ECR Hub

The Infrastructure Science Cannot Buy

Scientific institutions invest heavily in laboratories, instruments, and technology. But some of the most important infrastructure cannot be bought. My colleague Bosco Tjan taught me how much science depends on mentorship, generosity, and people who help others succeed.

When universities talk about research infrastructure, they usually mean things that can be purchased, installed, and placed on an inventory: buildings, laboratories, instruments, computers, data systems.

I have spent much of my scientific career helping to build and use exactly that kind of infrastructure. But one of my colleagues taught me that a successful research environment depends just as much on another kind—one that cannot simply be bought.

His name was Bosco Tjan.

Building a research center

I met Bosco in the 1990s, before he joined the University of Southern California. He was a brilliant vision scientist, technically gifted, intellectually adventurous, and unusually generous with his time.

Years later, we became deeply involved in building USC's brain-imaging research capability.

As scientific director, I thought of my job in very concrete terms. We needed the right MRI system. It had to be installed and tested. We needed research protocols, procedures, training, safety systems, technical expertise, and an organizational structure that would allow investigators from different fields to use the facility effectively.

This, to me, was infrastructure.

Bosco worked on these things too. But he was also doing something I did not fully appreciate at the time.

He was building the people.

If a student had a technical problem, Bosco helped. If an imaging sequence was not working, Bosco helped. If a colleague needed advice, Bosco helped. If someone was struggling to understand a method or develop an experiment, he rarely seemed too busy to talk.

I was his senior colleague, and sometimes I worried about him.

I told him that he could not solve everybody else's problems. He needed to protect his time. He needed to focus on his own research program. There would always be another student with a question or another colleague who needed technical help.

Bosco would listen.

Then he would continue helping people.

At the time, I thought he was giving away time that should have gone into his own science.

Years later, I realized that I had been using the wrong measure.

Bosco was not neglecting the scientific enterprise.

He was building it.

The infrastructure we do not put on inventories

A research center certainly needs instruments, funding, procedures, and technical expertise. Without them, many experiments simply cannot be done.

But equipment does not create a scientific community.

A laboratory also needs people who share what they know. Someone has to help the new student who does not yet know which question to ask. Someone has to remember why a procedure was designed a particular way. Someone has to notice when a junior researcher has lost confidence. Someone has to solve the problem that does not officially belong to anyone.

That is infrastructure too.

It is made of accumulated knowledge, mentorship, trust, generosity, technical judgment, institutional memory, and a willingness to contribute to work from which one may receive little direct credit.

Unlike an MRI scanner, it has no serial number.

Unlike a grant, it has no dollar amount.

Unlike a publication, it produces no citation count of its own.

And that makes it easy to overlook.

What science knows how to count

Scientific institutions are very good at measuring certain forms of accomplishment.

We count papers, grants, citations, patents, awards, and promotions. These are meaningful measures, and I am not arguing that we should abandon them.

But they capture only part of what makes science work.

Consider the senior scientist who repeatedly spends time advising students outside her own laboratory. Or the technician whose knowledge keeps a complex instrument functioning through several generations of trainees. Or the reviewer who spends hours improving someone else's paper anonymously. Or the administrator who quietly solves the logistical problems that make a large collaboration possible.

Those contributions may shape careers and influence science for years.

Yet they may leave almost no visible trace on the contributor's CV.

This raises a question I have increasingly found myself asking:

How much scientific progress depends on contributions that our systems of evaluation do not know how to count?

There is an irony here. If a university buys a multimillion-dollar instrument, it plans carefully for maintenance. Who will operate it? Who will repair it? How will users be trained? When will it need to be upgraded?

But what is the maintenance plan for the human infrastructure of science?

Who preserves knowledge when the person who knows how everything works leaves?

Who recognizes the faculty member whom everyone turns to for advice?

Who protects the technician who has become indispensable?

Who rewards the scientist who makes an entire community better, even when much of that contribution does not appear under his or her own name?

These are not peripheral questions about collegiality.

They are questions about scientific infrastructure.

What Bosco built

In 2016, Bosco was killed in his office at USC by a graduate student from his laboratory. His death was devastating to his family, friends, students, colleagues, and the broader scientific community.

In the aftermath, something became much clearer to me.

I had spent years thinking about what Bosco might have accomplished if he had protected more of his time.

But everywhere I looked, I saw evidence of what he had already built.

Students remembered the confidence he had given them. Colleagues remembered problems he had helped them solve. People remembered his patience, generosity, technical insight, humor, and willingness to make their problems his own.

The things I had once worried were distractions from his work had become part of his legacy.

I thought I had been building the imaging center.

Bosco had been building the people.

And without people, a research center is ultimately just a collection of rooms and machines.

What lasts

Scientific equipment has a predictable life cycle.

Scanners are replaced. Computers become obsolete. Software changes. Buildings are renovated. Today's cutting-edge technology eventually becomes yesterday's equipment.

Human infrastructure has a different life cycle.

A scientist teaches a student how to think, and that student later teaches others. A mentor gives a young researcher confidence at exactly the right moment, and decades later that researcher does the same for someone else. A colleague creates a culture of generosity, and people who experienced it carry that culture to new laboratories and institutions.

We rarely know how far those effects travel.

Universities should continue investing heavily in the physical and technical infrastructure that modern science requires.

But we should pay comparable attention to the infrastructure that cannot be purchased: the people who teach, mentor, troubleshoot, remember, connect, encourage, and quietly make everyone around them better.

Some of the most consequential contributions to science may be precisely the ones that are hardest to count.

Bosco taught me that.