Japan's demographic territory -- the people who belong to its changing geographic territory -- its nation -- since antiquity. Nevermind that "nationality" was not a formal legal status until 1899, the "nation" existed. Even during the Allied Occupation from 1945 to 1952, when the state ceased to exist, the nation continued to exist, and Japanese "nationality" continued to be recognized by the Allied Powers. The imperial nation, which in 1945 included the prefectural Interior, Taiwan, and Chōsen, was reduced to the Interior minus Karafuto and Okinwawa prefectures and few islands affiliated with Hokkaidō, Tokyo, and Kagoshima. Consequently, the Japanese nationality of "liberated" Taiwanese (Formosans) and Chosense (Koreans) was partly "alienated" until 1952, when the effects of treaties resulted in their loss of Japanese nationality. And from 1945 to 1972, during which Okinwawa was occupied and administered by the United States, Japan's sovereignty over the islands, and the Japanese nationality of affiliated inhabitants, were suspended.

Early origins of "Japanese" as a nation

Apparently Howell believes that the Tokugawa government did not think of the inhabitants of the domains within its dominion as a "nation". But it seems that the Tokugawa government defined the people it considered "Japanese" just as surely as it defined the dominion it considered "Japan" -- nevermind the precision of its "geographical"

Howell states that "The Tokugawa shogunate was the first regime in Japanese history to define the political boundaries of the state [sic] clearly, although it did not mark them unambiguously" (Howell 2005, page 4). He remarks that "the idea of a central state was nothing new in early modern Japan" for the "imperial house" [Yamato court] defined and defended the realms over which it claimed authority, but adds that, "although earlier regimes had claimed authority over all of 'Japan,' none had ever defined exactly what 'Japan' was, at least not in even remotely precise geographical terms." (Note 3, omitted here) -- and describes the precision of "early modern" Japan's borders like this (Howell 2005, page 5).

Early modern Japan's borders were not the unambiguous lines on a map that separate modern nation-states from one another. (Note 4, omitted here) Even within the inner boundaries of the core polity there existed zones that were autonomous yet subject to the authority of the shogunate. These internal automomies included the daimyo domains (whose physical borders were usually but not always clearly defined), territories under the authority of the Buddhist temples and Shinto shrines, and the more amorphous realm of the outcastes. The internal autonomies of the early modern polity were situationally defined according to the rules of the status system . . . .

Not only were the local domains semi-sovereign, but they maintained their own militias and local monetary systems. The Tokugawa authorities also had a militia, but it was precisely that -- a Tokugawa militia. The shogunate did create a national monetary system, but its use was limited. For there was no state as such -- only a central authority, represented by the most powerful clan that obliged local authorities to contribute taxes and cooperate in domestic and international commerce.

So it appears that the Japanese "nation" evolved before the "state". The leaders of some local clans warred over which clan should rule a united body of domains. They understood that local populations somewhat differed in their dialects, customs, and mannerisms, but they recognized the commanalities of the inhabitants in the various domains, and they believed that all the clans, and all the domains and their inhabitants, would ultimately benefit from the establishement of an overarching political and social order in the larger dominion of Japan.

Under the Tokugawa government, Japan was a confederation of semi-sovereign domains that recognized the authority of the Tokugawa shogunate to compartmentalize the nation, comprised of the aggregate of domain inhabitants, into social for the purpose of maintaining order both within and between the domains. The status system was therefore a "national" system. The government, however, was not quite a "state". A nation-centered political order of the kind that would qualify as a "state" would have to wait until the Meiji restoration -- when the local domains, with their own militias, were replaced by a central "national" government of a "nation" defended by "national" military forces.

Geographical knowledge and mapping methods icreased and improved -- and the political motivation to map became much stronger -- during the Tokugawa period, as Japanese encountered other people in the peripheral marine and terrestrial frontiers beyond "Japan" defined by the "dominion of domains" as I call them. The Tokugawa government marked Japan's boundaries as clearly as it needed to with the mapping technologies available to it.

