Tuesday, August 31, 2004


Looking at this list, you'll see both the strong connection between Turkish and the other two languages. More importantly, you'll see just how close together Uighur and Uzbek are. As is noted in the introduction, it really can be hard to tell whether they're separate languages or variants of one tongue.

English / Turkish / Uzbek / Uighur

black / kara / qora / qara
white / ak / oq / aq
red / kızıl / qizil / kizil
blue / mavi / ko’q / kök
green / yeşil / yashil / yeshil
yellow / sarı / sariq / sarik

A few "opposites"

(I / you, big / small, left / right, dog / cat, etc.)

English / Turkish / Uzbek / Uighur

I / ben / men / mæn
you / siz / siz / siz

big / büyük / katta / chong
small / küçük / kichkina / kichik

hot / sıcak / issiq / issiq
cold / soğuk / sovuq / soghoq

left sol / chap / sol
right sağ / o’ng / ong

north kuzey / shimol / shimal
south güney / janub / jänup

east doğu / sharq / shærq
west batı / ga’rb / ghærp

dog / it / it / it
cat / kedi / mushuk / müshük


This article gives you the phrases for a few basic civilities and, in the notes, a chance to see how Turkic, Persian and Arabic interact

English / Turkish / Uzbek / Uighur

Hello / Merhaba / Assalomu alaykum / æssalamu ælæykum
Goodbye / Iyi günler / Hair / Her xosh

How are you? / Nasılsınız? / Qandaysiz? / Qandaq ahvalingiz?
(I’m) fine / Iyiyim / Yaxshi / Yakhshi

Please / Lütfen / Marhamat / Mærhæmat
Thank you / Teşekkürler / Rahmat / Ræhmæt

What follows is a look at what makes up these vocabulary items, with an emphasis on the interplay between Turkic and Persian and the presence of Arabic loanwords

Ar=Arabic ; fr.=from ; Pers.=Persian ; T=Turkish ; Uz=Uzbek ; Ui=Uighur

T merhaba fr. Ar. “Welcome” (merhaba)
Uz/Ui “Good Day” fr. Ar. “Peace be with you.” (as-salaamu aleykum)

T iyi means “good” ; gün means “day” ; iyi günler = “good days”
Uz/Ui hair/her like T “blessing” (hayır)
Ui khosh like T “happiness” (hoş), taken into Pers. as khosh

T nasıl / Uz qanday / Ui qandaq = “how”
T siniz / Uz siz = “you”
T nasılsınız / Uz qandaysiz literally mean “how are you”
Ui ahval, like T ahval = “circumstances” ; ingiz = “your”
Ui qandaq ahvalingiz = “how are your circumstances”
T iyi / Uz yaxshi / Ui yakhshi = “good” ; the “yim” means “I am”

T lütfen fr. lütüf (“favor”) ; Persian and Dari, among others, have adopted the word
T teşekkürler probably fr. Ar. shokran (“thanks”) ; also adopted by Persian (with French derived merci)
Uz/Ui “Please” fr. Ar. rahmat (“compassion”) from root word meaning “mercy”
Uz/Ui “Thank you” fr. Ar. marhamat (“act of compassion”) ; same root as rahmat, namely r-h-m sequence as in rahman – merciful (reference to God and common Arabic last name)

A word on spelling in the vocabulary lists

In English, we’ve been make dictionaries and spelling guides for a few centuries. In Turkish, they’ve been making them since Attaturk. As a result, the spelling for Turkish, like the spelling in English, is pretty well fixed.

Uzbek and Uighur have been written with Arabic letters (adapted from the Persian usage), with Cyrillic (when the Soviet Union occupied major stretches of their homelands) and with Roman letters (in Uighur’s case, based on Chinese pin-yin’s use of that alphabet).

When you’re writing in English and Turkish, you consult a dictionary for spelling. In Uzbek and Uighur, things aren’t so clear. Spelling reforms are underway and it may even be that official spellings exist in the eyes of different governments and other institutions. But companies like Lonely Planet and Hippocrene and amateur web publishers have a way of offering their understandings of these (and many other) languages while government presses sit idle or turn out volumes that are too expensive to gain widespread use and set the standard, even among the people they ostensibly govern.

