Turkish written with Latin letters half a millennium ago

« previous post | next post »

In "Türkçe'nin 500 Yıl Önce Latin Harfleriyle Yazılışı" (7/26/16), Abdurrahman Onur Çalışır presents a Turkish text written in Latin letters together with a translation into Latin:

Here's a transcription of the older Turkish, plus a rendering in modern Turkish:

Turca – Ne habar scizum girlerden?
Christianus – Hits nesle bilmezom tsaa dimege.
Turca – Gioldassum varmı tsenumle?
Christianus – Ioch, ialanuz geldum.
Turca – Benumle gelurmitsun?
Christianus – Irachmider tsenum utaghom?
Türk – Ne haber sizin yerlerden?
Hristiyan – Hiçbir şey bilmezim sana demeye.
Türk – Yoldaşın var mı seninle?
Hristiyan – Yok, yalnız geldim.
Türk – Benimle gelir misin?
Hristiyan – Irak(uzak) mıdır senin otağın(evin)?

And here is a translation into English:

[Turk] What news from your country?
[Christian] I know nothing to tell you.
[T] Do you have a companion with you?
[Ch] No, I came alone.
[T] Do you want to come with me?
[Ch] Is your house far?

The translation is by Erika Hitzigrath Gilson, a native of Istanbul, who has taught Turkish for many years at Princeton.  She wrote a dissertation (1981) under Tibor Halasi-Kun at Penn that dealt with a "Transkriptionstext." It appeared as The Turkish grammar of Thomas Vaughan:  Ottoman-Turkish at the end of the XVIIth century according to an English "Transkriptionstext" (Wiesbaden: Harrassowitz, 1987).

Erika notes:

As contacts between the Europeans and Turks grew, to facilitate communication texts were prepared by the Europeans using the Latin alphabet to represent Turkic/Turkish. This in no way suggests that 'Turkish was using the Latin alphabet'. . ..  Older samples [Codex Cumanicus] going back to the 12-13th century exist for other Turkic languages.

Guides to learning Turkish very often included passages, dialogues between a Turk and 'Christian.' Some of these were debates on religion, but mostly on daily matters.

Such "transcription texts" are very famous from the middle of the 16th century.  One of the best known is Hieronymus Megiser's Institutionum linguae turcicae libri IV (Leipzig, 1612; over 300 pages), about which Heidi Stein has written extensively.  György Hazai is another specialist on this type of transcription text.   Stanisław Stachowski also worked on such texts for many years.

The oldest surviving Turkish source in Latin characters is one by an Italian called Argenti.  See Milan Adamović, Das Türkische des 16. Jahrhunderts: nach den Aufzeichnungen des Florentiners Filippo Argenti (1533) (Göttingen: Pontus Verlag, 2001).  Older still is the 14th-century Codex Cumanicus, written in the Crimea by Italian merchants and German clergymen, reflecting two distinct Middle Kipchak dialects.

Turks had only one "national script", so to say (the so-called Runic or Old Turkic alphabet), which was given up at the turn of 10th-11th centuries. Otherwise, they wrote their languages in different adopted scripts, as can be seen very clearly from the Old Uighur period when 7 or 8 scripts were in use (of course, not always contemporaneously).  Modern Uyghur has also been written in a dizzying succession of scripts. Therefore, Ottoman in Latin script is only one phenomenon in the history of Turkic languages.

[h.t. Arif Dirlik; thanks to Marcel Erdal, Peter Golden, Peter Zieme, Mehmet Olmez, and Veysel Batmaz]


  1. Levantine said,

    August 30, 2016 @ 3:01 am

    Very interesting!

    There's a slight mistake in the transcription given of the book: "nesle" should be "neste".

    "Tsaa" very effectively captures what we might now transcribe as ""sağa"/"saa", a provincial/non-standard pronunciation of "sana" (Ottoman "saña").

    It's curious that the second-person genitive suffixes, which should end in N/Ñ, end instead in M (e.g., "scizum" for "sizin"/"sizüñ"). This is certainly not how they were written in Ottoman Turkish. I wonder if we're dealing with the author's erroneous mishearing or a reliable record of a more widespread non-standard pronunciation of the time.

  2. R. Fenwick said,

    August 30, 2016 @ 10:23 am

    @Levantine: I don't think we should assume the Latinate transcription was precisely phonemic. Latin word-final -m has never been particularly strongly tied to a place of articulation, and was in the process of being lost epigraphically even in the pre-Republic period in favour of vowel nasalisation, which itself was eventually lost in the daughters. I wonder if the transcription scizum might be intended to depict to the learner a pronunciation [sizũ], which may be the closest a Latin-speaker could get to the velar nasal of old Turkic varieties.

  3. Victor Mair said,

    August 30, 2016 @ 11:34 am

    From Peter Golden:

    The Rasûlid Hexaglot project in which I was involved and ultimately published the full work (The King's Dictionary: The Rasûlid Hexaglot), in which Tibor Halasi-Kun, Lajos Ligeti, Ödön Schütz and I collaborated – I was the only one still above ground after 20 years work on the project and published the results), is a complicated Transkriptionstext (ca. 1370s) using the Arabic alphabet (Arabic, or rather its medieval Yemeni literary dialect, was the control language) with entries in Persian (rather close to modern Kabul Darî), Turkic (in three different dialects, Oghuz, Oghuz-Qïpchaq mix and Khwârazmian), Byzantine Greek (reflecting an Anatolian or possibly Cypriot dialect), Armenian (more Western than Eastern dialect) and Mongol (as spoken in Ilkhanid Iran). What made it fascinating was the fact that the languages represented spoken rather than literary forms. I asked Tom Allsen to write an essay for the volume on the different multilingual dictionaries that began to pop up in the Mongol world (reflecting a wider world view) as well as in the non-mongol-controlled Middle East, India etc. The Codex Cumanicus (on which I have also worked) was one example of this interest in languages and the problems in transcribing them.

  4. Levantine said,

    August 30, 2016 @ 12:41 pm

    R. Fenwick, thank you — that's helpful and makes sense.

  5. Victor Mair said,

    August 30, 2016 @ 7:08 pm

    From a colleague:

    Those Turkish transcriptions are fascinating. My initial response is that it's amazing how little Turkish has changed over the past 500 years, compared to, say, English or běifānghuà (Northern Sinitic), however you'd want to classify that historically, though I haven't actually checked, so that might not be accurate (& my knowledge of English between Chaucer & Shakespeare is sadly lacking—I'm not quite sure what it was like half a millennium ago—but even Shakespeare strikes me at first glance as more different).

RSS feed for comments on this post