The dominion of domains -- which corresponds to Howell's "core polity" -- had always been a growing entity that over the centuries expanded into the northern parts of Honshū. Ezochi, too, became an object of mapping, and it was only a matter of time before it, like the former fronteirlands in northern Honshū would also become sufficiently settled by Japanese to dominionize.

Early origins of "Japanese" as a nation

The history of Japan's early population regisistration practices makes it clear that "belonging to Japan" -- subjecthood in Japan, i.e., expecations of loyalty to the Yamato court or to its proxies, such as the Tokugawa shogunate -- was very much determined by criteria that were not at all unlike those which determine nationality status today. Provisions were also made at times to incorporate migrants into what amounted to the "national" population by settling them on land in Yamato localities and enrolling them in local registers.

This link between demographic registration and territory continued to define who was "Japanese" throughout the centuries, even during the Meiji period -- though, until the Meiji era, the registration systems varied with locality. The national regsiter system implemented by the Meiji government within 4 years of its founding continued to define Japanese as people who are affiliated with Japan through their status as registrants in local household registers.

Since "nationality" was of being a registrant of a household register in a locality within Japan's sovereign territory, Taiwanese and Chosense became Japanese subjects and nationals when Taiwan and Chōsen became part of Japan's sovereign dominion in respectively 1895 (the Treaty of Shimonoseki recognized Japan's Japan's right to subjectize Taiwanese from 1897) and 1905 (when Japan annxed the Empire of Korea as Ch%#333;sen) -- and they lost Japan's nationality when Japan formally lost Taiwan and Chō in 1952.

Never was Japan's "nation" in the past -- never has its "nationality" since the formalization of the status in 1899 -- been a matter of race or ethnicity. Primarily "nativity" yes -- born to parents or a parent who is Japanese -- but "nativity" regardless of the race or ethnicity of the Japanese parent(s). Nationality through birth in Japan, or through marriage or adoption (from 1873), or through change of allegiance (in early Japan, called "naturalization" from 1899), have also never matters of race or ethnicity.

Japan's earliest historical chronicles report cases in which children born to a Korean women a Yamato man had taken as a wive was regarded as a "Karako" (韓子) -- a "child of Kara" -- a Korean child. It also reports cases such children, when older, adopting titles and wearing clothing, and crossing borders, in such a way that brought their loyalties into question. And, of course, there are numerous accounts of refugees and others fleeing or migrating, or simply drifting in a wrecked ship, from the peninsula to Yamato, where they were likely to be given land and registered in returning for recognizing the moral authority of the Yamato court -- i.e., changing their allegiance -- the meaning of Sino-Japanese "kika" (帰化) -- the word that came to be used to mean "naturalization" in the nationality laws of Japan, Ching-dynasty China and the Republic of China (歸化 归化 kweihuà), and the Republic of Korea (歸化 귀화 kwihwa). Socialist states have used other expressions to avoid the classical connotations of submission to a ruling sovereign. However, all Asian states whose languages have been significantly influenced by Chinese writing, regardless of their governmental ideology, use their version of 國籍 (国籍) to express "nationality" as a civil status, hence Japanese kokuseki, Chinese guójí, Korean 국적 kukchŏk, and Vietnamese quốc tịch.

Nagasaki magistrates issued edicts that differentiated between people according to whether their mother was of "a seed of Japan" (日本の種子 Nihon no shushi) or their father was of "the bloodline of a barbarian" (蠻人の血脈 banjin no ketsumyaku, meaning Dutch or Portuguese. Other edicts established rules for Chinese men who wished to be entertained in their quarters by a Japanese courtesan, and for the care of children born through such unions. Some edicts also provided punishments -- death in some cases -- for Japanese who left or returned to Japan permission. ditto for aliens who entered Japan without permission.

