Learning a Domain From Its Own Categories
The hardest part of cross-industry research is learning a new field fast. There are two ways in, and one of them is built from the domain's own words.
The Real Job
Cross-industry research sounds like it’s about the research. Most of it is learning the domain, the part before you can ask a single useful question. Every new field is a fresh wall: its vocabulary, its categories, and the things practitioners consider so obvious they never say them out loud.
The bottleneck in learning a new domain isn’t information. It’s knowing which categories the domain uses to carve up its own world.
Two Ways In
Ethnographers have a name for the two stances you can take toward a culture you’re trying to understand.
The etic view is the outsider’s. You reconstruct a field from the outside, mining Reddit threads, blog posts, conference talks, whatever has been written about it. It’s available and it’s fast, but it’s noisy: secondhand, contradictory, and increasingly full of machine-generated slop.
The emic view is the insider’s. You try to understand a culture through its own categories, the distinctions its members actually make. It’s the harder path, and the truer one.
Etic is what people say about a field. Emic is how the field describes itself.
An Ontology Is a Domain Going Emic
Here’s the move this whole project rests on. A formal ontology (OWL, STIX, the CIM schema) is a domain describing itself, in its own terms, ratified by its own practitioners. It isn’t commentary about the field. It is the field’s account of what exists and how it all relates.
Reading the ontology is going emic. That’s the ground truth the project is built on, and it’s where the name comes from. Emic is ethnography’s word for the insider’s view.
The Catch
There’s a reason nobody reads ontologies for fun. DMTF’s Common Information Model, the proving ground for this project, is about 1,648 classes of enterprise IT management, written by a committee, for implementers. It’s the insider’s view rendered in the least inviting form imaginable: dense, exhaustively cross-referenced, and quietly assuming you already belong.
The ground truth is right there. It’s just unreadable.
An ontology is a near-perfect map of a domain, written in a font no human wants to read.
What Emic Does
Emic is an explorer that makes that map navigable. It runs graph analytics over the ontology to find the structure hiding in it, writes plain-language field notes on each concept, and lays the whole thing out as a territory you can actually walk.
The features borrow the ethnographer’s toolkit, almost one-to-one:
- The most-connected concepts are the key informants: the hubs everything else defers to.
- The clusters the graph forms on its own are villages, communities with their own character that rarely match the official table of contents.
- The plain-language summaries are field notes.
- The layout is the map, and the whole job is to make it faithful enough to pass for the territory.
Every feature here is an ethnographic idea wearing an engineer’s clothes.
The Takeaway
The Takeaway: The fastest way into an unfamiliar domain is to read how it describes itself. Ontologies are exactly that self-description. They’re just written for machines and committees, not newcomers. Emic is an attempt to make them legible. The rest of this series is how it got built, and what broke along the way.