From Here to Ontology

Continuing Stephen Clarke’s Course

In the short time that Stephen Clarke wrote on his From Here to Ontology Substack, he managed what many of us spend a career circling toward: a synthesis. He saw the future shape of our profession—its convergence with AI governance, its grounding in ontology, and its enduring commitment to evidence and meaning—and he began to chart it with both precision and grace.

Stephen’s passing leaves an immense loss for the records and information management community, particularly for those who shared his conviction that ontology is not a fringe concern but the intellectual architecture of our field. Yet in what he wrote, and in what he provoked others to think about, he has left a clear direction. His course deserves to be continued.

The Architecture of Meaning

His first post, A Call to Action: Ontology—the Future of Records and Information, read like a manifesto disguised as reflection.

“Ontology is not a new discipline; it is the oldest—the study of being. What’s new is our need to apply it to information.”

That sentence captures the bridge Stephen built between ancient questions of meaning and modern architectures of data. He argued that ontology is not an abstraction but a tool for understanding how we describe, connect, and act upon information. It is the framework through which metadata becomes more than description—it becomes understanding.

He saw metadata as the language of shared existence between humans and machines, and ontology as the grammar that makes that language intelligible. At a time when records professionals are asked to navigate complex data environments, from M365 workspaces to semantic graphs, Stephen reminded us that our core expertise is conceptual. We know how to structure meaning. The challenge is to make that visible and valuable in the age of AI.

In his writing, ontology became a form of ethics: a recognition that naming and classifying are acts of power, and that our choices about information structure shape what can be known, retrieved, or forgotten. He wanted us to be conscious of that and to design with integrity.

The Convergence of Governance

In From Charters to ChatGPT: Why the Ancient Science of Diplomatics Still Matters, Stephen addressed the accelerating conversation about how to regulate artificial intelligence.

“AI governance and information governance are not parallel tracks; they are the same conversation seen through different lenses.”

That observation was more than clever symmetry. It was a provocation. He insisted that our profession already holds the conceptual tools the AI community is now struggling to invent: accountability, provenance, transparency, lifecycle control, and ethical stewardship. In records terms, these are not new; they are our vocabulary.

Stephen called for a recentering of information professionals in the governance of AI systems—an insistence that evidence and explainability begin with the quality of the underlying information. To him, every AI model was only as trustworthy as the records and data that trained it. Without governed information, there could be no ethical automation. In this way, Stephen positioned the records discipline as the moral and structural foundation for responsible AI—a connection few have made with such clarity.

Information as a Living System

In The Illusion of Neutrality: Why Data Is Always Political, he wrote:

“Information management is not just about systems. It is about people—their language, their trust, their sense of continuity.”

That perspective gave his work a humanity too often missing in discussions of data and governance. Stephen’s background in anthropology informed his understanding that information is not neutral. Every record reflects a community of practice, a cultural vocabulary, a set of implicit agreements about what matters.

He viewed information ecosystems as living systems—adaptive, self-correcting, and deeply cultural. Ontology, in his view, was the ecology of information: the pattern of relationships that makes a knowledge environment sustainable. This was not sentimentality. It was systems thinking. He saw that technology alone cannot preserve meaning; only communities can. And our task, as records professionals, is to maintain the continuity between technological change and human purpose.

Standards, Stewardship, and the Social Contract

Stephen’s long engagement with standards work gave him a perspective that transcended mere compliance. In ISO Standards for Archives and Records Management, he wrote:

“Standards don’t tell us what to think; they give us a common language for thinking together.”

He treated standards as living documents—tools of collaboration and shared responsibility. For him, standards like ISO 15489 and 30301 were not checklists but expressions of professional ethics: the social contract that underpins trust in information.

Stephen also anticipated something many of us are now confronting: that AI will demand new standards for explainability, interoperability, and provenance—and that these must be built on the foundation of existing records management principles. His writing urged us to see continuity, not rupture, between traditional records management standards and emerging AI frameworks. In this way, he modeled what might be called integrative governance—the capacity to hold together technical precision, ethical reflection, and human consequence.

A Course Worth Continuing

Taken together, Stephen’s Substack essays mapped the next phase of records and information governance: ontology as infrastructure, giving structure to meaning; governance as ethics, linking records to accountability; information as ecology, recognizing human context as system design; and standards as dialogue, connecting disciplines through shared language.

Each of these themes remains open and unfinished, and that is both the challenge and the invitation he leaves us. Stephen Clarke’s course now becomes ours to continue: through the work of standardization, through interdisciplinary collaboration, and through the insistence that meaning and accountability must coexist in every digital system we build.

“We don’t manage information because it is easy to do so. We manage it because it anchors truth in a changing world.”

Haere rā, Stephen—go well. We’ll keep going from here.