Clearly the Tokugawa government, through its local magistrates, had a conception of who belonged to Japan and who didn't, and troubled itself to control its not entirely closed borders as best it could. This, too, suggests the extent that Japan already had a nation and was on its way to becoming a state -- or, if you wish, a "nation-state" -- though "nation" is really redundant, since all states are also essentially nations -- including the few that define multiple "nations" and "nationalities" within their general "state nation" and general "state nationality". Imperial Japan, with multiple legal jurisdictions within its sovereign dominion, was no different in this respect -- except that Japan chose not to define minority nationalities within its single imperial nationality. Even today, after the recognition of Ainu as a minority racioethnic nation, Japan does not extend this recognition to individuals or to any group of individuals. Ainu identification is a strictly unclassified personal matter, of no concern of the state, which defines and recognizes only one quality of Japanese nationality.


Ethnicity, ethnicization, ethnic negation

In his Introduction, Howell speaks of "ethnic and national identity" as though they amount to the same thing (Howell 2005, page 3, underscoring mine).

I use the phrase geographies of identity to highlight my premise that Japanese ethnic and national identity is the product of a long process of border drawing. The delineation of the Tokugawa state's political boundaries in the seventeenth century led to the formation of civilizational boundaries between the Japanese and the peoples on Japan's peripheries. Although the physical contours of the early modern polity were clearly bounded, the distinction between the state's subjects and the peoples on the state's peripheries was marked not by an identification with the nation but rather by a conception of civilization borrowed from China and adapted to fit Japanese circumstances.

Is Howell making a distinction between "ethnic" and "national"? Does "identification with the nation" -- rather than "identification with the nation [national polity] and/or an ethnos [ethnic polity] -- imply that "ethnicity" and "nationality" are the same?

Howell uses and indexes "ethnicity" and contends that there were no "ethnic" distinctions in Japan until the Meiji period. He implies that "ethnic" distinctions were made only after defined itself as a "nation" -- i.e., after it became a "modern nation-state" as opposed to a premodern "state". He develops the distinction when explaining why he avoids using the term "ethnicity" in reference to the status of "Ainu" during the Tokugawa period (Howell 2005, page 6, underscoring mine).

The key word missing from the chapters that focus on the Tokugawa period is ethnicity. (Note 6) I avoid the term [in these chapters] because it misrepresents the way difference between the Japanese and the peoples on the immediate peripheries of the Tokugawa state, including the Ainu people of Hokkaido, was understood. Early modern Japanese did not see essential cultural or racial differences between themselves and the Ainu, but rather perceived Ainu identity as an expression of the Ainu's customs (fūzoku), specifically visible features like hairstyles and clothing. Customs were fluid and pliable. So too were the identities and allegiances marked by them.

Note 6 (page 206)   Here I depart from my earlier work on this subject. Compare David L. Howell, "Ainu Ethnicity and the Boundaries of the Early Modern Japanese State," Past and Present 142 (Feb. 1994): 69-93. Parts of this article have been incorporated into chapters 5 and 6.

because it misrepresents the way difference [sic = differences] between the Japanese and the peoples on the immediate peripheries of the Tokugawa state, including the Ainu people of Hokkaido, was [sic = were] understood   Howell makes two comments here, (1) about "ethnicity" as an appropriate historical term, and (2) about the appropriateness of "ethnicity" as a term for describing the differences between, say, "Japanese" and "Ainu" during the centuries leading up to the Meiji period. Regarding (1), he implies that it would be appropriate to use the term "ethnicity" in reference to the understanding of differences between people in and around Japan during the Tokugawa period , if the differences were "ethnic" in nature. Regarding (2), he states that such differences as were not understood as being "ethnic" in nature.

Early modern Japanese did not see essential cultural or racial differences between themselves and the Ainu   But then Howell states that Japanese did see visible differences in customs between themselves and those identified as Ainu -- and are not customs "ethnic" in nature? Apparently both "identities" and "allegiances" were "marked" by customs. Were "allegiances" independent of "identities"? Were either every indepenent of "customs"?

Barth on "ethnic boundaries"

Howell elaborates like this (Howell 2005, pages 6-7, underscoring mine).

