ESL Writing & AI Detection Glossary

The vocabulary of writing in English as a second language — and of the AI detectors that judge it. Each definition is written to stand alone: quote it, cite it, send it to your professor.

• Flagxiety

  Flagxiety is the fear of having your own writing falsely flagged as AI-generated by detection tools such as Turnitin or GPTZero.

• Perplexity (AI detection)

  In AI detection, perplexity measures how predictable a text's word choices are to a language model.

• Burstiness (AI detection)

  In AI detection, burstiness measures how much sentence length and structure vary across a text.

• False cognate

  A false cognate (often called a false friend) is a word that looks similar in two languages but means something different in each — Spanish actualmente means "currently," not "actually." The resulting sentence is usually grammatical, which is why grammar checkers miss the error..

• L1 interference

  L1 interference, or negative language transfer, occurs when the grammar, vocabulary, or rhythm of your first language shapes what you write in a second language — dropped articles, word-for-word phrases, imported word meanings.

• Calque

  A calque is a phrase translated word-for-word from another language into English, such as "make a photo" instead of "take a photo." Every word is correct on its own; the combination is borrowed from the writer's first language..

• Code-switching

  Code-switching is the practice of alternating between two or more languages within a single conversation, sentence, or text.

• Register mismatch

  A register mismatch is using vocabulary or phrasing whose formality level does not fit the context — casual wording in an academic paper, or stacked formal words ("utilize, moreover, plethora") in an everyday email.

• Translation ghost

  A translation ghost is a sentence that is grammatically correct English but whose word order and emphasis still follow the writer's first language — the source language shows through like a ghost image.

• Hedging (academic writing)

  Hedging is the use of softening language — may, might, suggest, appear, tend to — to match the strength of a claim to the strength of the evidence behind it.

• Passive voice transfer

  Passive voice transfer is an L1-interference pattern in which the prestige of passive and impersonal constructions in a writer's first academic language carries into English, producing far more passive voice than current English academic style expects..

• Authorship certificate

  An authorship certificate is a cryptographically signed record of how a document was written — its timestamped edit and revision history — that lets a writer prove a text is their own work rather than pasted AI output..

• Diglot Weave Method

  The Diglot Weave Method is a language-learning technique in which foreign-language words are woven into text written in the learner's native language, so vocabulary is absorbed in context rather than memorized from lists.

• Vibe writing

  Vibe writing is the writing analog of vibe coding: you supply the intent and the meaning, an AI shapes the prose, and your job shifts to steering, judging, and verifying.

• Translationese

  Translationese is the statistical and stylistic fingerprint that translated text carries: word-for-word calques, over-explicit phrasing, and a flatter, more uniform rhythm than text composed directly in the target language.

• Lexical gap

  A lexical gap is a concept that one language names with a single word and another language does not: Russian сутки (a 24-hour period), Portuguese saudade, German Termin.

• Patchwriting

  Patchwriting is paraphrase that stays too close to its source: synonyms are swapped in and phrases lightly reshuffled, but the original sentence skeleton remains.

• Tortured phrases

  Tortured phrases are absurd synonym-spun rewordings of fixed technical terms — «counterfeit consciousness» for «artificial intelligence» — left behind when text is run through automatic paraphrasing tools.

• Stylometry

  Stylometry is the statistical study of writing style — function-word frequencies, sentence-length rhythm, punctuation habits — used to identify or verify the author of a text.

• Exophony

  Exophony is the practice of writing in a language that is not your mother tongue.

• Similarity score

  A similarity score is the percentage of a document that matches text in a plagiarism checker's database — Turnitin's is the best-known.

• Foreign language anxiety

  Foreign language anxiety is the fear and tension tied specifically to performing in a language you are still learning — a distinct research construct identified by Elaine Horwitz and colleagues in 1986.

• Cognitive debt

  Cognitive debt is the cost of outsourcing thinking to AI: the task gets done, but the understanding, memory, and skill that doing it would have built are missing when you need them later.

• Runglish

  Runglish is the informal name for Russian-influenced English: dropped articles (Russian has none), word-for-word calques like "make a photo," and false friends such as магазин (a shop) shadowing magazine.