Terminology Management 5 September 2026
Term Validation: How Do You Know an Equivalent Is Reliable?
How do you know that a translation equivalent is correct? We work through the steps of term validation, source reliability and the role of expert approval, with examples.
How do you know that a translation equivalent is correct? Term validation is the process of testing systematically whether an equivalent really is reliable. The purpose is to stop a wrong or out-of-context equivalent from entering the termbase, because an error that gets into the base is repeated in every translation that follows.
Contents
Hann sets the priorities of technical translation as precision and intelligibility, and stresses that choosing the wrong term can have serious consequences. Validation is therefore the most critical yet most frequently skipped step in term management. This article looks at what term validation is, how source reliability is measured, the role of expert approval and how all of it is put on record.
What Does Term Validation Guarantee?
Validation moves an equivalent from “probably correct” to “proven correct”. When a translator finds an equivalent for a term, that is an initial hypothesis. Validation tests the hypothesis: does the equivalent really name this concept, does it hold in this field, is it confirmed in a reliable source?
This guarantee prevents errors from spreading. If an unvalidated equivalent enters the base, every translator who uses it inherits the same error and the error hardens into a corporate “fact”. An error caught at the validation stage, by contrast, is resolved with a single correction. Early validation is the cheapest correction there is.
Validation also gives the translator confidence. A translator who knows the equivalent has been tested uses the term without hesitation and does not research it afresh each time. Validation therefore both raises quality and increases the speed of later work. A term validated well once is used with confidence for a long time and becomes common ground for everyone on the team.
How Is Source Reliability Measured?
Not every source is equally reliable. When you validate an equivalent, you look at whether the source is expert in its field, how current it is, and whether it agrees with other independent sources. An equivalent prepared by field specialists, kept current and confirmed by more than one source is preferred to one taken from a single general dictionary.
| Criterion | Reliable source | Weak source |
|---|---|---|
| Expertise | Compiled by a field specialist | General purpose |
| Currency | Recent and maintained | Out of date |
| Corroboration | Several sources agree | A single source |
| Context | Gives field and usage | A bare list |
Context information is in itself an indicator of reliability. A source that gives the field, the usage and an example alongside the equivalent is far safer than one that merely lists equivalences. This is exactly why Hann’s dictionary attaches a field code and a context to every equivalent: the equivalent is presented not in a vacuum but inside its use.
Who prepared a source, and for what purpose, matters too. A product manufacturer’s official documentation is the most reliable source for terms specific to that product; the glossary of a specialist association reflects the established usage of that discipline. Amateur translation forums or lists of unclear provenance, by contrast, may be quick but are low in reliability. Validation measures a source’s competence, not its speed.
The Risk of Trusting a Single Source
The most common mistake in validation is trusting a single source blindly. Dictionaries, especially alphabetical ones, list the possible equivalents of a term but often fail to show which one holds in which field. Taking the first entry and moving on is a near-guaranteed error with polysemous terms.
The homonymy examples Hann gives illustrate the risk well: the German Leiter means “conductor” when used as a masculine noun and “ladder” when used as a feminine one. A translator who does not check the context and the grammatical information may take the first equivalent in the dictionary and stray into an entirely different concept. A single source keeps this trap invisible.
Validation therefore looks for agreement between at least two independent sources. If two different reliable sources give the same equivalent in the same field, confidence rises markedly. Where there is no agreement, the equivalent is suspect and deeper research is needed. Cross-validation is the golden rule of term work.
Context and Field Checking
The first question to ask when validating an equivalent is which field it holds in. The same term may name one concept in the automotive field and quite another in semiconductors. If the field of the equivalent does not match the field of the text being translated, the equivalent has been taken from the right dictionary but from the wrong context.
The second check is how the equivalent behaves inside a sentence. Does the term occur with natural companions in the target language, or does it carry the pattern of the source language? Context checking tests the equivalent not only at dictionary level but at the level of use. A correct term used in the wrong collocation is still wrong.
Field codes are a powerful instrument here. Tagging every equivalent with a field, as Hann’s system does, makes context checking almost automatic. Once you select the field along with the equivalent, it becomes hard to pick up a term from the wrong context. The material in our knowledge centre shows this check step by step.
Context checking also brings the inconsistencies of the source text itself to light. If the source author has used a concept loosely, that looseness is noticed while the equivalent is being validated, and it is put right. Validation therefore tests not only the translation but the understanding the translation rests on. Good validation can leave a text clearer than its source.