If customs had been relevant only as markers of civilization and barbarism -- if they had served only to indicate a dichotomy between barbarian Ainu and civilized Japanese -- ethnicity would indeed be a useful way to conceive of difference in early modern Japan. In their formation, civilizational boundaries were akin to what Fredrik Barth calls ethnic boundaries. (Note 7) As Barth argues, there is no universal litmus test of ethnicity, be it language, race, or culture: ethnic boundaries are marked by essentially arbitray (if hardly random) attributes, including the features of physical appearance covered by the notion of customs. However, customs in early modern Japan marked not only the difference between civilization and barbarism, but distinctions in social status within the core polity as well. Civilizing the Ainu meant assigning them a place within the status order. As a result, attempts to intervene in "ethnic" relations necessarily implicated the status ssytem and its dual role in demarcating differences among social groups while at the same time binding them as the constituent entities of the core polity.

Note 7 (page 206)   Fredrik Barth, "Introduction," in Ethnic Groups and Boundaries: The social Organization of Culture Difference, ed. Fredrik Barth (Boston: Little, Brown and Co., 1969), 9-38.

Howell thus dodges hazarding a definition of "ethnicity" for his purposes. He does not explain why -- a decade after writing about "Ainu ethnicity" in relation to boundaries during the Tokugawa period -- he changed his mind about the appropriateness perceiving differences between people on either side of the boundary betwen "the Japanese" and "the Ainu" as "ethnic". Did he change his definition of "ethnicity"? Or did he change his understanding of how "Japanese" and "Ainu" perceived each other?

Howell arbitarily -- and possibly even randomly, if he's flipping coins on whether to "ethnicize" or "not ethnicize" the difference between "Japanese" and "Ainu" customs -- determines that everything had to do with status up to the Meiji period, after which he claims that status lost it's pivotal role in determining who was "Japanese" (see more under "Status" below).

Turn the page and Howell gets into deeper trouble with his thesis of "ethnicity" and "status" (Howell 2005, page 8, underscoring mine).

See "Status" and "Subjects, nationals, and citizens" (both below) for a closer look at why Howell's

It is important to note that the Meiji state asserted sovereignty over the core polity and the old Tokugawa peripheries but not over other areas, with the sole exception of the Ogasawara Islands. The people of the peripheries -- the Ainu and Ryukyuans -- became imperial subjects and were thereby subsumed politically and institutionally within a broader, homogeneous Japanese identity. This was accomplished through a systematic process of ethnic negation: the state denied the validity of minorities' non-Japanese identities and promoted their acculturation (and eventual assimilation) as the ultimate goal, though it did not immediately require it.

There is nothing important to note with regard to the assertion of sovereignty over what Howell calls the "core polity" -- meaning the domains, soon to become prefectures, of Japan. Howell had already stated that "The Tokugawa shogunate was the first regime in Japanese history to define the political boundaries of the state [sic] clearly, although it did not mark them unambiguously" (Howell 2005, page 4).

In fact, the Tokugawa government marked Japan's boundaries as clearly as it needed to with the mapping technologies available to it. Ezochi, too, was an object of mapping, The "core polity" had always been a growing polity, which over the centuries expanded into the northern parts of Honshū, and it was only a matter of time before Ezochi would also become sufficiently settled by Japanese to dominionize the territory.

By the Meiji period, Japan was sufficiently involved in territories beyond Ezochi that were also objects of Russian expansionism. Such territorial rivalry naturally led to conflicts over interests, and created pressure to sovereignize contested territory, which required declaring and defending borders.

Okinawa was a contested territory of a different sort, and its nationalization by Japan posed different problems for Japan, both diplomatic and domestic. The size and ethnographic quality of the Ryukyuan population -- and the quality of its history as a so-called "kingdom" that both China and Japan treated as a subordinate polity -- required very different approaches to nationalization, which is clear from the manner in which Japan proceded to extend its legal jurisdiction to these newly acquired, eventually prefecturized territories.

The Ogasawaras, among other remote islands that Japan formally nationalized, were not entirely new to Japan. Japan's interest in the islands increased dramatically when they were partly settled by Sandwich islanders and others. Here, too, Japan's designs on the islands did not go uncontested, but the stakes were simply too small for other interested states to oppose.

