Documentation / QueryTek Review glossary / Linguistic Quality Assurance
Linguistic quality assurance is structured evaluation of language quality against defined criteria — accuracy, terminology, style, and fitness for the intended reader. It differs from proofreading by categorising findings against a rubric rather than simply correcting them, which makes the results measurable.
LQA originated in translation and localization, where quality had to be assessed across languages nobody on the requesting side could read. That constraint forced a discipline that turns out to be useful for any content quality work: findings are classified, not just fixed.
An LQA pass produces categorised findings. Typical categories include accuracy (does it convey the source meaning), terminology (are approved terms used consistently), language (grammar and mechanics), style (does it match the required register), and audience fit (can the intended reader use it). Findings are usually also weighted by severity, because a mistranslated dosage and an inconsistent capitalisation are not comparable failures.
That classification is what separates LQA from editing. An editor improves the text and hands it back. An LQA pass says the text has three accuracy findings and one critical terminology finding, in a form that can be compared across documents, vendors, and time.
The discipline transfers well beyond translation. Regulated communications, learning content, and AI-generated output all raise the same question — is this language fit for its purpose and can we show it — and all benefit from findings that aggregate.
Two constraints keep LQA honest. Categories must be mutually exclusive enough that reviewers classify the same finding the same way, or the aggregate numbers measure classification habits instead of quality. And severity must reflect consequence rather than conspicuousness, since the most visible error is frequently not the most damaging.
"The writing is good" is not something an organization can act on, verify, or hold a supplier to. LQA converts language quality into categorised, severity-weighted findings that support a decision.
It also makes quality trackable. Because findings aggregate, an organization can see whether a recurring problem is improving — impossible when review output is a series of corrected drafts.
QueryTek Review applies LQA-style structured evaluation where language quality is the subject, classifying findings against the engagement rubric so results are comparable across documents. Category sets and severity weightings are engagement artefacts.