The Role of Expert Approval
Some terms call for an approval that goes beyond linguistic validation. With safety-critical, legal or highly technical terms, only a field expert can confirm that the equivalent really holds in context. The expert catches fine distinctions that the translator would not see.
Hann notes that the technical translator often has to work without fully understanding the source text, a situation that almost never arises in general translation. Expert approval closes precisely this gap in understanding. The translator validates the language and the expert validates the concept; together they form a complete guarantee.
Expert approval is not needed for every term; that would slow the process down needlessly. Approval is reserved for terms that are risky or uncertain. Deciding in advance which terms go for approval keeps validation both quick and safe. Picking out what is critical and concentrating on it is the intelligent use of resources.
Nor is the collaboration between expert and translator one-way. The translator catches linguistic inconsistencies the expert has overlooked; the expert catches conceptual errors the translator could not see. The soundest termbases emerge where these two perspectives meet regularly. Approval is not a formality but the joint product of two kinds of expertise.
Putting Validation on Record
A validated term should be recorded together with its validation information: which sources confirmed it, who approved it and on what date. Without that information, the whole validation has to be done again from scratch the next time the term is questioned. The record makes the effort permanent.
The record also delivers accountability. When the reason a term is tied to a particular equivalent is visible along with its sources, there is something concrete to fall back on whenever the decision is disputed. This matters especially in texts that are legally and technically sensitive; the validation trail turns into a document of assurance.
Finally, validation is not a one-off. Sources may age and a field may change; an equivalent that was once correct can lose its validity over time. A record of the validation date shows which terms need to be reviewed again. A base that is maintained keeps validation a living process. The base then remains not a time capsule but a live resource that is updated along with its field.
The Türkiye context
In Türkiye, term validation takes extra effort because reliable, field-coded Turkish sources are scarce. In many technical fields there is as yet no comprehensive, maintained Turkish termbase, and the translator frequently has to derive an equivalent by working from an English source. That makes cross-validation all the more necessary.
The glossaries of the Turkish Standards Institution, the publications of the professional chambers and university resources are all valuable, but their coverage is scattered. Trusting a single Turkish source is therefore risky; the equivalent has to be confirmed both against established usage and with a field expert.
The agglutinative structure of Turkish calls for extra care during validation. A term may look different in its inflected forms, and these differences in shape can be missed when it is compared with a source. Newly coined equivalents may also not have settled yet; in that case the target reader has to be taken into account in order to test that the equivalent really is understood. At our office in Ankara Kızılay we validate critical technical terms together with the client’s field expert and put the approval information on record. The equivalent then rests on ground that is sound both linguistically and technically.
What does this mean for you?
If you are having technical or legal documents translated, term validation is a direct risk management instrument for you. An unvalidated equivalent is the source of the most expensive errors. To reduce that risk:
- Ask for critical terms to be confirmed by at least two independent sources.
- Plan for field-expert approval of safety-critical, legal and highly technical terms.
- Require the record to state which field each equivalent holds in.
- Keep the validation information in your termbase and reuse it on later jobs.
We validate critical terms through cross-referenced sources and expert approval, and we put the approval information on record; our translation services include that assurance. To have us assess a sensitive text, you can reach us through our contact page. This article is for general information and is not legal advice.
Frequently asked questions
What is term validation?
It is the process of testing systematically that the equivalent of a term really is correct and reliable. The purpose is to stop a wrong or out-of-context equivalent from entering the base and to prevent that error from being repeated in later translations.
How do I tell whether a source is reliable?
A source being expert in its field, being current and agreeing with more than one independent source are all indicators of reliability. Trusting a single dictionary blindly is risky.
Is a single dictionary not enough?
Usually not. Dictionaries often fail to show context and field; the same term means different things in different fields. The equivalent has to be validated against at least a second source and against the context of the field.
When is expert approval needed?
It is needed above all for safety-critical, legal or highly technical terms. A field expert confirms that the equivalent really holds in context in a way a linguist could not. Approval is requested not for every term, but only for those that are risky or uncertain.
Sources
- Hann, M. (1992). The Key to Technical Translation, Vol. 2. John Benjamins.
- Sager, J. C. (1990). A Practical Course in Terminology Processing. John Benjamins.
- ISO 704:2022. Terminology work – Principles and methods.