When Two Departments Disagree
The fight over a term is almost never about language.
Most terminology disputes are not linguistic; they are two parts of an organisation having meant different things for years.
The Dispute Behind the Definition
A translation project stalls. Two glossaries exist, and they contradict each other. Legal uses one word; Engineering uses another; nobody agreed because nobody checked. The terminologist hired to resolve the problem discovers that the dispute is not linguistic at all — the two departments have been describing subtly different things for years, and the terminology merely made the disagreement visible.
This is the normal case. Terminology work, as the discipline was formalised by Eugen Wüster in Vienna, rests on a foundational principle: a term is not a label you argue about; it is a marker attached to a concept. If two parts of an organisation cannot agree on a term, the productive question is not which word is right but whether they are pointing at the same concept in the first place. Wüster's foundational work on the General Theory of Terminology, developed through the mid-twentieth century, made this the axiomatic starting point for the discipline.
A term base only works if it is in the translator's tool at the moment of writing, not in a document somebody must remember to open. The database nobody reads
The first task, then, is conceptual comparison — not voting on words, and not checking which term appeared in print first.
The Structural Reasons It Happens
Organisations grow sideways. A company that began with an engineering department adds procurement, then legal, then a compliance function, each inheriting the language of whoever wrote the founding documents for that unit. The International Organization for Standardization addresses this in the ISO 704 standard on terminology work principles, which specifies that concept systems — the logical relationships between terms in a domain — must be mapped explicitly before terms are assigned, precisely because intuitive naming diverges without that structure.
When the divergence is discovered late, usually at translation or at audit, the cost is significant. A translator rendering a legal contract will choose one of the competing terms and thereby appear to endorse a position in a dispute they were unaware of. The Court of Justice of the European Union, which handles texts across multiple legal systems and multiple languages simultaneously, requires controlled terminology precisely because an inconsistency in a source text becomes a proportionally larger inconsistency when that text is reproduced in twenty-four languages. The same pressure operates in any multinational organisation, at smaller scale.
Terminology work is deciding that within this document set a thing has exactly one designation, and enforcing it. One object, one name
The institutional terminologists who work within large bodies — at the European Commission's Directorate-General for Translation, at the United Nations Office at Geneva, or at bodies connected with the terminology standardisation work housed at Infoterm in Vienna — do not usually impose terms. They facilitate the process by which a responsible subject-matter authority within the organisation makes a documented decision. The documentation matters as much as the decision itself: a term agreed verbally and never written into a managed term base will be re-litigated the next time a new employee encounters the conflicting usage.
The first task, then, is conceptual comparison — not voting on words, and not checking which term appeared in print first
What Resolution Actually Requires
A terminology dispute with an organisational cause needs organisational resolution. That means identifying who has authority over the concept — not over the word — and securing a decision from that authority in writing. It then means propagating the decision into every system where the old term persists: style guides, translation memories, contract templates, technical specifications, training materials. The terminologist's job is to map the chain, not to adjudicate the argument.
This is unglamorous and frequently resisted. Subject-matter experts in each department generally believe their usage is obviously correct, and they are often right that it is correct within their own conceptual system. The problem is that two locally correct systems can be mutually inconsistent — and the place where that inconsistency bites is always the text that crosses the boundary: the contract, the translated manual, the regulatory submission reviewed by an external authority who has access to both departments' documents and no reason to guess which one governs.
The terminologist's real work is making that boundary visible before the text does it for them.