1. Forum
  2. >
  3. Topic: Turkish
  4. >
  5. Alphabet, Pronunciation and t…


Alphabet, Pronunciation and the TTS

The Turkish Alphabet

It consists of 29 letters.



The spelling of Turkish is phonetic, so once you learn the pronunciation of each letter, the reading should not be a problem. There are some exceptions though they are all caused by loanwords (mostly Arabic).

Turkish also uses a, i and u with the circumflex BUT the circumflex is often not used (The Turkish language association once kicked it out of the language, and tried to bring it back after a long time, but almost nobody uses them anymore, except for confusing words.) We will still give you the necessary information:

  • â for /aː/ and/or to indicate that the consonant before â is palatalized
  • î for /iː/ (no palatalization implied)
  • û for /uː/ and/or to indicate palatalization.

Here is the pronunciation guide taken from Wikipedia. Please note that the letter ğ has almost no sound, it is used to lengthen the preceding vowel. If you don’t get how to pronounce it, better skip it instead of pronouncing it as “g” :)

We will be accepting c-ç, ı-i, g-ğ, o-ö, s-ş, u-ü, not because the sounds are very close, because we know it is difficult to get used to a new keyboard. You’ll see a typo message when you don’t use the right character (although you are only allowed 1 typo normally, you are allowed to make as many “special character” typos as you want).


This was the best TTS we could find – and it is definitely not perfect. We can say that the pronunciation itself is correct most of the time, however, there are many problems with the intonation. We tried to disable several “bad” audio for the listening exercises. Unfortunately there is nothing more we can do right now.

March 23, 2015



Thanks for this. I have to say i was expecting far worse from the TTS. It seems significantly better than the one google translate uses, which is contrary to my general experience when comparing it to duolingo's in other languages... I haven't tried much of either though, perhaps it gets hairier with more complex things...


When listening to the Turkish TTS, its 'e' often sounds closer to /a/ (as in father) than to /e/ (as in red) to me.


It depends on the position of the "e."

If the "e" is followed by R, L, M or N within the same syllable, you pronounce it as the /a/ in "man", "sad" , "can" and so on. Otherwise it's the "e" in "bed", "red", "set", etc.

For example:

"Ben" (first person singular 'I') rhymes with the English word "can". It's an open "e", because it's followed by an N, one of the four consonants I've indicated above.

"Beni" (first person singular object 'me') rhymes with "Benny". The "e" is followed by an "n", that's true. But they're not in the same syllable as the syllable construction is: be-ni. Therefore the "e" is closed.


This was truly helpful, Ektoraskan. Like mizinamo, I had noticed the change in the sound of the "e" as well, but had not identified that it was consistently changing depending on the letters that followed it. Thanks to you, I'm much more likely to pay attention to things of that nature in the future.

To help me memorize the letters that influence how the Turkish "e" is pronounced, I made a pictogram of sorts and am adding it here in case it helps others. Hope you like it and thanks again!


This was bothering me, too, so thank you for the helpful explanation.

Doesn't this mean that Turkish isn't a phonetic language? I don't imagine all of the words with er, el, em, or en are loanwords!

Is it that it is really close to being a phonetic language, or are Turks in denial? :)

  • 2660

I assume you mean that it has a phonetic writing system, since all languages are phonetic inasmuch as they have phonemes. If I'm understanding correctly, Turkish has a phoneme /e/, which has a default allophone [ɛ] that occurs in most contexts, and another allophone [æ] that occurs before some consonants (what we call sonorants). Because that alteration is rule-governed, the two sounds are still one phoneme, and native speakers of Turkish wouldn't necessarily notice the difference between them. They also can't form minimal pairs, like in English between "ban" and "Ben". To a native speaker, I believe that pronouncing ben like English "Ben" instead of "ban" doesn't sound like a different word, just like pronouncing it with an accent.

Allophones occur in English without us noticing, like the /t/ in "stop" and "top". In the latter, it's pronounced with a puff of air (called aspiration), but after an /s/ in the same syllable, it's not. English speakers don't notice the difference between the [t] in [stɑp] and the [tʰ] in [tʰɑp]. Another example of allophony is in Spanish (which is usually considered to have a phonemic writing system). The voiced stops (/b/, /d/, /g/) become fricatives (more continuous sounds) after vowels or most consonants (except nasals like /m/, /n/; and /l/ in the case of /d/). So tengo "I have" is [teŋgo], while lago "lake" is [laɣo]. But to a native Spanish speaker, the difference between [g] and [ɣ] wouldn't be noticable; they're both the phoneme /g/ and they're both written with the letter "g".


