On Monday, China’s decennial census began, sending six million census workers door-to-door in a quest to record and count the country’s population over the course of only ten days. A key issue in this census, according to some observers, will be placing China’s population in terms of place of residence. One thing analysts are waiting to find out, for example, is how many citizens of the PRC are described as living in cities rather than villages, as this census, which comes after a period of massive rural-to-urban migration, is supposed to describe where people physically live and work, not their official place of residence, which is still a village for many urban factory workers. This was not the primary concern of the census takers of the early 1950s. For them, a different kind of placing of the population—into ethnic categories—was a bigger issue, or at least turned out to be an especially complicated one. Here, in two excerpts from his brand-new book Coming to Terms with the Nation: Ethnic Classification in Modern China (UC Press), Tom Mullaney explains why the Communist Party sought to conduct this ethnic classification project, and how the 1953-54 census made this a particularly difficult task.
Fifty-six brothers and sisters together form one family
Fifty-six national languages together form one sentence:
I love my China, I love my China
—Love My China, lyrics by Qiao Yu
From the sacred to the profane, the idea of China as a “unified, multinational country” (tongyi de duo minzu guojia) is a central, load-bearing concept within a wide and heterogeneous array of discourses and practices in the contemporary People’s Republic. China is a plural singularity, this orthodoxy maintains, composed of exactly fifty-six ethnonational groups (minzu): the Han ethnic majority, which constitutes over ninety percent of the population, and a long list of fifty-five minority nationalities who account for the rest. Wherever the question of diversity is raised, this same taxonomic orthodoxy is reproduced, forming a carefully monitored orchestra of remarkable reach and consistency: anthropology museums with the requisite fifty-six displays, “nationalities doll sets” with the requisite fifty-six figurines, book series with the requisite fifty-six “brief histories” of each group, Olympic ceremonies with fifty-six delightfully costumed children, and the list goes on. Fifty-six stars, fifty-six flowers, fifty-six minzu, one China.
China has not always been home to fifty-six officially recognized groups, however. In the late Qing (1644–1911), gazetteerists reported to the imperial center about a wide variety of “barbarians” living in the frontier regions. For one province, Yunnan, such accounts portrayed the region as home to over one hundred distinct peoples, with nearly one hundred more in the neighboring province of Guizhou. Only a few decades later, however, in the China of Chiang Kai-shek, the Nationalist regime vociferously argued that the country was home to only one people, “the Chinese people” (Zhonghua minzu), and that the supposedly distinct groups of the republic were merely subvarieties of a common stock. At the same time, a counterdiscourse emerged among Chinese scholars in the newly formed disciplines of ethnology and linguistics, a discourse in which China was reimagined as home to many dozens of unique ethnic groups—a newly imported concept also translated using the term minzu. Early Chinese Communists began mounting a comparable argument, railing against Chiang Kai-shek’s vision of a mono-minzu China, and on behalf of one in which the country was seen as a composite of politically and economically equal ethnonational constituencies.
Following the revolution of 1949, this ethnotaxonomic volatility persisted. In the first census of the People’s Republic of China (PRC), carried out in late 1953 and early 1954, officials tabulated over four hundred different responses to the question of minzu identity. This deluge came in response to the Communist Party’s promise of ethnonational equality, which entailed a commitment to recognizing the existence of ethnonational diversity to a greater extent than their predecessors had ever been willing to do. Over the course of the subsequent three decades, however, only fifty-five of these were officially recognized, which entailed a remarkable level of categorical compression: from four hundred potential categories of minzu identity to under sixty. The most dramatic case, again, was that of Yunnan Province. Out of the four-hundred-plus names recorded in the 1953–54 census, more than half came from Yunnan alone. Over the following years, however, only twenty-five of these were ultimately recognized by the state.
