Spelling and scripts

Karachay is written in Cyrillic. Every word here is stored in Cyrillic, and the Latin spelling you may be reading is derived from it.

One direction only

The Cyrillic spelling is the word. The engine reads it, builds the forms in Cyrillic, and renders each form to Cyrillic without loss and to Latin as display. There is no Latin-to-Cyrillic conversion anywhere on the site: it was written once, it guessed, and it was removed. So a form you see in Latin is a rendering of a Cyrillic form the dictionary holds, and never the other way round.

We write дж, not ж: джол, джангы, джаз. Both spellings are in print, and the ж spelling belongs to the Balkar standard and to some Soviet-era editions. Where a source spells a word with ж, the entry keeps our дж spelling and the source is cited as it stands.

The alphabet

The Latin scheme is one letter for one sound. Digraphs on the Cyrillic side — къ, гъ, нг, дж — are single letters here, which is what lets the engine apply the sound rules without guessing where a letter begins.

• ng and ŋ. The nasal нг is one sound. The site writes it ng; the scholarly display scripts and the engine's own internal code for it are both ŋ. The doubled spelling ннг keeps both letters: меннге is mennge, сеннге is sennge.
• İ and ı. The dotless ı and the dotted i are different letters, as in Turkish, and case is handled in the Turkish locale throughout: Ыйыкъ comes back as Iyıq, never İyıq. Sorting uses the same collation, which is why ç, ğ, ı, ö, ş and ü each have their own place in the alphabet and their own page under A–Z.
• One Cyrillic letter, two Latin ones. у is the vowel u and the consonant w, and ю is ü and yu, decided by the letters around them. The engine reads the Cyrillic, so the Latin form always says which.
• э and е. The same sound is written э at the start of a word and е inside it, so the renderer places them by position rather than by a letter map.
• Capitals. Proper nouns are capitalized in both scripts — Къарачай, Ислам, Ингуш. URL slugs stay lowercase.

Other Latin spellings

Three scholars' own romanizations — Seegmiller 1996, Nevruz 1991, Sipos & Tavkul 2015 — are offered as display scripts from the gear in the corner, so a reader working from one of those books can read the site in its spelling. Each is a letter map out of the canonical form; they are named after the systems, not attributed to the scholars as our writing.

Pronunciation

The line under a head-word, between slashes, is a broad phonemic transcription. It is derived from the spelling, letter by letter, by the engine: the canonical scheme is one letter per sound, so no guessing is involved and no exception is recorded. It is not a recording, not narrow, and does not show stress or how a particular speaker says the word. Recordings are a separate piece of work.

Karachay and Balkar

The dictionary lists Karachay. A Balkar-only form is left out of the Karachay lists; where one is shown, it carries a Balkar label, and the label is what tells you not to use it as the Karachay word. The conventions page explains the rest of the labels.