Incomplete Reform

Last blog post talked about the upcoming engineering education reform. Since it’s not starting from scratch but patching things up, success is unlikely. Today let’s look at an example of starting anew — the model software schools.

The Ministry of Education, spurred by the Indian example, wanted to cultivate engineering-oriented software talent. Why through model software schools? Why not reform existing computer science departments with “high-level, practical, international software talent education reform”? Clearly, the resistance was too great. A bunch of old experts, old scholars — their thinking was already固化 like BIOS. Upgrading BIOS 1.0 to 1.1 is possible, but upgrading to UEFI? No way. What to do when reform isn’t possible? Start fresh. Give them policies: high fees, self-funded, engineering teaching, international cooperation, enterprise cooperation, not allowed to be co-located with CS departments (though私下 collusion happens, with authorities turning a blind eye)…

Model software schools have been running since 2002. The benefits have been well publicized — everyone has seen them. I won’t repeat them. In my personal view, model software schools are as significant as the restoration of the college entrance exam in Chinese education history. Today let’s talk about the problems.

Political textbooks say: the bourgeoisie has compromise, reform is incomplete, resulting in constitutional monarchy. In reality, every class has untouchable “red lines.” Reform always happens within the confines of those red lines. Try crossing them — look at Shang Yang, Liang Qichao, Wen Yiduo. Education reform is the same.

After all these years of software school reform, the problems that could surface have basically surfaced. My feeling is that the Ministry of Education landed a left jab on target, but the right jab never came. The opponent isn’t knocked down, so I throw another left jab. But relying only on left jabs makes it hard to knock anyone down… Why not throw a right hook? Because the right hand is tied by “red lines.”

Let me talk about a few red line issues:

Issue One: Economic lifeline — “big money” can’t beat “public money.” Software schools charge high fees and are self-funded. Does high revenue mean better conditions for students and better treatment for teachers? Not at all. A peasant entrepreneur, a Shanxi coal boss — they’re rich, but compare them to PetroChina or Huaneng? Software schools are like Shanxi coal bosses, relying on tuition fees (especially engineering master’s fees). There’s considerable income. (Insert elementary school math problem: Given that in 2002, undergraduate tuition for 4 years was 40k, and 40k could buy 10 sqm of inner ring housing, find: if tuition grew at the same rate as housing, how much would tuition be now?) But nobody sees the expenses — from millions in annual rent to 80 RMB per network port monthly — all self-funded. And everyone outside sees software schools as Shanxi coal bosses, so every department comes to “skim.” Are software schools really rich? Take a look. On the entire Jiading campus, how many schools don’t have their own buildings? How many are still renting? Like anyone sane, if I could choose, I’d rather be the boss of PetroChina than a Shanxi coal boss. Ultimately, Shanxi coal bosses haven’t found a sustainable self-funding mechanism, always worrying about their “coal mines” being “nationalized.”

Issue Two: Political status — “President Ma” or “Governor Ma.” Engineering talent cultivation requires engineering-oriented teachers. The “three-thirds system” for faculty was the Ministry of Education’s far-sighted vision. One-third with enterprise backgrounds for practical courses, one-third from academia for foundational courses, one-third with foreign backgrounds for internationalization. Over the years, this has been tough on enterprise-background teachers. The university’s evaluation indicators completely ignore their issues. To this day, some enterprise teachers haven’t been promoted to associate professor after eight years. A certain school hired a CTO from industry — still a lecturer. Same for schools: annual evaluations always rank software schools at the bottom. Why? Because what you do has nothing to do with current evaluation metrics. No papers, no research, no patents, no projects. Student employment rates don’t count for anything. High salaries for graduates don’t count for anything. English-taught courses and international students don’t count for anything. In short, you’re a P-grade school. Not only that — I want to fire a teacher? Can’t do it. Public institutions aren’t enterprises. I want to hire someone? Can’t — no PhD, doesn’t meet requirements. I want to pay teachers more? Can’t — must align with university average salaries. I don’t want to offer so many useless “general education” courses? Can’t — that’s a university-wide requirement. I want to offer more practical courses? Can’t — the university isn’t a corporate cheerleader… In the end, a software school is a “province,” not a “country.” It has no sovereignty, and “one country, two systems” doesn’t work here.

Issue Three: Cultural status — “Two Alls” or “Seeking Truth from Facts.” All graduate students must belong to some disciplinary program; all disciplinary programs must be evaluated by uniform criteria. These two statements are unshakeable in today’s universities. The higher education system built over many years in China is based on “disciplines.” From this perspective, the software engineering undergraduate major and the software engineering master’s degree are freaks. They don’t belong to any discipline. Software schools have no discipline! And lacking a discipline isn’t just a stumbling block — it’s the “three mountains” weighing down software schools’ development. This explains why graduate students must go through CS departments, why software schools can’t have PhD programs, why our master’s graduates are evaluated by the Telecom School, why faculty promotions depend on others’ approval, why… No discipline? Two options: first, dismantle the entire discipline-based system; second, find ways to create a new discipline for software schools. Which approach is more thorough reform? Obviously the first. The second approach uses old systems to solve new problems from reform — going backward. Ultimately, software schools remain “wanderers” outside the system. And this “system,” even with emancipation of the mind and seeking truth from facts, cannot be broken.

As in elementary school essays, I must summarize: Incomplete reform means “new shoes, old path.”