Thank you for such a thoughtful response!

I guess "phonetic" was not the correct technical term.

I was referring to the fact that all the textbooks and YouTube videos I've seen have said that the letters of the Turkish alphabet are always pronounced the same way, regardless of context. But that clearly is not true (and is probably not strictly true in any language).

Are there other letters in the Turkish alphabet that follow rule-based changes in pronunciation?

  • 2660

I think the important distinction here is between phonological context, which is predictable, versus semantic context (i.e. the meaning of the word), which is not. So, to know the pronunciations of the English words "rough" and "bough", you need to have the words in your vocabulary, whereas you can determine what the "e"s in benzemek sound like without knowing the word.

To answer your question, "ğ" is also pronounced differently in different context, although I'm not a native speaker, and to be honest, I'm a little unclear on what the various contexts are.


I'm replying here because there was not option to reply to your last message.

I came across this site from another post, which gives a pretty comprehensive set of examples for each letter of the alphabet: https://www.turkishtextbook.com/pronunciation/


Ah, that makes so much more sense! Teşekkürler!


Aaaaah! That makes sense. I had wondered why "e" sometimes sounded like "a" (e.g. "öğretman") but hadn't looked at the environments that might be making that happen.

That makes sense!


Thanks. I didn't know, but I had already remarked it without understand why. Now, it is clear.


thanks a bunch! the first satisfying answer I got after a while (and asking countless Turkish people)


That trips me up too, but I hear it in real speech too, so I don't think it's the TTS. Although to me, it sounds more like the a in "apple" than in "father". The e in güzel often sounds like that to me.


Swedish is my mother tongue and we have a extra letter for this sound "ä". We pronounce "a" like in father and "ä" sounds like a in apple. When I listen to turkish "e" seems to be very similar to our "ä".


I was wondering about the pronunciation of the alphabet so this is very useful! Teşekkürler!


German and French speakers, ü is the same sound as the German ü (same letter) and French u. The almost-r listed as the Turkish pronunciation of R is made by flapping the tongue against where you make 'd' sounds; Japanese speakers, it's the same sound as what we transliterate as R (or sometimes L) from Japanese.


As a native Japanese speaker, I'd have to disagree. Turkish R sounds very different than Japanese R, especially when a word ends with an R, it almost sounds like there's "sh" sound. From my experience, Japanese R sounds a lot like Spanish R. Turkish R is somewhere between Spanish R and RR, I think.


The Japanese r is somehow a mixture of l and r; it‘s a alveolar lateral flap, as far as I know. The Turkish r as described above [ɾ] is an alveolar flap.


To be more precise, the R is like the one in American English which tends to replace T intervocalically - it is sounds like ir is.


Also, if a word ends in R, many speakers have the tendency to "hiss it". What happens is, they pronounce an American R, then stop producing a sound but keep exhaling just one second longer, which creates a faint hissing sound which actually sometimes sounds like a "sh". So if you ever hear "bir-sh", it's actually just "bir" (one). Don't get puzzled. :)


Is this hiss something we should aim to replicate? Or is it mainly a regional thing?


Nah, don't worry about it. Many people are oblivious to it. It doesn't happen all the time either.

Here's what I'm talking about:


Listen to all those recordings, you'll hear a faint "sh" sound. I bet the people who recorded those aren't even aware they do it. ;p

It's not something you should replicate or anything. I just wanted to give a heads-up so that you don't get puzzled about it at some point. :)


Ok thanks for clarifying


I was very confused because I heard that "sh" sound although I saw in other comments I shouldn't hear it. I tried to pronounce as you described "keep exhaling just one second longer" and now I definitely know where that sound comes from. Since I'm no longer confused about it and your comment made me feel like I'm improving my understanding of pronunciation, I'm giving you a lingot. ^^ Thank you. :))


another good info from u. thanks. so besides the hissing R on the end of words, the r in the middle of words are more frequently than not pronounced like American "r"? I mean for the general population


Where in the US does "r" replace "t" intervocalically? I'm used to it turning into a "d" or being dropped.