How do we account for this polyphony of ethnotaxonomic theories? Were there in fact more distinct ethnocultural groups living in the territories of China during the Qing than in the early twentieth century? Had there been a mass exodus? On October 1, 1949, did these communities return, eager to be recognized by the new Communist regime? Clearly, this is not the case. These differences in ethnotaxonomy cannot be accounted for at the level of the categorized. Rather, what changed over the course of this period were the ethnopolitical worldviews of the different Chinese regimes, the modes and methods of categorization they employed, and the political commitments that guided their respective efforts to reconceptualize China in the postimperial era. There was no single “search for a nation in modern Chinese nationalism”—rather, there were searches, in the plural. The Nationalists did not assimilate or expel hundreds of minority groups from the country following the 1911 Revolution. Rather, late republican Nationalists adopted and promoted an ethnotaxonomic worldview wherein the very meaning of the operative term, minzu, was defined in such a way so as to disallow the very possibility of a multi-minzu China. Like a “four-sided triangle,” a multi-minzu China was for Chiang Kai-shek and others a logical impossibility, a contradiction in terms. Continuing into the Communist period, it is clear that the revolution of 1949 did not prompt an influx of minority communities. Rather, the emergence of the “unified multinational” People’s Republic is understandable only when we take into account the radical changes in the very meaning of term minzu and the new regime’s distinct approach to the “national question.”
* * *
The 1953 Experiment with Self-Categorization
Unlike the Soviet census upon which the inaugural PRC registration campaign was loosely based, one which presented respondents with a predetermined set of nationality categories from which to choose, the Chinese census of 1953–54 posed the question of minzu as an unbounded, fill-in-the-blank query wherein a registrant dictated his or her minzu name to the census taker, who then transcribed it into Chinese characters. This policy was known colloquially as “names are to be chosen by the bearers,” or ming cong zhuren. The underlying principle of this policy was a commitment to self-categorization: a political ideal that granted citizens the unfettered right to select their own minzu designations. More specifically, it specified that any individual over the age of eighteen would be free to select his or her own minzu status, which would then be officially recognized by the state. Whatever people called themselves, so too would the government.
The remarkably open nature of the minzu question is most readily appreciated when one compares it against the other questions posed during the registration campaign. When it came to documenting personal names, for example, census takers received instructions on how to sift through the multiple appellations that the average person in China adopted and carried at different stages of life. In certain cases, respondents lacked names altogether, particularly in the recently banned but still persistent practice of betrothal adoption. In such instances, census takers were instructed to assign the person with a name for the sake of documentation.
Age was also a complex issue. According to local custom, an infant in China was considered one year old at birth, and then one year older at the passage of the child’s first Lunar New Year. An infant born shortly before the New Year, therefore, would be quickly be counted as two years old in China, whereas in other contexts the same child would not yet be considered one. To assist census officials in translating between customary modes of chronology and the Gregorian calendar upon which the census was based, participants were outfitted with conversion charts listing dates from 1853 to 1953 in terms of the Western calendar, the lunar calendar, and imperial reign dates.
Place of residence was an especially complicated issue, insofar as the double-counting of absent family members could lead to vast overestimations of the country’s population. Customary practice rendered the issue quite sensitive, however, since it was common for family members to refer to their kin as “temporarily absent” even when they had been away for extended periods of time or had established lives elsewhere. Realizing that it would be an affront not to count these “temporarily absent” individuals, census officials developed a face-saving way of procuring the desired data without bruising the sensibilities of their respondents. The census form was divided into two sections: the right-hand side would include information on permanent residents and those who had been absent for less than six months; the left-hand side would include information about “temporarily absent” family members who had been away for more than six months (and thus, according to census procedures, would ultimately be counted as residents of another location). After recording both, the census office would use the right half and discard the left.
In contrast to the care and attention that the architects of the census paid to the questions of personal names, age, and residency, almost no effort was made to anticipate or neutralize ambiguities associated with the fifth question, that of nationality or minzu. Not only was the question left wide open, but in the event that one’s respondent did not understand the question being posed, census takers were not allowed to assign a minzu designation on the person’s behalf. Rather, they were instructed to approach the issue indirectly, with leading questions such as “Are you Hui or are you Han?” (Huimin haishi Hanmin?) or perhaps “Are you a bannerman or are you Han?” (Zaiqi haishi Hanmin?). By invoking such well-known groups, and contrasting them against the Han majority, these auxiliary questions were designed to induce a corresponding response from the interviewee. Once an answer was provided, however, no further provisions were made to insure commensurability or comparability among the names being procured. Census takers were not provided with any standardized list of nationalities, nor any conversion charts with which they could translate between customary and official modes of identification. No official ethnotaxonomy existed, after all. Thus, whatever a person called him or herself, so too would the state. . . .
