Some thoughts on Dravidian-Turkic-Sanskrit lexical comparisons

Türkbilig TÜRKOLOJİ ARAŞTIRMALARI 2012/24 Ankara <Güz> 2012 Türkbilig TÜRKOLOJİ ARAŞTIRMALARI ISSN : 1302-6011 Kuruluş Tarihi: 2000 Kurucusu / Founder Prof. Dr. Dursun YILDIRIM Sahibi ve Sorumlu Yazı İşleri Müdürü Doç. Dr. Bülent GÜL YAYIN DANIŞMA KURULU / PUBLICATION BOARD OF OVERSEERS Prof. Dr. Sema BARUTCU ÖZÖNDER [TÜRKİYE] • Prof. Dr. Chetin CUMAGULOV [KIRGIZİSTAN] • Prof. Dr. Nikolay İvanoviç EGOROV [ÇUVAŞISTAN-RUSYA] • Prof. Dr. Ahmet Bican ERCİLASUN [TÜRKİYE] • Prof. Dr. Shimin GENG [ÇİN] • Doç. Dr. G. Gonca GÖKALP ALPASLAN [TÜRKİYE] • Doç. Dr. Bülent GÜL [TÜRKİYE] • Prof. Dr. Şakir İBRAYEV [KAZAKİSTAN] • Prof. Dr. G. JORİGT [İÇ MOĞOLİSTAN-ÇİN] • Prof. Dr. Marek STACHOWSKI [POLONYA] • Prof. Dr. Dursun YILDIRIM [TÜRKİYE] YAYIN HAKEM KURULU / BOARD OF REFEREES Doç. Dr. Ferruh AĞCA • Prof. Dr. Şerif AKTAŞ • Doç. Dr. Nuray ALAGÖZLÜ • Prof. Dr. Mustafa ARGUNŞAH • Prof. Dr. F. Sema BARUTCU-ÖZÖNDER • Yrd. Doç. Dr. Bülent BAYRAM • Prof. Dr. Süleyman Hayri BOLAY • Prof. Dr. Nalan BÜYÜKKANTARCIOĞLU • Prof. Dr. İsmet ÇETİN • Prof. Dr. Özkul ÇOBANOĞLU • Prof. Dr. Çetin CUMAGULOV • Prof. Dr. Cem DİLÇİN • Doç. Dr. İbrahim DİLEK • Prof. Dr. Abide DOĞAN • Yrd. Doç. Dr. Oğuzhan DURMUŞ • Prof. Dr. Ahmet Bican ERCİLASUN • Prof. Dr. Bilge ERCİLASUN • Doç. Dr. Ruhi ERSOY • Yrd. Doç. Dr. Genç Osman GEÇER • Prof. Dr. Shimin GENG • Doç. Dr. G. Gonca GÖKALP ALPASLAN • Yrd. Doç. Dr. Faruk GÖKÇE • Prof. Dr. İsmail GÖRKEM • Doç. Dr. Bülent GÜL • Yrd. Doç. Dr. Galip GÜNER • Prof. Dr. Şakir İBRAYEV • Prof. Dr. Ayşe KIRAN • Doç. Dr. Fatma Sabiha KUTLAR • Yrd. Doç. Dr. Serdar ODACI • Prof. Dr. İsa ÖZKAN • Prof. Dr. Nevzat ÖZKAN • Prof. Dr. Musa Yaşar SAĞLAM • Prof. Dr. Ahmet TAŞAĞIL • Doç. Dr. Mete TAŞLIOVA • Prof. Dr. İsenbike TOGAN • Prof. Dr. Fikret TÜRKMEN • Prof. Dr. Alemdar YALÇIN • Doç. Dr. Paşa YAVUZARSLAN • Prof. Dr. Dursun YILDIRIM • Prof. Dr. Emine YILMAZ • Prof. Dr. Ayşe YÜCEL-ÇETİN • YAYIN KURULU / EDITORIAL BOARD Editör / Editor Dursun YILDIRIM Editör Yardımcıları / Assistants To Editor Bülent GÜL • Ferruh AĞCA • Mete TAŞLIOVA • Bülent BAYRAM • Oğuzhan DURMUŞ • Mehmet ÇERİBAŞ • Ali ILGIN • Sinan GÜZEL • Mikail CENGİZ • Haydar YALÇIN • Abdulkadir ATICI Yayın Kurulu Başkanı / Chairman of Editorial Board Bülent GÜL Yayın Kurulu Üyeleri / Members of Editorial Board Esra BİLGE • Gülnaz ÇETİNKAYA • Sibel HATİPOĞLU • Serdar ODACI • İsa SARI • Kadriye TÜRKAN İngilizce Editörü Doç. Dr. Nuray ALAGÖZLÜ Kapak Tasarımı: Mehmet Fidancı İç Tasarım: Bülent Gül * Yayının Türü Türkbilig, hakemli uluslar arası süreli (altı aylık) yayındır. Türkbilig, Uluslar Arası Türkoloji Dergisidir. ** Taranan İndeksler / Indekses Türkbilig, MLA (Modern Language Association) International Bibliography, EBSCO PUBLISHING, European Reference Index for the Humanites (ERIH), Ulrich’s Periodicals Directory ve ULAKBİM Sosyal Bilimler Veri Tabanı (SBVT) tarafından taranmaktadır. *** Türkbilig’de yayımlanan makaleler yayımcının yazılı izni olmadan tamamı veya bir kısmı herhangi bir yolla çoğaltılamaz. Yazıların fikrî sorumluluğu ve imlâ tercihi yazarlarına aittir. Başka kaynaklardan alınmış tablo, resim ve benzeri şeylerin yazılarda kullanım sorumluluğu yazara aittir. İLETİŞİM ADRESİ / INFORMATION ADRESS Türkbilig Türkoloji Araştırmaları H.