China Just Changed What Counts as a Graduate Degree

The headline writes itself: some graduate students in China may no longer have to write a traditional thesis. The policy behind the headline does something more specific. It rewrites what counts as proof that a professional-degree candidate deserves the degree. The requirement of proof stays. Its form changes, and the places where proof must be produced multiply.

The General Office of the Academic Degrees Committee of the State Council has issued a guiding opinion allowing professional master's and doctoral students to apply for degrees on the basis of practice outcomes. Five categories are slated for the first wave of pilots: translation, agriculture, clinical medicine, accounting, and art. The policy follows an earlier experiment in engineering professional degrees, in which more than 140 master's and doctoral students applied for degrees through practice outcomes rather than conventional dissertations. The official framing is blunt about the direction: professional-degree training should move "from a knowledge-based orientation to a competence-and-contribution-based orientation."

That phrase is the policy's centre of gravity. It explains why the five pilots are not a soft landing for students who dislike writing. They are a test of whether professional degrees can be assessed by something closer to what professionals do, and whether universities, supervisors, and industry partners can build a credible apparatus around that assessment.

The thesis was never the only problem. The evidence was.

The old system had a tidy logic. A dissertation is a self-contained artefact that a committee can read, interrogate, and judge. Its weakness was equally tidy: it rewarded the ability to construct a defensible written argument, which is not the same thing as the ability to solve a problem in a hospital, a translation bureau, a farm, an audit file, or a concert hall.

Practice-based applications invert the evidentiary burden. The candidate must show not only that something was done, but that it was done by them, under live constraints, with observable effects, and to a standard that industry peers recognise. The Opinion requires the practice outcome to be independently completed by the student, or to represent the core content of work in which the student was the principal contributor. It also requires a full chain of documentation: topic justification, feasibility review, process supervision, outcome appraisal, expert review, and contribution verification.

The "no thesis" reading misses what is happening. The text may shrink. The evidence burden does not. A practice-based candidate may end up producing more written material, not less. Not a single narrative, but a portfolio of justification, records, and appraisal that surrounds the outcome and makes it legible to a committee.

What "harder than a thesis" looks like in practice

The engineering pilot offers the clearest evidence that this route is not a shortcut. Wang Haoran, a professional doctoral student at Nanjing University, applied for a doctorate in electronic information on the basis of a distributed fibre-optic sensing system for pipeline safety monitoring. The system had been deployed in the South-to-North Water Diversion Project and a regional water supply project, and had also been exported to a European customer. His supervisor, Professor Zhang Xuping, described the evidentiary threshold in terms that leave little room for fantasy: the outcome required a user application report, third-party testing, and authoritative industry appraisal. "People think it is easier without high-level papers," she said. "It is actually harder. Without a real application, you cannot do it at all."

Wang put the comparison directly. The practice-based defence was more demanding than a conventional dissertation, because the work had to be installed on site, negotiated with owners, and sustained in live operating conditions. The committee did not evaluate a text. It evaluated a system that had to function.

The agriculture pilot adds a second kind of evidence: the field. China Agricultural University's first cohort of five master's students to defend practice outcomes included a candidate who developed a sulphur-based fertiliser for saline-alkali wheat fields, another who worked on a phosphate-tailings conditioner, and another who ran four consecutive growth seasons of household surveys in the Erhai Lake basin to identify yield constraints among smallholders. The defence committee was not composed solely of academics. It included a chief expert from the National Agricultural Technology Extension Service Centre and a deputy secretary-general of the China Phosphate and Compound Fertiliser Industry Association, who questioned cost accounting, environmental indicators, the logic of the process research, and the representativeness of the field data. The university's vice-president stated the principle explicitly: "change the form, do not lower the standard."

These cases are not ornaments. They show what the policy looks like when it is working: an outcome that exists outside the library, a paper trail that makes the candidate's contribution traceable, and evaluators who bring industry standards into the room.

