Most Turkish letters are easy for English speakers: each one has a single, consistent sound. Three cause almost all the trouble: the dotless ı, its partner dotted i, and the soft g (ğ). Mixing up ı and i changes words (dış “outside” vs diş “tooth”), and pronouncing ğ as a hard g is one of the most noticeable foreign accents in Turkish.
This guide covers how to make each sound, the minimal pairs to practice, the capitalisation rule that famously breaks software, and what ğ actually does. To hear every example, use the alphabet soundboard as you read.
Dotted i: the easy one
The Turkish i is the vowel in English see or machine, only a little shorter. It never sounds like the i in bike or bit.
| Turkish | Sounds like | English |
|---|---|---|
| bir | beer (short) | one |
| iki | ee-kee | two |
| için | ee-cheen | for |
| bilet | bee-let | ticket |
| kim | keem | who |
Dotless ı: the sound English doesn’t have
ı is a high vowel made with the tongue pulled back and the lips spread, not rounded. In phonetic terms it’s [ɯ]. English has nothing quite the same, but the closest sounds are the unstressed vowel in roses or the a in about.
How to find it:
- Say “oo” as in food. Notice your lips are rounded.
- Keep your tongue exactly where it is and spread your lips into a flat smile.
- The sound that comes out is ı.
| Turkish | Approximate | English |
|---|---|---|
| kız | kuhz | girl, daughter |
| ışık | uh-shuhk | light |
| kapı | kah-puh | door |
| balık | bah-luhk | fish |
| altı | ahl-tuh | six |
| sıcak | suh-jahk | hot |
Minimal pairs: ı vs i
These words differ only in that one vowel. Say each pair aloud until you can hear and make the difference:
| With ı | English | With i | English |
|---|---|---|---|
| dış | outside | diş | tooth |
| kır | countryside | kir | dirt |
| kıl | (body) hair | kil | clay |
| sınır | border | sinir | nerve |
| ılık | lukewarm | ilik | bone marrow |
Where ı fits in the vowel system
ı isn’t an odd extra letter. It fills a gap in a very tidy system. Turkish has eight vowels, sorted by three features: front or back (tongue position), rounded or unrounded (lips), and high or low (jaw):
| Unrounded, low | Unrounded, high | Rounded, low | Rounded, high | |
|---|---|---|---|---|
| Back | a | ı | o | u |
| Front | e | i | ö | ü |
Read across the high unrounded column: ı is simply the back partner of i, just as u is the back partner of ü. If you can already say i and u, you have both ingredients: ı uses the tongue position of u with the flat lips of i.
Why it matters for grammar
The ı/i difference isn’t just pronunciation: it drives vowel harmony. Four-way suffixes take ı after a and ı, but i after e and i:
- kız → kızım (my daughter), kızın (your daughter)
- diş → dişim (my tooth), dişin (your tooth)
If you hear or write the root vowel wrong, every suffix after it comes out wrong too.
Capital letters: İ and I
English has one i with two shapes (i, I). Turkish has two separate letters, each with its own capital:
| Lowercase | Uppercase | Example |
|---|---|---|
| i | İ | istanbul → İSTANBUL, izmir → İZMİR |
| ı | I | ılık → ILIK, ırmak (river) → IRMAK |
So the capital of i keeps its dot (İ), and the plain capital I belongs to dotless ı. That’s why the city is written İstanbul, and why Turkish signs show İZMİR, never IZMIR. Proper names starting with dotless ı keep plain I: the province of Iğdır is pronounced with ı.
The “Turkish I” problem in software
This rule causes one of the best-known bugs in internationalisation. In several programming languages, including Java and C#, the default case conversion follows the locale (language settings) of the computer. Under Turkish rules:
- uppercasing i gives İ, not I;
- lowercasing I gives ı, not i.
