I'm one of those teachers who plays off the energy of the crowd. I'm constantly checking my students' faces to see how they're following — when they're interested, when they're overloaded or bored. Their faces inform a constant stream of micro-decisions, telling me when to speed up, slow down, maybe tell a stupid joke to shift the mood. I want to see them startled by an unexpected question. I'm a philosophy professor, and I basically live for the sight of a new idea dawning over a student's face.

All that changed during the pandemic. My university switched to remote learning, and I was left desperately trying to figure out how to make my classes work in this new environment. I was hardly alone. My Facebook feed was suddenly full of teachers posting advice, pleas for help — and some pretty heated debates about student camera use, in particular. Some argued that forcing students to keep their cameras on during class constituted a clear invasion of privacy. I was convinced by that argument.

My role was to be their teacher, not some behaviour-enforcement officer peering into their personal space. So I told my students that I had no problem with them leaving their cameras off. The result, in retrospect, was entirely predictable: Every single student chose that option, and I ended up lecturing to a sea of black squares. This turned out to be exhausting, isolating, miserable, and my teaching suffered badly.

I hadn't even realised how much my teaching — and the joy I get from it — depended on that facial feedback. Without it, I became just a guy, alone in his basement, shouting into a blank laptop, hour after hour, day after day.

So I tried opening up to my students. I didn't want to place too much pressure on them, but I also felt they deserved to know that too many black squares on Zoom created an unhappy and floundering teacher. The result both radically improved my teaching on this medium and reconnected me with the students. Faces began to reappear, and not just temporarily. In the end, we reached a workable equilibrium — and the success of this mutual effort was a rare bright spot in a very difficult year.

Last year, even as I realised that the cameras-off norm was making me miserable, I wasn't sure how to proceed. I didn't want to impose a strict policy because I could see how wildly different people's lives were in the pandemic and how little I understood about those differences. There are so many reasons students might not want to go on camera — from difficult housing situations to poor technology.

Students are having to hold children and sit with sick parents. Some are worried about having their faces recorded by other students. The utter novelty of the pandemic makes it hard for anyone to grasp one another's circumstances. (A small irony: One of my classes is literally on this very topic. We explore "social epistemology", which includes questions about the limits of our mutual understanding.)

At least one teacher I deeply respect argues that, even without an official policy, exerting any kind of emotional pressure — even the slightest bit — on students to turn on their cameras is a bad practice. And there's a part of me that wants to be enough of a teaching saint to follow that advice. But I'm also a person; isolation is taking its toll on me, too. And that toll rose when I tried to be a fully engaged teacher getting zero response from the other end.

So this semester, on the first day of class, I gave honesty a shot. I told my students that I totally understood their privacy concerns and that they were free to turn off their cameras — either all the time, or occasionally. There were hundreds of good reasons for doing it, I said. They didn't need to explain themselves.

But I also told them that, as more cameras go off, my teaching inevitably suffers, and I gave them my best estimates about the effects: When fewer than a third of students have their cameras off, there is no impact. When the blank screens start to reach the 50 per cent mark, I begin to lose my feel for the class. And once the proportion of turned-off cameras hits two thirds, I feel profoundly cut off from the feedback I need to do even a vaguely decent job.

Some students told me they had never thought about what it was like to teach remotely. (I knew that I had trouble putting myself in their shoes.) And they listened. This semester is light-years better than the last. On most days, I can see somewhere between half and two thirds of my students. No policy, no further comments from me, no enforcement or pleading. I think I see some self-regulation at work in our pedagogical ecosystem. Whenever the ratio of black squares drifts up towards the two-thirds mark, a few students will turn their cameras back on.

I feel so much more alive and engaged. And I've heard from some of the students that they feel more invested in the material when they can see the reactions on other students' faces, even as they appreciate the freedom to vanish when they need to. I suspect that many people have only a mild preference for disabling the cameras — and now some of these students have forsaken that preference to further the communal good.

In the end, I see this as a question of informed choice. Given who I am, it's very predicable that my teaching will get worse as more cameras go off. Students deserve to know that, and take that into account, in their own choices. I suspect that honesty is the best we can do right now.

This experience has also changed how I behave when I'm on the other side of the exchange — in the audience of an online lecture. In that situation, I would almost always prefer to turn my camera off. But now I go camera-on most of the time because of my understanding of the impact of my decision on the speaker.

Right now, our knowledge of one another's lives is slim, gathered as it is through impoverished channels such as Zoom. When our connections are so tenuous, a little trust can go a long way.

• C. Thi Nguyen, an associate professor of philosophy at the University of Utah, is the author of Games: Agency as Art