At multilingua.info, we had the choice, therefore, of creating our own spelling conventions or deciding whom to follow. For Uzbek, we are for the most part following the Hippocrene Uzbek Dictionary and Phrasebook by Awde et al. For Uighur, we are following the Lonely Planet Central Asian Phrasebook by Rudelson, with the following exception:

For Uighur, we are writing the front “a” as “æ” and leaving the back “a” as “a”. We like the idea of using the umlaut for all front vowels, but remark that “ä” carries a variety of different sounds in different Germanic languages (especially “eh” in German), and preferred the a-e ligature for the sound of “a” in “cat”; “ö” and “ü” are roughly the same in German and in Rudelson’s transcription system, so we have left them alone.

Regardless of source, we will do our best to stick to the spelling conventions mentioned here, adapting material as necessary.

Therefore, a word to the wise: whether you follow our conventions or someone else’s, for Uzbek and Uighur you’re going to have to be conscious of different ways of representing the same word depending on where you’re reading.

Monday, August 30, 2004

Vowel Harmony II – What is vowel harmony?

The problem with agglutinating languages is they don’t do a very good job of helping people talk quickly and easily. With all those particles to tack on, it’s hard to find a convenient way of slurring through the whole mess so that you don’t sound like a robot talking. Or at least it would be, were it not for the marvel of “vowel harmony.” Try saying this three times fast:

“I saw ye all last week.”

Did you like that “saw ye all” combination? If you’re from the Southern U.S., I’ll bet you said “y’all” without even thinking about it.

What’s wrong with “saw ye all”? Well, the “aw” of “saw” is way in the back of your mouth. You can practically feel it resonate in your throat when you say it. The “e” in “ye,” on the other hand, is at the front of the mouth. If you just say “ye,” your nose might even tickle a little because the sound resonates right underneath it. When you say “all,” you’re at the back of the mouth again. This is hard work – your breath has to go scurrying all over the place to resonate in one spot, then the next, then back again. In fact, that whole sentence –

I saw ye all last week –

goes front-back-front-back-back-front. Which is why, speaking quickly, you might say something like, “ah saw y’all last way-uck.”

Sentences like “I saw ye all last week” are to vowels what tongue twisters are to consonants. Call them breath-busters. Thing is, if you speak a language like English that sort of marches along, they’re not fun and you sound funny if you try to say them fast but they’re manageable. But if you speak a language chock full of little words and wish to ease through them, something has to be done so that speech will be fast but distinct.

Vowel harmony languages do their best to keep their particles/endings/suffixes clear and distinct while recognizing the need for easier speech. They do this with vowel harmony – regular rules for how and where to slur – modify – vowels to suit the phonetic environment.

We saw above the difficulties involved in saying “saw ye all” really fast. But “saw ya” and “saw y’all” are much easier to say. In this case, you’ve got all back vowels – vowels that resonate nearer the throat. Let’s create a system. Instead of saying “you,” we’re going to say “ye” if the vowel from the syllable before is a front vowel (resonates in the mouth just below the nose) and “ya” in the case of preceding back vowels. Here are some questions:

Saw ya that? / See ye there?
Spoke ya well? / Speak-ye kindly.
What have-ya there? / Were-ye happy?

These questions follow a sort of vowel harmony. Don’t believe me? Say “saw-ya that,” then “saw-ye that.” Betcha had to pause between “saw” and “ye”. Try flipping things with all six questions. You’ll notice that you have to pause to relocate your breath in each case, which adds time, if only half a second, to each and every utterance. Over time, this adds up.

In English, as I’ve hinted, we have a sort of vowel harmony. “I” becomes “ah” (or “ayh” if the next word has a back vowel. “You” shifts between “ya,” “you” and “yuh.” But this is all very informal and in a limited number of phonetic environments (actually, English has “consonant consonance” in greater measure – we rearrange consonants to make them easier to pronounce together, as with “dontcha,” “betcha,” “dunno,” “gonna,” etc.).

