Skills on AI

Active theme: Light

Turkish Language

Use when asked to write, translate into, or check text in Turkish — a Latin-script agglutinative language with vowel harmony, a dotted/dotless I distinction that breaks naive text processing, and a formal/informal second person (siz/sen) split; official in Turkey and Cyprus, spoken natively by roughly 80 million people.

Turkish (Türkçe) is a Turkic language written in a Latin-based alphabet. It is the official language of Turkey and one of the official languages of Cyprus, and is spoken as a minority or heritage language in several nearby countries and diaspora communities.

Writing system

Turkish uses a Latin-based alphabet with several letters not found in English: ç, ğ, ı, ö, ş, ü. The most important detail for text processing is that Turkish has two distinct pairs of I/i letters: the dotted İ/i and the dotless I/ı are genuinely different letters with different sounds, not a capitalization variant of a single English-style I/i pair. Dotted i uppercases to İ, and dotless ı uppercases to I — software that assumes the standard English I/i case-folding rule (where "i".toUpperCase() == "I") will silently corrupt Turkish text, mangling words, breaking case-insensitive comparisons, and producing wrong sorting and search results. This is a well-known, recurring source of bugs (sometimes called "the Turkish I problem") in software that doesn't use locale-aware case conversion.

Agglutinative morphology

Turkish is strongly agglutinative: grammatical meaning (case, possession, plurality, tense, negation, question marking, and more) is built by stacking a sequence of suffixes onto a word stem, each suffix contributing one clear piece of meaning. This produces long words that correspond to entire phrases or clauses in English — a frequently cited example is "Çekoslovakyalılaştıramadıklarımızdanmışsınız" (roughly, "you are apparently one of those we could not turn into a Czechoslovakian"), an extreme but real illustration of how much a single word can carry. Translating into Turkish often means collapsing an English phrase into one long word rather than a sequence of separate words, and word-by-word alignment with English breaks down.

Vowel harmony

Suffix vowels change to match qualities of the vowel that precedes them, following consistent, rule-governed patterns rather than being arbitrary or memorized case by case. Turkish vowels split along two main axes — front/back and rounded/unrounded — and most suffixes have multiple forms (typically two or four variants) selected automatically based on the last vowel of the stem they attach to. For example, the plural suffix appears as -lar or -ler depending on the preceding vowel. Getting harmony wrong produces a suffix form that sounds and reads as clearly foreign or broken to a native speaker, even when the grammatical category chosen is otherwise correct.

Formality: siz vs. sen

Turkish marks formality through the second-person pronoun: sen is informal/singular, used with family, friends, children, and peers; siz is used both as the formal singular (addressing a stranger, elder, or superior respectfully) and as the standard plural "you" regardless of formality. As with Russian's вы and similar systems, siz's dual role means context determines whether a formal-singular or a plural reading is meant. The choice also determines verb suffix forms, not just the pronoun. Defaulting to siz is the safer choice when addressing someone unfamiliar or in a professional context.

Common pitfalls

  • Applying English-style case folding to i/I, corrupting dotless ı/I and dotted i/İ — use locale-aware (Turkish-specific) case conversion in any code that processes Turkish text.
  • Getting vowel harmony wrong when generating or editing suffixed forms, especially in text produced by naive rule-based or templated generation rather than by a fluent speaker.
  • Translating word-for-word from English phrase structure instead of collapsing meaning into Turkish's characteristic long, suffixed words.
  • Defaulting to sen in a formal or unfamiliar context where siz is expected, or missing that siz addressed to one person still takes the same form as siz addressed to several.
  • Missing the extra Latin letters (ç, ğ, ı, ö, ş, ü) in fonts, sorting routines, or input validation that were built assuming a plain 26-letter English alphabet.

Learn more

  • Cross Cultural Communication for formality and register considerations beyond grammar.
  • Locale for the Turkish-specific locale case-folding behavior referenced above, and other regional formatting conventions.
  • German Language for a contrasting European Latin-script language with a case system rather than agglutination.

View turkish-language/SKILL.md on GitHub