On the Rectification of Names

A Provincial Examination Essay on the Governance of Interfaces Between Systems


The following essay was submitted at the Zhejiang provincial examination (xiangshi, 鄉試) in the autumn of 1893 by one Chen Zhaowen (陳昭文, 1867–1931), a candidate from Hangzhou sitting the examination for the third time.

The question, drawn from Analects 13:3, required candidates to expound upon the Master’s teaching on the rectification of names (zhengming, 正名):

If names are not correct, speech does not accord with reality. If speech does not accord with reality, affairs cannot be brought to success.

Chen’s response is formally competent — a standard eight-legged essay (baguwen, 八股文) of orthodox structure. What has attracted attention since its rediscovery during the digitisation of late Qing provincial archives in 2019 is the unusual concreteness of the argument. Where most candidates treated the rectification of names as an exercise in moral philosophy, Chen describes what appears to be the actual governance of communication between administrative systems, with a specificity that has led several commentators to note a presumably coincidental resemblance to modern principles of systems interface design.

The chief examiner, Wang Qichang (汪啟昌), noted in his evaluation: “The candidate’s argument is correct in substance but excessively particular for a question of philosophical principle. He writes as though the Master were issuing administrative regulations rather than teaching the foundations of governance. The form is adequate. The spirit is that of a bureau clerk.”

Chen received a failing mark. He passed the examination on his fourth attempt, in 1896, with a less unusual essay on the duty of the minister to remonstrate with the sovereign.

The translation preserves the eight-legged structure. Section headings have been added for the convenience of readers unfamiliar with the form.


破題 pò tí — Breaking Open the Topic

The Master teaches that the ordering of names is the foundation upon which all communication between systems depends. Where names are disordered, no exchange of information — however sincere in intention, however necessary in substance — can arrive at the result that both parties intend.

承題 chéng tí — Receiving the Topic

For the affairs of governance require that many systems, each serving its distinct function, communicate with one another in terms that are mutually understood. The bureau of revenue must speak to the bureau of works; the bureau of works must speak to the bureau of war; each must receive from the others information in a form it can act upon. The reliability of these exchanges depends not upon the goodwill of the officials — for officials are generally well-intentioned — but upon the precision of the names by which things are called. A name that means one thing within the bureau of revenue and another within the bureau of works is not a shared name at all. It is a source of error disguised as a term of art.

起講 qǐ jiǎng — Beginning the Discussion

I venture to say that the Master’s teaching applies with particular force to the governance of interfaces — those points at which one system must address another. Within a single bureau, names may be informal, abbreviated, even idiosyncratic, for those who use them share a common understanding born of daily proximity. But at the boundary, where one system speaks to another, every name must bear exactly the meaning that both parties have agreed upon, and no other. The interface is the point of greatest vulnerability in the governance of systems, for it is where the assumptions of one bureau meet the assumptions of another, and where any discrepancy between them becomes not a local inconvenience but a failure of the whole.

入手 rù shǒu — The Initial Argument

Consider the practice of the well-governed bureau. When it receives a petition, the form of the petition is prescribed: what information must be provided, in what order, in what terms. When it issues a response, the form of the response is likewise known. The petitioner does not invent his own format and hope the bureau will understand it; the bureau does not alter its format from one season to the next and expect its petitioners to adapt without notice. Both parties submit to a shared convention that is published, stable, and enforced, and it is this convention — not the cleverness or adaptability of either party — that makes the exchange reliable.

The bureau that changes the form of its responses without warning is not exercising authority. It is abdicating it. For the authority of a bureau resides not in its power to act as it pleases but in the reliability of its conduct — in the certainty that what it promised yesterday it will honour today. A response whose form cannot be predicted is not a response. It is an event.

起股 qǐ gǔ — First Parallel Pair

When the name of a resource is stable, the communication of the realm proceeds in good order. If a granary is called a granary today and will be called a granary tomorrow, and if every bureau that must inquire about grain knows to address its inquiry to “granary” and will receive in return a reply of known form — then the exchange requires no negotiation, no preliminary correspondence, no clarification. The name is sufficient. It is a reliable guide to the thing it designates and to the manner in which that thing may be addressed. The name carries the convention within itself.