So Java code that checks whether "FILE".toLowerCase() equals "file" fails on a computer set to Turkish, because the result is “fıle”. In 2008, developer Jeff Moser called this “the Turkey Test”, and Microsoft’s own guidance warns about it. Typical symptoms are keywords that aren’t recognised, settings that don’t load and case-insensitive searches that miss matches, but only on machines set to Turkish.
There’s a mirror-image problem too. Under the default (non-Turkish) Unicode rules, lowercasing İ produces i followed by a combining dot, which can look odd or fail to match. Apps that capitalise names or headings without Turkish rules also produce mistakes like “ISTANBUL” for İstanbul.
For developers: use ordinal or culture-invariant comparisons for identifiers and internal keys, and use explicit Turkish rules (for example toLocaleUpperCase('tr') in JavaScript) for text that people read.
For learners: if an app or website turns your İ into I or adds strange dots, that’s a software bug, not a spelling rule. And when you type Turkish, make sure your keyboard really produces ı and İ. Our guide to typing Turkish characters shows how.
Soft g (ğ): the letter that’s almost never heard
ğ (yumuşak ge, “soft g”) is the strangest letter in the alphabet: it’s usually not pronounced as a consonant at all. In modern standard Turkish it mostly affects the vowel before it. Three rules cover nearly every case:
1. It never starts a word. No Turkish word begins with ğ, so you’ll only ever meet it after a vowel.
2. After a back vowel (a, ı, o, u), it lengthens the vowel.
| Turkish | Sounds like | English |
|---|---|---|
| dağ | daa | mountain |
| yağmur | yaa-moor | rain |
| ağır | aa-uhr | heavy |
| soğuk | so-ook | cold |
| doğru | doh-roo (long o) | true, correct |
| uğurlar olsun | oo-oor-lar ol-soon | farewell, safe travels |
3. Next to a front vowel (e, i, ö, ü), it often sounds like a light y or simply lengthens the vowel.
| Turkish | Sounds like | English |
|---|---|---|
| değil | dey-il | not |
| eğer | ey-er | if |
| öğretmen | öy-ret-men / öö-ret-men | teacher |
| iğne | ii-ne | needle |
| düğün | dü-ün | wedding |
The most common mistakes with ğ
- Saying a hard g: dag, yagmur, degil. This is the most noticeable error. Treat ğ as “make the vowel longer” instead.
- Saying a throaty “gh”: ğ developed from a consonant like that, and some regional accents keep traces of it, but standard Istanbul Turkish doesn’t use it.
- Adding a full y everywhere: değil is a light glide, not a strong “deh-YEEL”.
Where ğ comes from in grammar
You’ll often meet ğ in the middle of words because of consonant mutation: many words ending in k after a vowel change it to ğ before a suffix that starts with a vowel.
- çocuk (child) → çocuğu (his/her child) → said “cho-joo-oo”
- ekmek (bread) → ekmeği (the bread, as object) → said “ek-mey-i”
- ayak (foot) → ayağım (my foot) → said “a-yaa-uhm”
A 5-minute daily drill
- Say the five ı/i minimal pairs twice, slowly, then at normal speed.
- Read the six ğ words in rule 2, stretching each vowel.
- Record yourself saying Bu doğru değil (“This isn’t true”) and Kızım dişçiye gitti (“My daughter went to the dentist”) and compare with a native recording.
Quick check
- Which word means “tooth”: dış or diş?
- Write ılık and istanbul in capital letters.
- How is dağ pronounced: “dag”, “dagh” or “daa”?
- Add “my” (-ım/-im/-um/-üm) to kız and to diş.
- What happens to the k in çocuk when you add -u?
Answers: 1. diş (dış means “outside”). 2. ILIK, İSTANBUL. 3. “daa”: ğ lengthens the a. 4. kızım, dişim. 5. It becomes ğ: çocuğu, said roughly “cho-joo-oo”.
Practice this
Listen to every letter in the alphabet soundboard, then check your own pronunciation with the pronunciation trainer. For the full sound system, read the Turkish alphabet and pronunciation guide.