People who speak vowel harmony languages are as attuned to the problem of getting through breath-busters (rapid shifting between front and back vowels) as we are to negotiating consonant clusters. If you’re going to become a speaker of a language with vowel harmony (Finno-Ugric – eg Hungarian, Finnish, Lapp; and Altaic – Turkic and Mongolian are the most common), you’ve got something new and different, but doable, to learn. In our example, we just shifted between “ye” and “ya.” In vowel harmony languages, just about every ending has two forms like this, with rules for which one to pick. And at first it will be tricky. But pretty soon you’ll feel the difference between front and back (redo the examples if you’re still unclear on this) and you’ll start making the distinction naturally.

One word about actually learning to use vowel harmony: The suggestions in this primer draw on English to make the concept clearer to English speakers. For actually learning a vowel harmony language, the first thing you’ll need to do is to consult a decent grammar to find out which vowels change in which ways and when. Using what I’ve explained, you should be able to make some real sense out of the tables, physically sensing where the distinctions are among sounds once they’ve been shown to you, but getting to where you can do this intuitively requires practice, more than anything else, once you’ve figured out the basic idea behind what you’re doing and why.

Vowel Harmony I – The Trouble with Endings

When creating a language, people (or should I say peoples) often have a really bright idea. To keep it simple, they think, I’ll come up with some little words that clarify what I’m talking about. We call these words particles, and some languages use them with reckless abandon.

The particle plan usually works so well that quicker ways of saying them are created to make speech even better. If you’ve ever said, “See ya later” instead of “See you later” (and you know you have!), you’ll understand where these people are coming from.

Pretty soon, just as surely as we say “can’t” because “cannot” takes too long, particles get run into the words before or after them to make things quicker. If you’ve ever said “see ya” as a two-syllable word (“see-ya!”) instead of as two distinct words (you probably have!), you know what drives these people.

At this point, one of two things can happen. 1) The grammar police can show up and drill into young minds that whether you say “see-ya” or not, you have to write “see you” as separate words. 2) Things can run together so well that by the time the grammar police get there, it’s to drill into young minds that even if it looks like it came from “see you,” you have to write “see-ya” (and not “see-yu,” “see-yuh” or “se-yuh”).

When particles get tacked on the ends of other words, we like to call them endings. (Some people prefer “suffix,” but that’s a bit highfalutin.) Endings are bad news. The problem is, people don’t like to take a long time to say them. So they slur over them.

Once an ending gets slurred, there are two things that can happen. 1) The grammar police can insist it be enunciated correctly. 2) The grammar police can come up with rules for the proper way to slur it.

In Latin, if you wanted to say “I will love,” you’d say amabo. This was ama (love) + b (future marker) + o (“I” form marker). People thought that was too hard, so they said, “amare habeo” (ama – love + re – infinitive marker; habe – have + o – “I” form marker).

By the time the French got done with “amare habeo”, they were saying “aimerai” (aim – love + er – infinitive marker + ai = “I have”). More precisely, they were saying, “j’aimerai” – I will love. Now that they’ve been doing that 1000 years or so, some people think the rule is too rigid or confusing. They just say, “je vais aimer” (I am going to love). One day grammar teachers will designate the proper way to run together what is pronounced “zh’vay zemmay” in rapid speech.

In the Romance languages (and most Indo-European languages), when the grammar police pick endings for verbs, it’s called “conjugation” (Latin for “tie together”). With nouns, it’s called “declension” (like “decline,” because you have to read down the list of endings – I’m not joking on this point!). So basically, by their own admission, Romance language speakers handle endings by sticking stuff together and memorizing the lists. The Turkic languages do it differently. Their grammar police had a different idea. The idea was “vowel harmony.”

What is vowel harmony?

Click here to find out.

Agglutination in Turkic with Uzbek examples

Look at this Uzbek conversation:

Men tushunyapmanmi? Yöq! Tushunmayapman.