As census research neared the halfway mark, however, problems began to surface that called into question this unbounded approach to minzu registration. The problems surfaced, not in the north, where the Communists had developed their ethnopolitical platform, but in the southwestern province of Yunnan where the CCP had virtually no experience with the question of cultural and linguistic diversity. In the county of Yunlong, local census officials reported the discovery of ten additional groups, or at the very least, ten ethnonyms that had never before been recorded by authorities. In Yongsheng County, the situation was identical. Prior to the inaugural census, county officials had recorded the names of some twenty-eight minority groups. During the 1953–54 registration, this number had grown to thirty-eight. In Yuanmou County, local officials reported to the provincial government that many of their respondents did not know how to answer the question of minzu. Even with the auxiliary prompts, many locals simply did not know what the term meant.
In addition to those who did not understand the question being asked, there was an even larger number of people who, from the perspective of state authorities, posed an even bigger problem: those who understood minzu in ways that diverged from the official interpretation. In their report on registration work in Kunming, census takers described two “chaotic” taxonomic phenomena: “single minzu with multiple names” (yi zhong minzu ji zhong mingcheng) and “single names with multiple ways of writing them” (yi zhong mingcheng ji zhong xiefa).
In early 1954, as the tabulation of census data neared completion, it quickly became apparent that local reports such as these, at first seemingly isolated and anecdotal, were signs of a much broader crisis. Back in Beijing, the scope of the problem soon became evident: nationwide, over four hundred distinct entries appeared in the census registers in response to the question of minzu identity. Among these, over two hundred of them were recorded in one province alone: China’s south-westernmost province of Yunnan. . . .
For state authorities, the 1953–54 census data was befuddling. To call it “illegible,” in the James Scott sense, would be a vast understatement. Indeed, to read the list of self-registered minzu in the inaugural census is primarily an exercise in forensic reconstruction, not categorization. The case of one entry, the Xiang, is particularly illustrative. Based on the census results, what could and could not be inferred about this “group,” one whose “population” equaled one? First and most basically, the appearance of the commonplace Chinese character xiang (elephant) tells us almost nothing about the encounter that took place between the registrant and the census worker. We do not and cannot know what the individual said, only what one particular Chinese-speaking census taker considered to be a reasonable approximation of the individual’s declaration of self. Due to vast dialectical variations within Chinese, moreover, the character inscription and the original response are doubly mediated, first from the respondent’s language into the census taker’s Chinese, and from there to state authorities elsewhere in China speaking different dialects. As the Xiang, the nineteen other isolates, and many of the other single-digit entries demonstrate, the fill-in-the-blank census exposed the fact that in contrast to large self-identified groups such as the Lemo and the Jingpo, there were at the same time many people and communities whose understanding of minzu departed radically from that of the state. Because each respondent was permitted to answer the question of minzu improvisationally—not guided by various options that might have provided the person hints as to the specific type of answer sought by the census taker—it left the door wide open for people to answer, in essence, their own versions of the minzu question. Faced with the radically open-ended nature of the census data, then, central and provincial authorities were presented with one “Xiang,” one “Fuduo,” four “Donglan,” and yet absolutely no way to know what a Xiang, Fuduo, or Donglan was. Were they ethnonyms, patronyms, pseudonyms, toponyms, eponyms, or perhaps something else still? The fact is that when confronted with the radical uncertainty of the entry, one has little choice but to acknowledge this uncertainty and recognize: Xiang, Fuduo, Donglan, and many of the other entries in the inaugural census were simply nyms, floating taxa whose classificatory frameworks are lost to us.
Excerpt published with the permission of the University of California Press.