Turkish/Azeri Characters Alt Codes & Unicode Cheatsheet

Turkish adds six letters to the Latin alphabet, and one of them creates a software bug so well known it has its own name. Turkish distinguishes two i letters: the dotted pair i and İ, and the dotless pair ı and I. These are four separate letters, not two letters with optional dots. The lowercase of I is ı, and the uppercase of i is İ. That single fact breaks a startling amount of software.

The problem, usually called the Turkish-I problem, appears whenever code uppercases or lowercases a string using the current locale. Under a Turkish locale, uppercasing the ASCII letter i produces İ rather than I, so a case-insensitive comparison of file, FILE or an HTML tag name silently fails. This has broken configuration parsers, login systems and file handling in real products for decades. The fix is to use an invariant or ordinal comparison for anything machine-facing and reserve locale-aware casing for text shown to humans. If a system misbehaves only for Turkish users, this is the first thing to check.

The other letters are more ordinary but still meaningful. Ç and ş are consonants with their own sounds, ch and sh respectively. Ö and ü are front vowels shared with German. The soft g, ğ, is the unusual one: it is never pronounced as a distinct consonant in modern standard Turkish and instead lengthens the preceding vowel or acts as a glide, which is why it never begins a word — there is no Turkish word starting with ğ.

Turkish vowel harmony makes the vowel letters structurally important rather than decorative. Suffixes change their vowels to match the stem, so the plural is -ler after front vowels and -lar after back vowels, and the dotted-versus-dotless i distinction is part of that system. Replacing ı with i does not merely misspell a word; it puts it in the wrong harmony class and makes every suffix that follows look wrong to a Turkish reader.

A note on ş: Turkish ş takes a true cedilla, and Romanian ș takes a comma below. They look similar, they are different characters, and text that mixes them will not match. If you are handling data from both languages, normalise deliberately rather than assuming they are the same letter. On Windows, hold Alt with Num Lock on: ç is Alt+0231, ö Alt+0246, ü Alt+0252, ş Alt+0351, ğ Alt+0287, ı Alt+0305 and İ Alt+0304. On macOS, select the Turkish input source, or use Option+C for ç and Option+U then the vowel for the umlauts. The Turkish Q layout is the common choice and keeps QWERTY largely intact while adding the six letters around the edges.