Pretty scary, huh? Let’s put in some hyphens.

Men tushun-yap-man-mi? Yöq! Tushun-ma-yap-man.

Here’s a little glossary:

ma - not
man – “me” ending of the verb.
men – I
mi – question marker
siz – you
tushun – understand
yap – -ing
yöq - no

Here’s our conversation with the secret decoder ring read-out after:

Tushun-yap-siz-mi? (understand-ing-you-do-?)
Men tushun-yap-man-mi? (I understand-ing-I-do-?)
Yöq! Tushun-ma-yap-man. (no! understand-not-ing-I!)

As you can see, those weird long words are not completely (just largely) incomprehensible. If you know to look for the little pieces, you can start to follow what’s going on. Let’s try one more. Here is the conversation:

Ruscha bilasizmi?
Ruscha bilamanmi? Yöq! Ruscha bilmayman.
Inglizcha bilasizmi?
Ha, inglizcha bilaman.

Now in digestible pieces:

Rus-cha bila-siz-mi?
Rus-cha bila-man-mi? Yöq! Rus-cha bil-may-man.
Ingliz-cha bila-siz-mi?
Ha, ingliz-cha bila-man.

The glossary of new words:
bil(a) – to speak/to know a language
cha – language
ha – yes
may = maingliz – English

Now with the secret decoder ring read-out:

Rus-cha bila-siz-mi? (Russian-language speak-you-?)
Rus-cha bila-man-mi? (Russian-language speak-I-?)
Yöq! Rus-cha bil-may-man. (no! Russian-language speak-not-I.)
Ingliz-cha bila-siz-mi? (English-language speak-you-?)
Ha, ingliz-cha bila-man. (yes, English-language speak-I.)

Incidentally, throwing in our “yap” particle would give, for example:
Ha, inglizcha bilayapman. = Yes, I am speaking English, which is not the same as “I do speak English” (in response to “Do you speak English?”).

Got all this now? Ready to go to Uzbekistan and negotiate a major business deal? Maybe not. But you should have some idea how agglutination works and what it looks like.

Of course, tacking regular particles/endings together to create large but logical words looks like a great way to run a language. Why don’t we all do it? There are complications…

Click here for the trouble with endings, followed by what the Turks have done about it.

Agglutination – What’s it all about?

Let’s make up a language. With our language, instead of using a whole bunch of little words all over the place, we’re going to tack a bunch of things on to the main words. Ready?

The man sees your mother.

Man-the mother-your sees.

That “s” on “sees” stands for “he/she/it,” by the way. So, just to make things clear…

Man-the mother-your see-s.

Let’s make it in the past.

Man-the mother-your saw.

Oops! Where’s the “s” in “saw”?

Man-the mother-your saw-s.

That doesn’t look very good either. Let’s make “see” a regular verb.

Man-the mother-your see-s-ed.

It looks funny, but now we’ve got something we can work with. Try to change this sentence:

The horse heard my mother.

How’s this?

Horse-the mother-my hear-s-ed.

Let’s go back to our first sentence:

Man-the mother-your see-s-ed.

Let’s expand it, mentioning that this happened in Edinburgh.

Man-the mother-your Edinburgh-in see-s-ed.

Suppose the horse saw my mother in London…

Horse-the mother-my London-in hear-s-ed.

Now, let’s suppose the man was seeing your mother when the horse heard my mother. We’re going to add an “ing” in “see” to indicate an ongoing action (like “was seeing” or “is seeing”).

Man-the mother-your see-s-ing-ed Edinburgh-in when horse-the mother-my hear-s-ed.

This may look like gibberish, but what you’re looking at is “agglutination,” which boils down to “gluing” a bunch of piece onto a word to specify its meaning in context. Let’s look at our sentence again.

Manthe motheryour seesinged Edinburghin when horsethe mothermy hearsed.


Would you believe this is how Turkic languages work? It is, sort of.

On the next page, we’re going to do an exercise like this with Uzbek.

Click for the next page.

