Semu was a collective designation used under the Mongol-led Yuan dynasty (1271–1368) for people of diverse origins who were distinguished administratively from Mongols and from the populations generally classified as Hanren and Nanren. Many came from Central and Western Asia, although the category also included people from other regions. Semu was not a single ethnic group, religious community, or nationality. It was a context-dependent classification associated with the recruitment of officials, legal distinctions, and the organization of the Yuan empire’s diverse subjects.
Terminology and meaning
The Chinese term semu (色目) had a broader meaning of “categories,” “kinds,” or “designations.” In Yuan contexts, semuren (色目人) identified people belonging to a range of groups treated separately from Mongols and the main categories of Chinese subjects.
The expression is sometimes translated as “colored-eye people,” but this misleading interpretation treats the characters as a description of physical appearance. Eye color was not the defining criterion of the category. Translations such as “people of various categories” better indicate its classificatory character, although no brief English rendering fully captures its historical usage.
The meaning of the term depended on the document and institution involved. Earlier Chinese uses of semu as an ordinary word for categories should not automatically be understood as references to the later Yuan population designation.
Origins in Mongol imperial expansion
The category developed against the background of the Mongol Empire and its incorporation of populations across Eurasia during the thirteenth century. Conquest, military service, commerce, and the movement of skilled workers brought Central and Western Asians into northern China and subsequently into other territories ruled by the Yuan.
These movements took several forms. Some people entered Mongol service as allies, administrators, soldiers, or merchants. Others were captives or belonged to communities transferred under imperial direction. Their descendants could retain an inherited group affiliation even after settlement in China.
Mongol rulers drew on personnel from different regions to govern a large empire. Individuals with experience in finance, languages, commerce, and administration could serve as intermediaries between the ruling court and its subjects. Appointment also depended on patronage, family service, and political loyalty rather than on a uniform policy of employing foreigners.
The formation of the Semu category therefore reflected both migration and the administrative practices of conquest. It should not be understood simply as a label for all foreigners living in China.
Peoples and religious diversity
Groups commonly included among the Semu were Uyghurs, people of Tangut origin, and various Central and Western Asian populations, including people from Persian-speaking regions. Some people of Caucasian and other origins also appeared within the category. Its boundaries were not identical in every administrative setting.
Semu communities were religiously diverse. They included followers of Islam, Buddhism, and Christianity, among other traditions. Religious affiliation and administrative classification did not coincide: not all Semu were Muslim, and “Semu” was not a synonym for any one religious community.
Likewise, the category did not identify a common language. People classified as Semu could use Turkic, Iranian, Mongolian, Chinese, or other languages. Their linguistic practices could change across generations, particularly among families established in Chinese-speaking regions.
Position within Yuan society
The four-category model
Yuan society is often described through a four-part hierarchy:
- Mongols, the ruling conquest population.
- Semu, a heterogeneous collection of other groups.
- Hanren, generally the populations of northern China previously ruled by the Jin dynasty, including groups not equivalent to the modern Han ethnic category.
- Nanren, generally the former subjects of the Southern Song dynasty.
In this conventional account, Semu occupied a favored position below Mongols but above Hanren and Nanren. The model captures important inequalities in access to office and aspects of legal treatment. However, it is an analytical simplification, not a complete description of every Yuan institution.
The categories combined political history, inherited affiliation, and geographical origin. They should not be mapped directly onto modern ethnic identities or citizenship. Nor did the Yuan establish a single comprehensive code that organized every aspect of society into four uniformly applied ranks.
Officeholding and examinations
Semu personnel held positions in the Yuan bureaucracy, including fiscal administration, provincial government, and court service. Some families accumulated influence through service over several generations. Other Semu people were soldiers, artisans, merchants, or ordinary households without substantial political power.
The imperial examination system provides a clear example of institutional differentiation. After the court ordered its restoration in 1313, examinations were held in 1315 with separate provisions for Mongols and Semu on the one hand and Hanren and Nanren on the other. Examination requirements and selection arrangements favored the former groups relative to the much larger Chinese populations.
Examinations were nevertheless only one route into Yuan office. Hereditary service, recommendation, clerical promotion, and direct appointment also mattered. Examination privileges therefore explain only part of Semu participation in government.
Law and social position
Yuan law sometimes distinguished people by group affiliation in judicial procedure, punishment, and eligibility for particular offices. These distinctions were not identical across all periods or types of case.
Administrative advantage also did not mean that every Semu individual was wealthier or more powerful than every Hanren or Nanren individual. Rank, occupation, wealth, patronage, and household status intersected with population classifications. The experiences of a senior official, a merchant, and a dependent artisan could differ greatly despite their inclusion under the same broad designation.
Economic and cultural roles
Some Semu merchants participated in commercial partnerships associated with Mongol princes and officials. Central and Western Asian personnel also contributed to fiscal administration and long-distance trade. Such activities could bring considerable influence, but they should not be treated as characteristic of everyone classified as Semu.
Specialists from these communities participated in astronomy, calendrical work, medicine, and translation. Their careers formed part of wider exchanges across Mongol-ruled Eurasia rather than the activity of a single unified Semu culture.
Settlement also encouraged adaptation to local institutions and practices. Some families adopted Chinese surnames, acquired Chinese literary learning, or formed local ties through marriage. These developments did not necessarily erase inherited affiliations immediately, and cultural adaptation varied considerably between communities.
After the Yuan
The Yuan court’s loss of most of China in 1368 ended the political framework in which Semu had functioned as a major administrative designation. Under the succeeding Ming dynasty, people descended from Yuan-period migrant communities followed different trajectories: some remained in China, while others moved or continued to serve Mongol rulers.
Over time, many resident families became integrated into local society. Some Muslim lineages associated with these migrations formed part of the historical background of later Hui communities. This was not a simple transformation of Semu into Hui: Semu included non-Muslims, and Muslim settlement in China both preceded and continued beyond the Yuan period.
Historical interpretation
The central interpretive problem is how to describe real inequalities without turning a flexible classification into a rigid, all-encompassing social system. Describing the Yuan order as a four-tier “caste system” can obscure differences between institutions, changes over time, and the importance of office, household affiliation, and personal connections.
Conversely, the variability of the term does not make its consequences merely nominal. Examination rules and appointment practices show that group classification could have practical effects. The most useful distinction is between Semu as an administrative umbrella, the particular communities placed beneath it, and the individual careers and social relationships that often crossed those boundaries.