Ü. Edebiyat Fakültesi Türk Dili ve Edebiyatı Bölümü 06800 Beytepe/Ankara TÜRKİYE web sayfası: http://www.turkbilig.com e-posta: [email protected] /[email protected] Türkbilig, 2012/24 Öncü Basımevi’nde (Sor.Müd. Yusuf AVAN; Kazımkarabekir Cad. 85/2 İskitler/ANKARA, Tel: 3843120) basılmıştır. Ankara <Güz>2012 Türkbilig’ten Merhaba, Türkbilig dergisinin takipçilerine ve teşvikçilerine, zorlukları büyük bir özveri ile göğüsleyen arkadaşlarımın her birine ayrı ayrı yürekten teşekkür ediyorum. Önceki sayımıza kadar, derginin yabancı dil editörlüğünü yürüten arkadaşımız Nalan Büyükkantarcıoğlu’ya karşılıksız verdiği emekler için teşekkür ederim. Bu görevi, aynı yüreklilikle üstlenen ve aramıza katılan Nuray Alagözlü arkadaşımıza da hoş geldin der, dergi için bu güzel yol arkadaşlığının uzun sürmesini dilerim. Yıllardır Türkbilig dergisinin sahipliğini ve yazı işlerini karşılıksız, hiç yüksünmeden yürüten Bülent Gül’ü ve Yazı Kurulu’nda görev alan arkadaşları, yazılara ilmi rapor yazan değerli bilim adamlarımızı, yazıları ile Türk bilim hayatına değerli katkılar veren bilim adamlarını ve varlığı ile aramıza katılıp Türkbilig’e güç veren G. Ahmetcan Asena’yı burada şükranla zikretmek isterim. Onların bu yürekli davranışı ve bu ağır yükü özveri ile götürebilecekleri yere kadar götürme kararlılığı, Türk bilim hayatına, gerçek bir bilim dergisi ile yürüme çabaları, elbette daha önceki sayılarda da olduğu gibi devam edecektir. Ne gündelik hayatta, ne bilim ve ne fikir hayatında yoldaşını yolda bırakanlardan olmadığımıza göre, imkânımız ve gücümüz olduğu ve Tanrı fırsat verdiği sürece, Türklük bilimine doğru yoldaşlarla at salıp yürümek kutlu bir görevdir, yürüyoruz. Bu derginin görevi, amacı, sadece ve sadece bilimsel ölçütlere ve bilimsel etiğe uygun biçimde Türk’e ait ne var ise, doğru, kanıtları ile ortaya konmuş, tenkidi, tahlili yapılmış gerçeklerin bütün yönleri ile, yeterli ve gerekli düzeyde tasvirini yapıp okuyucusunu aydınlatmaktan ve sosyal ve beşeri bilimler planında Türk bilim hayatına katkıda bulunmaktan ibarettir. Yunus Emre ne güzel söylemiş: ”İlim, ilim bilmektir/İlim kendin bilmektir/Sen kendini bilmezsen/Bu nasıl okumaktır”. Dünya üzerinde gördüğünüz her erişik devlet aygıtı, kurum ve kuruluşları bu gerçek deyişi, sözde değil, özde gerçeğe dönüştürdükleri için bu düzeye varmışlardır. Erişik olmayan toplumların devlet aygıtları, kurum ve kuruluşları ise, Yunus Emre’nin söylediği sözlerin özüne sahip olamayışın hafifliği içinde, kendi toplumlarına yalan veya yanlış bilgiler üzerine kurulu afura/tafura yüklü sözler ile şişinip mahallelerinde çalım satarlar, alem güler. Dünya böyle bir yer. Yunus Emre’yi anlamak, bu sözünün nereye uzandığını bilmek yetmez, bilim düzeyini o seviyeye çıkarmak, toplumun geçmişini, hâlini ve geleceğini buna göre hazırlamaya azmetmiş bir zihin berraklığı, hassasiyet de olması icap eder. Bu çabayı fiiliyat sahasına çıkaramayan veya aklı ile ruhu çalınmış toplumlarda, Yunus Emre’nin bu sözleri yerine çakma Köroğlu’ların esip yağıp gürlemeleri hükûm sürer ve bilim hayatı da buna göre şekillenir ve toplumlarda uğradıkları akıl tutulması nedeniyle akibetlerine doğru yürürler. Türkbilig dergisi, bilimin ışığına ve ölçütlerine sâdık biçimde yoluna devam kararı almıştır. Karanlıkta küçük bir