What's so tricky about Turkic vocabulary?

What is so tricky about Turkic vocabulary?
A lot of it is Persian and/or Arabic.

Speakers of Persian dialects have been in competition with Turkic speakers for dominance in the region for quite some time. Neither has achieved truly lasting dominance the way that, say, the Romance languages dominate the South of Europe. A quick look around the region turns up Persian in Iran, Dari (almost identical) and Pashto (same family) in Afghanistan and Tajik (almost identical to Persian) in Tajikastan, just to name a few. On the Turkic side, one finds Turkish, Uzbek, Uighur, Kazakh and Khirgiz, among others, these associated with Turkey, Uzbekistan, Western China, Kazakhstan and Khirgizstan. Being in close proximity, the different languages have exchanged a lot of vocabulary over the years.

What about Arabic?
Both the Turkic and the Persian-speaking peoples tend to be Muslims. Because of the extent to which Islamic societies are supposed to follow the teachings of the Koran and the Hadith, a lot of Arabic terms have shown up for use discussing religion, society and law (which, strictly speaking, are aspects of the same thing). Even basic phrases like hello and please and thank you may have their roots in Arabic depending on the language.

So, to speak Turkic languages, I have to know Persian and Arabic?
No, but you’ll learn a fair measure of both if you undertake study of a Turkic language.

Like learning the French word “restaurant” to talk about a place to eat?
Very much like that.

So, what’s the hard part?
Take the phrase for giving thanks in Turkish, Uzbek and Uighur. In Turkish, it’s teşekkürler. This is almost certainly connected to the Arabic shokran (root sh-k-r, see the shehk-kur in the middle) with the plural ending –ler. It means, simply, thanks – the plural for a word meaning the giving of thanks. In Uzbek and Uighur, they say rahmat, or something similar. This is from the Arabic root r-h-m, which connotes mercy. Rahmat means act of compassion. In both cases, you’ve got an Arabic word playing a role, but which Arabic role and what word is a thornier issue.

So different languages might use different foreign words or different variants of their own?
Which means that if you know a word in one language, you’ll probably know words in the others, but they won’t always mean the same thing in the same context. On the other hand, the words that are equivalent in meaning in context might be different in and of themselves.

Sounds hard.
But it does provide an interesting and useful entrée into Islamic society on the Central Asian plain.

Sunday, August 29, 2004

The Two Great Challenges to Learning Turkic Languages

In approaching the Turkic languages, I have found, there are only two major difficulties: 1) the words... 2) what to do with them...

I can see you wiping the sweat off your brow already! Actually, while these two items - i.e. the whole of the languages - are going to prove a bit challenging, they don't have to prove insurmountable obstacles to the learner who knows what he or she is getting into. Coming up, we have:
  • a look at how geography, culture and history shaped Turkic vocabulary
  • thoughts on the morphology (word patterns) of Turkic vs. common European languages
  • what the dread vowel harmony has to do with all this
Sound like a lot? When you're done reading, you'll have seen it's a piece of cake.

Welcome to multilingua.info's Turkic page

multilingua.info seeks to improve your language learning by putting you face to face with not one, but two or preferably three languages. By approaching multiple related languages, you'll come in contact with lots of self-reinforcing material, helping your brain to build stronger, more meaningful associations.

For the Turkic languages, this project has added relevance: While linguists like to tie down what is and isn't a language and why or why not, the Turkic languages are not so much discreet entitities as speech patterns along a continuum. Experts can identify standard Uzbek vs. Uighur, but it's less clear at what point you stop speaking one and start speaking another. Undertaking the study of Turkish, Uzbek and Uighur, you won't just learn three languages - you'll be ready to take a stab at five or six others and engage speakers from Constantinople to Xianjiang.

Unfortunately, until the editors at multilingua.info have picked up more Turkic themselves, new articles will appear slowly. However, we'll be talking about recommended resources very soon, as well as explaining the two greatest challenges in learning Turkic languages for English speakers and some useful ways to think about them.

Geoffrey Barto
editor, multilingua.info