However, Ryukyuans and Ogasawarans were quickly enrolled into local household (family) registers of the kind prescribed by the 1871 family register law, implemented from 1872. The registration of Ainu proceeded somewhat differently, and Ainu registers wouldn't be integrated into general registers until later. The different early treatment of Ainu residing in local Hokkaidō polities -- and the accomodation of Karafuto Ainu after 1905, when Russia ceded Karafuto to Japan, as well as the treatment of Karafuto Ainu in 1875, when Japan agreed that Karafuto would be part of Russia -- would eventually provide both the political and legal grounds for recognizing Ainu, by the 1990s, over a century later, as a "minority [racioethnic] nation" (少数民族 shōsū minzoku). However, note that even for Japanese who style themselves as being of Ainu descent, the quality of Japan's nationality is the same for all its nationals -- as it was from the beginning of the Meiji period. And in fact, Japanese status as a matter of law has never been based on putative race or ethnicity.


Population (household, family) registers





Howell gives the following overview of the role of "status" in "early modern" Japan (Howell 2005, Introduction, page 4)

Rather than subscribe to a monolithic conception of Japaneseness, the early modern Japanese perceived their identities in terms of social status; they considered themselves samurai, warriors, peasants, outcastes, and so on. Identity in early modern Japan was rooted in the code of status (mibun), and so the early modern roots of modern Japanese identity lay in the workings of the status system (mibunsei).

Howell fails to note that the "status system" applied only to people who were considered inhabitants ot Japan's domains, which did not include the territories that later becaume Hokkaidō and Okinawa. And aliens who arrived at Japanese ports or washed ashore, including those who resided at designated quarters in Nagasaki, or in the extraterritorial foreign settlements established in treaty ports during last decade of the Tokugawa period, were treated as aliens. During the Tokugawa period, only a few aliens were allowed to change change their allegiance and join the "nation" as it were.

Defining impact of "status"

Howell explains the importance of "status" in his scheme of "geographies" like this (Howell 2005, page 7, underscoring mine).

In my discussion of the Ainu, and indeed throughout the book, I emphasize status as the central institution of the early modern political order. . . . few scholars have considered the tremendous -- indeed defining -- impact of the status system on the political, social, and economic structure of Japan in the early modern period . . . no study in any language has ever attempted systematically to elucidate the connections between status and nation building in Japanese history.

Here, as elsewhere, Howell fails to locate the so-called "status system" within the overarching status of belonging to one or another of Japan's domains, which made one Japanese. The "status system" was merely a compartmentalization of people who shared the singular status of belonging to the essentially national dominion defined by the local domains that constituted the territorial and demographic "core" of Howell's Tokugawa "state".

Howell then makes the following statement (Howell 2005, page 7, underscoring mine).

After the collapse of the Tokugawa regime in 1868 status ceased to be the principal criterion of membership in the polity, and eventually it was replaced by the notion of imperial subjecthood. Thus the outcastes became known widely if unofficially as "new commoners" (shinheimin) in 1871, and the samurai lost their status privileges by 1876; both groups formally joined commoners and members of other status groups as more or less homogeneous imperial subjects.

After the collapse of the Tokugawa regime . . .   Contrary to what Howell writes, after the collapse of the Tokugawa regime in 1868, the overarching status of being affiliated with localities within what I call the Tokugawa shogunate's "dominion of domains" -- the collectivity of local domains ruled by semi-sovereign lords -- was replaced by the overarching status of being affiliated with local polities within what I call the the emperor's "dominion of prefectures" -- as the domains began to become in 1869, when their lords returned their lands and registers to the emperor (版籍奉還 hanseki hōkan).

The effect of the return of the lands and registers to the emperor was to restore the emperor's original sovereignty over Japan and its people. Imperial subjecthood did not become a constitutional reality until 1890, and popular sovereignty would not become a constitutional fact until 1947. But the effects of the collective sovereignty of the "dominion of domains" on the status of affiliated individuals did not change. People who had been regarded as "Japanese" before the imperial restoration in 1868 continued to be regarded as so.

the outcastes became known widely if unofficially as "new commoners"   Later, on the same page, Howell states that "the officials maintaining the actual [household] registers occasionally listed former outcastes as 'new commoners' (shinheimin) and Ainu as 'former aborigines' (kyūdojin)." In an endnote to this remark, he cites a buraku history source and states that "officials in only a few districts added notations like 'former eta' and 'new commoner' to entries in the 1872 registry" (page 214, note 48). The "known widely" is clearly an exaggeration. Even so-called "eta" and "hinin" were not that "widely" known in Japan.