  • 2660

American English speakers don't ever think of [ɾ] as an "r" sound, but in many dialects (mine included), "latter" and "ladder" are homophones pronounced [læɾɹ].


Just as brjaga said it, and the dropping of it is called glottalization. You don't actually lose the sound, it just becomes a glottal stop.


I just finished the locative skill, and two words sounded a bit off to me: orada and burada. In both cases i feel like the first 'a' is pretty much omitted and there is a big stress on the first vowel. So I'd spell what i hear: oğrda and buğrda, pretty much.

Is that the TTS, or are these really pronounced that way? Or is it all in my head?


Orada, Burada and Şurada are usually pronounced without the a in the middle. I personally do pronounce them as my speech isn't exactly standard, but know that a great majority of the speakers just say burda, şurda and orda.


Please note that the letter ğ has almost no sound, it is used to lengthen the preceding vowel. If you don’t get how to pronounce it, better skip it instead of pronouncing it as “g” :)

I remember getting quite a surprise the first time I heard Erdoğan's name spoken on TV.


One can also listen to native speakers here: http://www.forvo.com/languages/tr/

and someone posted this link to listen to the alphabet as well: https://www.youtube.com/watch?v=6v7JH0DWXmE

Oh, here was that post I just mentioned: https://www.duolingo.com/comment/7703357


The table above lists both /ɫ/ and /l/ as possible pronounciations for the letter l. From what I understand, /ɫ/ would be closer to the hard l in Russian as in написа́л and /l/ closer to the soft Russian l as in написа́ли? (The English l in love that is given as an example seems to be somewhere in between..?) So is there a rule when to use /ɫ/ or /l/ in Turkish, or is this a matter of dialect?


I have some questions.

  1. Does TTS stand for Text To Speach ?
  2. In English we have the names of the letters, which is different in some instances to sounds that the letters make when used in words. Is this the same in Turkish. Are there names for letters, while the sounds they make in words can be different to their names.
  3. Does Turkish have an order that their alphabet appears in - in the same way that English does? If it does, how are letters with diacritics dealt with ? What is the order that affects their position when - for example - looking up a word in a dictionary ?

Thank you for assistance with these questions.


The other letters also have names - for example, the letter v does not simply have the name vvvvv (which is the sound that it makes) but, I think, ve (which is different from the sound that the letter makes because the "e" is not part of the sound, but part of the name). Similarly, be ce çe etc.


Thank you SO much ! You are brilliant !


Thank you so much for this - you have given me the passion to search - to really search - for the information I was after.
This is the closest I can find at the moment for the resource that I am after.
I will now work on making the resource I ultimately want (unless I stumble on it first that is)


First of all: Happy (0th) birthday and thanks for all your work.

One question about the TTS: Do you think "erkek" is pronounced similar to what it should be?



I agree with yalcintarkan. It's passable. I think the problem is that it's just a bit too fast and pronounced out of context.


Thanks to you too


it is not perfect but it is ok. surprisingly google translate pronounce it better : https://translate.google.com/#auto/es/erkek


Thanks for your thoughts.


Somehow I keep hearing egg-yeg


Thank you for the post. I did some lessons by curiosity, but didn't try the "special character typos". The only strange test I did was to put 1 in place of the "dotless i", which I saw somewhere in the past.

By the way, as a Portuguese speaker my keyboard haves many of those characters (I will struggle only with ı, ğ, ş, and dotted I) =]


Some time back, I modified the US-International keyboard layout using the Microsoft Keyboard Layout Creator so that I could type the additional Turkish characters. The "ı" that I type is still considered a misspelling, though. Is there a different character I could be mixing this up with?


I recommend adding the Turkish keyboard. It's not as difficult to get used to as one might think (speaking as someone who just added it today). Though if you insist on the US International, the Unicode character code for a dotless i is U+0131.


When it says ü is like the e in new, does it mean it's pronounced like the German ü? As an American, I pronounce "new" like "noo"


it is pronounced exactly like the German ü.


really helping. Tesekkurlar


Merhaba! can someone tell me what the difference in pronounciation between the these two words ending ( -eyi and eyı )


both are in accusative case


You should pronounce "i" like "i" in "split", while "ı" is like a schwa sound, that is - more like "a" in "about".

Learn Turkish in just 5 minutes a day. For free.