But when names shift, the communication of the realm is thrown into disorder. If the bureau of revenue calls it a “granary” while the bureau of works calls it a “grain depot” and the bureau of war calls it a “provisions facility,” then three inquiries are sent regarding the same resource under three names, and three replies are received that cannot be reconciled — not because the replies are wrong, but because no one recognises that they refer to a single thing. Disorder enters not through malice or incompetence but through the quiet divergence of nomenclature, which is the most common and least dramatic form of administrative failure.

中股 zhōng gǔ — Middle Parallel Pair

Some have proposed that efficiency would be better served if the form of the reply were determined not by the bureau that responds but by the bureau that requests. Under this arrangement, the bureau of works, which cares only about the quantity of materials on hand, would ask for quantity alone; the bureau of war, which cares only about the location of provisions, would ask for location alone; and each would receive only what it needs, in the arrangement it prefers. The argument is that the present system obliges the requesting bureau to receive much information it does not require, and that this excess is wasteful.

But this is a false economy, for it transfers the burden of accommodation from the many to the one. The responding bureau must now maintain not a single form of reply but a separate form for each petitioner, and must understand not only its own affairs but the specific requirements and preferred arrangements of every bureau that addresses it. The servant is made to attend individually to each master. The result is not efficiency but fragmentation, and fragmentation is the first stage of exhaustion, and exhaustion is the condition in which errors are most likely and least likely to be detected. The Master’s teaching is clear: it is the name that must be rectified, not the listener’s expectations that must be individually accommodated.

後股 hòu gǔ — Rear Parallel Pair

The well-governed state maintains a registry of its bureaus. Each is named; its function is described; the form of address to it is published; the form of its reply is documented. A new official, arriving at court for the first time, consults the registry and knows at once how to address every bureau in the realm, though he has never communicated with any of them. He requires no introduction, no private instruction, no period of tentative inquiry. The registry does not perform the work of governance. It makes the work of governance possible, by ensuring that every participant knows the names and protocols of every other before the first petition is sent. It is, in effect, the rectification of names made institutional — the Master’s teaching given permanent administrative form.

The disordered state maintains no such registry, or maintains one that is incomplete, or maintains several that contradict one another. Each bureau must be discovered by inquiry, introduction, or accident. The form of address to the bureau of rites is known only to those who have previously addressed it and received a reply; the form changes when a new director is appointed; and the bureau itself may be reorganised, renamed, or absorbed into another without notice to those who depend upon it. In such a state, officials spend more of their effort discovering how to communicate than in communicating, and the business of governance is subordinated to the mechanics of finding one’s interlocutor. This is the condition the Master warns against: the state in which the people have nowhere to put hand or foot.

束股 shù gǔ — Binding Pair

The Master’s teaching is therefore not a principle of philosophy alone but a principle of continuous administration. To rectify the names once — to establish a convention, publish a registry, and consider the matter settled — is to have understood only the beginning of the teaching. Names decay. A name that was correct in the reign of one sovereign may be misleading in the reign of the next, as the thing it designates changes while the name remains. The bureau of works that once oversaw construction may now oversee procurement; the name persists, the function has shifted, and every bureau that addresses it under the old understanding will receive a reply it does not expect, from a bureau that is no longer what its name declares.

The rectification of names is therefore not a task but a practice — perpetual, attentive, without a terminal point. The governor who rectified the names once and considers his work complete has not understood the Master’s teaching. He has understood only its first application. The teaching is not that names should be corrected, as though correction were an event. It is that names must be corrected — continuously, as a discipline of governance, because the relationship between name and reality is not a problem to be solved but a condition to be maintained, and the moment one ceases to maintain it is the moment at which the affairs of the state begin, imperceptibly at first and then with gathering momentum, to fail.


Note on the candidate’s subsequent career: Chen Zhaowen passed the provincial examination in 1896 and received an appointment to the Bureau of Telegraph Administration (電報總局) in Shanghai, where he spent eleven years managing the coordination of message formats between regional telegraph offices. In a letter to his younger brother, dated 1903, he described his duties as follows: “I ensure that the stations call things by the same names. It sounds like a small thing. It is my entire occupation, and it never concludes.” He retired to Hangzhou following the fall of the dynasty and maintained an extensive correspondence with former colleagues on the subject of terminological standardisation until his death in 1931. He does not appear to have owned or operated a computing machine.