False Fluency Principle
Knowing a language is not the same as understanding one.
False fluency is the mechanism by which translational injustice occurs. A fluent sentence can still be a failed act of understanding, and institutions keep treating that failure as if it were neutral.
False fluency assumes that language knowledge is language understanding.
That assumption is false. Knowing words, grammar, or a plausible equivalent does not prove that the speaker, interpreter, machine, student, or institution has understood what is being communicated.
False fluency is therefore not only a principle. It is the mechanism that turns a gap in understanding into a translation, record, decision, lesson, or system output that looks complete.
The principle comes from the fourth fieldwork interview: a professional interpreter explained that interpreting means understanding two languages. Knowing them is only the start.
How the gap is made.
False fluency is created by smaller mechanisms that make knowing look like understanding. Bias is the first developed mechanism; the others are mapped here while their pages are being built.
Translation choices are shaped by ideals, context and stereotypes: what is prioritised, omitted, edited, domesticated, or normalised.
02 / In development Lack of metacognitive awarenessThe belief that one understands can hide the limits of that understanding, especially in institutional or educational settings.
03 / In development Knowledge deficitsMissing domain, cultural, contextual, or source-language knowledge can produce fluent but shallow interpretation.
False fluency is attractive because it lowers institutional friction.
It lets an institution say that language has been dealt with: an interpreter was booked, a machine returned output, a student read a translation, a system processed text.
But cost is not the same as competence. A freedom of information response showed nearly £400,000 spent on employment tribunal interpreters. The problem is not whether interpreters are hired. It is whether the system monitors whether the right person can do the job.
The machine can be fluent without being personal.
Machine translation can look clean and still remove tone, register, dialect, hesitation, idiom, and human connection. Refugee fieldwork already shows this: the words can be accurate while the person disappears.
AI makes the principle sharper. Whether AI can understand at all remains debated, so systems such as ACS should be treated with care when they are used around testimony, vulnerability, credibility, or legal records.
The False Fluency Principle extends the whole framework.
The record can sound coherent while the person has been mistranslated.
Case File 02 Language technologyThe machine can return words without understanding the person using them.
Case File 03 AI and low-resource languagesUnconscious systems should not be treated as if they understand high-stakes language.
Case File 04 Classics and educationStudents can become falsely fluent in a translation while missing the source text.