Five pilots, five different evidentiary problems

The five categories were not chosen at random. Each has a strong industry orientation and a weak fit with the conventional dissertation. Each also presents a distinct assessment problem, and the policy's insistence on category-specific Basic Requirements reflects that.

Translation is the clearest case of a professional practice that cannot be reduced to a thesis about translation. The secretary-general of the national MTI steering committee, Zhang Ailing, has proposed a formula for a valid practice outcome: translation output produced under a commissioned scenario, real or simulated, plus a practice-based analytical report. The hard part is not the translation itself. It is contribution verification in team projects, which are the norm in the industry. A team translation of United Nations documents, for example, cannot be attributed wholesale to one student. The candidate must show which module they independently handled, and must retain division-of-labour records, draft versions, and revision traces. The assessment standard is equivalence of competence, not identity of format: the practice route must test professional practice, problem-solving, and professional reflection at the same level that the thesis route tests theoretical synthesis and scholarly argument.

Agriculture has the advantage of measurable outcomes: yield, fertiliser efficiency, adoption area. It has the disadvantage of uncontrollable variables. Weather, pests, and soil heterogeneity can swamp individual competence. The China Agricultural University defence handled this by requiring field validation and by asking industry experts to probe whether the results held beyond the demonstration plot. The structural risk remains: a candidate can do everything right and still lose a season to hail. The assessment regime must distinguish bad luck from bad work, which is easier to say than to institutionalise.

Clinical medicine carries constraints that the other pilots do not. Patient safety, ethics review, and data privacy limit what can be publicly demonstrated. A practice outcome in clinical medicine might be a redesigned care pathway, a diagnostic protocol, or a device, but the evidence must be assembled without exposing patients or breaching confidentiality. The category-specific requirements will have to specify what counts as verifiable evidence under those conditions.

Accounting faces the opposite problem. The most valuable practice outcomes are often buried inside commercial confidentiality. An audit case, a financial diagnostic report, or an internal-control redesign may be exactly the right kind of outcome, but the evidence that would persuade an external reviewer may be the evidence the client will not release. The pilot will have to develop redaction and verification protocols that preserve assessability without destroying the commercial value of the work.

Art has the most intuitive outcomes (works, performances, design projects) and the most contestable evaluative criteria. Industry experts, curators, and critics can judge artistic quality, but the risk of circle-based or personality-driven evaluation is genuine. The policy's emphasis on "just and fair evaluation mechanisms" and conflict-of-interest avoidance is especially relevant here.

The common thread is a shared discipline, not a single template. Every category must specify where the outcome comes from, what form it takes, how individual contribution is established, and what evidence makes it assessable.

The quiet shift in who holds evaluation power

Who gets to judge the work is changing alongside what gets submitted. That shift is quieter than the headline about theses, and more consequential.

Under the conventional dissertation regime, the academic community largely controls the standard. External examiners are academics. The criteria are scholarly. Industry input, where it exists, is advisory. The Opinion moves deliberately in the other direction. It calls for deeper industry participation, clearer responsibilities for industry supervisors, and shared standards across training, practice, degree awarding, and commercialisation.

In the agriculture defence, industry experts did not merely observe. They cross-examined cost structures and environmental claims. In translation, the proposed model makes the industry supervisor a source of live projects, a bearer of industry standards, and a participant in the defence.

This is a transfer of evaluative authority, not a rhetorical gesture. Once industry practitioners sit inside the assessment process with substantive questioning power, the boundary of what counts as a legitimate professional-degree outcome shifts. A dissertation that satisfies academic conventions but solves nothing that a practitioner would recognise becomes harder to defend. Conversely, a practice outcome that works in the field but cannot be articulated in professional terms also fails, because the candidate must still produce a reflective account that makes the work intelligible and defensible.

The risk is that industry participation becomes decorative. If industry experts are chosen for their willingness to sign off rather than their willingness to challenge, the reform collapses into a new formality. The policy's emphasis on conflict-of-interest rules and fair evaluation acknowledges that risk. The actual safeguard will be in the composition of committees and the willingness of universities to tolerate uncomfortable questions.