RESUME Nocites a source later repeats this with an endnote that states just the opposite (shinheimin) in 1871, and the samurai lost their status privileges by 1876; both groups formally joined commoners and members of other status groups as more or less homogeneous imperial subjects

Howell's characterization of the remark about the creation of the commoner status resulted Howell goes on to describe the changes in personal statuses, which could not have effected the overall and the samurai lost their status privileges by 1876; both groups formally joined commoners and members of other status groups as more or less homogeneous imperial subjects. which remained no more and no less "homogeneous" than it had been during the Tokugawa period. Subjecthood -- before and after the Meiji restoration -- was precisely that. All subjects were equal as subjects before subjecting authorities. A subject's status determined how authorities would treat a given subject in a given matter, but subjecthood was an overarching either-or status. You had to be a subject before you could be assigned a status. The same principle applies today. And today, as in the past, subjectivity to authority and differential treatment based on status particulars determines the parameters of an individuals relationship with the state and nested prefectural and municipal polities.

Status within the overarching status of being Japanese changed pretty much as Howell writes (Howell 2005, page 67).

[The implimentation (in 1872) of the Household Registration Law of 1871] confirmed an earlier [1829] wet of reforms that had divided society into four broad categories: the imperial household (kōzoku), the peerage (kizoku), the gentry (shizoku), and commoners (heimin). The peerage comprised the court nobility, former daimyo, and, after 1884, a handful of persons who had rendered extraordinary service to the state. The genry consisted overwhelmingly of former samurai. . . . Everyone else, regardless of previous status, was legally considered a commoner,

In 1869, the domain lords returned their lands and registers to the emperor (版籍奉還 hanseki hōkan). The former court nobility (公卿 kugyō, 公家 kuge), and and the domain lords (諸侯 shokō, 大名 daimyō), who had been the highest ranking members of the warrior caste or (武士 bushi, 武家 buke), became peers (華族 kazoku). The Imperial Family (皇族 kōzoku) remained at the top of this new caste order. In 1870, other higher ranking samurai were became shizoku (士族), while lower ranking "servant" or "pawn" samurai became sotsuzoku (卒族). Some lower ranking samurai and some commoners also became shizoku. Shizoku and sotsuzoku, who had lost their hereditary stipends when disenfranchised as samurai, were given lump sums of money to pursue new occupations. In 1871, upper and lower castes were allowed to intermarry, shizoku were no longer allowed to wear topknots or swords, and the eta and hinin subcastes were abolished. The domains, with some merging, became prefectures (廃藩置県 haihan chiken), and many former daimyo became governors.


Subjects, nationals, citizens


describes the changes Howell describes the

The farmer, craftsman, and merchant castes, and eta, hinin, and other subcastes, generally became commoners (平民 heimin). Most lower ranking members of the samurai caste became gentry (士族 shizoku). Domain lords and some other high-ranking samurai became peers of the nobility (貴族 kizoku). And members of the imperial family continued to form a caste unto themselves (皇族 kōzoku) statuses were every bit as important as their former staemperor as the monarch Meiji nominal monarch of continued ). a domain status ceased to be the principal criterion of membership in the polity, and eventually it was replaced by the notion of imperial subjecthood  

Howell, having failed to recognize thatpp>

How naturally fails to recognize that the 1971 registration law, implemented from 1872, labeled all people who were expected to register -- essentially all inhabitants of Japan except aliens -- as subjects (臣民 shinmin), nationals (国民 kokumin), and people (人民 jinmin) of Japan -- and these three terms together defined the overarching status of all registrants.





Outcastes, eta, hinin, Burakumin



Ezo, Ainu



Ryukyuans, Okinawans






Koreans and Taiwanese



Metis, ainoko, haafu