For most of my working life, I have sat in academic meetings and said very little.

It was not because I had no view. It was because I often could not find the thing we were supposed to be discussing. A proposal would arrive wrapped in familiar terms: stakeholders, value creation, impact, strategic alignment, delivery. Everyone seemed to understand them. I usually felt that they had delayed understanding.

The language sounded adult and professional. It also sounded base and puerile. It reduced people, projects and disagreements to a few approved tokens.

A researcher became a stakeholder. Teaching became delivery. A difficult question became a challenge. A disputed objective became an opportunity for alignment. At the end of it, the sentence often made no claim that could be tested or resisted.

This is the language often called MBA speak. That label is unfair if it is taken literally. The problem is broader than business schools. It is a managerial register that has spread through universities, charities, public bodies and large organisations. It promises clarity while making specific responsibility harder to see.

Why the language survives

There is a practical reason this language thrives in meetings.

Specific statements create obligations. If someone says, “We will cut laboratory and seminar sessions to save money,” people can argue about the decision. They can ask how much money will be saved, what students will lose, and whether the saving justifies the loss.

If the same decision becomes “a strategic realignment of the teaching offer”, discussion becomes harder. The phrase invites agreement before it invites comprehension.

Managerial language makes disagreement more orderly. It can smooth over conflicts of interest, uncertainty and power. It can also allow a meeting to finish without anyone having plainly said what they want done.

That may be useful when the aim is to maintain a fragile consensus. It is a poor habit for a university.

Universities are supposed to be places where claims are defined, evidence is weighed, terms are challenged and arguments change in response. Their internal language should reflect that work. Too often, it does the opposite.

Where AI comes in

AI did not create this problem.

The language was already there, in reports, committee papers, grant applications, strategic plans and emails. Large language models absorbed it because it is common, predictable and rewarded. Give an AI a prompt about a university initiative and it will readily offer “stakeholder engagement”, “meaningful impact” and “a robust framework”. Those phrases are statistically safe. They also travel well from one context to another.

AI changes the speed and volume.

A vague register that once took an afternoon to produce can now fill a page in seconds. A paper can sound complete before its author has worked out what it says. A committee can receive a fluent summary that conceals every important choice. The prose acquires the appearance of work before the work has been done.

That is why the argument about AI and academic writing often misses the point. We are treating the machine as the source of a weakness it learned from us.

The danger is not a robotic voice. The danger is a familiar institutional voice, reproduced at industrial scale.

A different standard

There is no mystery about the remedy. It is the ordinary discipline of writing well.

Name the people involved. Say what they will do. Use verbs that describe an action. Give an example. Put a number on a claim where a number is available. Admit a disagreement where there is one. State the judgment you are asking others to accept.

A colleague should be able to return a draft with a few blunt questions:

What does this mean?

Who is responsible?

What will happen?

How will we know?

What is your view?

Those questions are not hostile to administration. They are a minimum standard for it.

Some academic writing has always been difficult, for good reasons. A difficult subject may require technical terms, careful distinctions and slow reading. That is different from language that creates a fog around a simple decision.

The first kind asks the reader to think. The second asks the reader to nod.

I know which one I would rather encounter in a meeting.

Concrete nouns. Real verbs. A point of view.

The video that prompted part of this reflection is below.