The equity problem nobody wants to name

Practice-based routes are resource-intensive. They require live projects, industry partners, patient supervisors, and often money for fieldwork, testing, or prototyping. A student with a well-connected supervisor and an active industry project can assemble the evidence. A student without those advantages may find the practice route riskier than the thesis route.

This is already visible in the early pilots. The engineering cases come from well-resourced teams with deployed systems. The agriculture cases come from a leading agricultural university with established field stations and industry links. The translation discussion explicitly acknowledges that the 374 MTI programmes vary enormously in their access to high-level projects, and argues that local resources (regional publicity materials, tourism texts, small-enterprise manuals) can legitimately serve as practice outcomes as long as the task involves genuine work and the difficulty is professionally meaningful. That argument is plausible, but it does not dissolve the underlying inequality. A regional tourism translation is not the same evidentiary challenge as a United Nations document project, and the assessment system will have to ensure that "equivalence" is not achieved by quietly lowering the bar for less-resourced programmes.

Student risk tolerance adds a second dimension. The early evidence suggests that students with stronger projects and better support will self-select into the practice route, while students with weaker projects or a lower tolerance for uncertainty will stay with the thesis. That is not necessarily a problem in a pilot. It becomes a problem if the practice route becomes a premium option available mainly to students who already have advantages, while the thesis route becomes the residual option for everyone else. The policy's "pilot, evaluate, scale" sequence is designed to prevent premature universalisation, but it does not by itself solve the resource problem.

What the international comparison actually shows

The international experience is often invoked as evidence that practice-based doctorates are normal. The comparison is instructive, but not in the way a simple "they do it too" argument suggests.

In the United Kingdom, the United States, and Australia, professional doctorates have increasingly used portfolio-based assessment rather than a single dissertation. The portfolio typically combines practice artefacts, reflective commentary, and evidence of impact, and it is assessed as a whole rather than as a practice component plus a separate written component. The rationale is not that writing is unimportant. It is that the relationship between practice and written reflection is itself part of what is being assessed: the candidate must show that the practice informed the writing and the writing illuminated the practice.

That model is a useful corrective to two failure modes the Chinese pilot could fall into. The first is treating the practice outcome as a deliverable and the written material as a bureaucratic appendix. The second is treating the written material as a conventional thesis in disguise, with the practice outcome merely illustrating it. The portfolio model insists that both are constitutive of the claim to a degree.

The international literature also indicates that practice-based routes are not inherently more "applied" or less rigorous. They are differently rigorous. They demand a different kind of reflexivity: the ability to explain why a decision was made under conditions of uncertainty, what alternatives were rejected, and what the candidate learned from the gap between intention and outcome. That is a professional competence, not a scholarly ornament.

A different gate, not a lower one

The policy's significance is that it changes the gate, not the height of the wall. The old gate asked: can you produce a defensible written argument about a problem? The new gate asks: can you produce a defensible practice outcome, prove that it is yours, show that it works under live constraints, and articulate what it means in professional terms?

That is a harder gate to pass for students who have been trained only to write. It is also a harder gate for universities to administer, because it requires evidence infrastructure, industry relationships, and supervisors who can judge practice as well as prose. The five pilots will not settle the question of whether professional degrees in China are ready for that shift. They will show where the evidence chain holds and where it snaps.

The first cohorts are already showing what the difference looks like. A student defending a fertiliser that was tested across four growing seasons in saline-alkali fields is not doing something easier than a student defending a literature review. A doctoral candidate whose sensing system is monitoring a live water pipeline is not avoiding rigour; he is being held to a standard that a committee can verify against the pipeline itself. The question is whether the system can make that standard available to the students who do not have a pipeline, a field station, or an industry partner waiting for them.

Until that question is answered, the honest description of the reform is a different kind of proof, with a different kind of risk.