ışık, Türk bilim hayatına küçük bir pencere olabilir ise, dünya Türkoloji çalışmalarına küçük bir ‘meydan’ açabilir ise, bu derginin ve onun çıkmasına hayat verenlerin alacağı en büyük ödül olacaktır, diye düşünüyorum. Hey gidi Yunus Emre heey. Senin gibi düşünüp Türk milletini kendini bulmaya adamış, ilim/ilim diye cihanı sesi çınlatmış Türk beyleri o güzel atlara binip varsan dönülmez şehrine gittiler. Gittiler de, o sese icabet eden hiç kimse bu diyarda kalmadı mı? Elbette kaldı, bu kervan onlarla yürüyor. Bu yüzden Türk milletine Tanrı bağışı M. Kemâl Atatürk, ‘Hayatta hakiki mürşid ilimdir” diyerek Peygamberin ‘İlim Çin’de ise, ara bul, edin” sözüne sadakat göstermiştir. Bilimden nasibi olmayanlar, bu sözlerin anlamını da, Yunus Emre’nin söylemiş olduğu sözler gibi, anlamamış ve her biri bir ‘Molla Kasım’ mertebesinde cehalet alıp satmıştır. Bugün karşılaştığımız her ne zorluk var ise, bu zihniyet yüzündendir. Türk aydını, bilim adamı, sosyal ve beşeri bilimler alanında yapacağı araştırma ve incelemeler ile, karanlıkta boğulmak istenen Türk milletini aydınlığa çıkarmakla yükümlüdür. Erişik devlet aygıtına, kurum ve kuruluşlara bu toplum sahip olmalıdır deniyor ise, Türk aydını bilimden, bilimsel ölçütlerden ve bilimsel etikten ödün vermeden toplumu aydınlatma işlevini sürdürmelidir. Merkezi Avrasya ve Önavrasya üzerinde yüzyıllar boyu iki yurt kurup medeniyete önderlik eden Türkler, bugün yeryüzünde paramparça edilip birbirine yabancılaştırılmış, farklı milletlere çevrilmek isteniyor ise, bu Türk aydınlarının ve bilim adamlarının, devlet aygıtlarının, kurum ve kuruluşlarının akıl tutulması içinde bulunmalarından, bilim hayatlarının yetersizliği nedeniyledir. Yeryüzünde bir de Araplar aynı durumdadır. Onlarda da bölünen toplumlara yönetim aygıtları bölenler eliyle verilmiş, o toplumcukları yönetecek köleler tayin edilmiştir. Bilim hayatı, Yunus Emre’nin sözleri henüz oralara da erişmemiştir. Halleri bu yüzdendir. Türkbilig, bilim yolu ile kendini arayanlara, öğrenmek isteyenlere, birliğe ve dirliğe bu şekilde ihtiyaç duyanlara bir bilimsel meydandır. Gücü ve imkânı nisbetinde bu görevi sürdürmeye devam edecektir. Dergi, bilimsel bağlamda ve belirtilmiş çerçevede yapılacak her özgün araştırmaya, her araştırmacıya açıktır. Bu dergi, her türlü siyasetten uzaktır ve sadece, geniş anlamda Türkolojiye bilimsel çalışmalar yönünde gönül verenlerin dergisidir. Türkbilig dergisinin çıkmasına imkân verenlere, onu size ulaştıran değerli arkadaşlarıma, siz okuyucularımıza yürekten teşekkür eder; Tanrı’dan, gelecek yılın aziz Türk milleti ve sizler, hepimiz için, Türkoloji için aydınlık, refah, huzur ve saadet getirmesini dilerim. Hoşça kalınız. Dursun YILDIRIM Editör Türkbilig Yayın İlkeleri • Türkbilig, yılda iki kez yayınlanır. • Türkbilig, hakemli bir yayındır. Türkbilig’de yayınlanmak üzere gönderilen özgün makaleler, • • • • • • • • • • • YAZI KURULU tarafından incelendikten sonra konunun uzmanı üç hakem tarafından değerlendirilir ve üç hakemin en az ikisinden olumlu rapor gelmesi halinde yayınlanır. Türkbilig’e gönderilen yazılar daha önce hiçbir yerde yayınlanmamış olmalıdır. Derginin yayın dili Türkçe’dir. Yayın ilkelerine uygun olmak koşuluyla yabancı bilim adamlarının yazıları, İngilizce, Almanca, Fransızca veya Rusça yazılmış olabilir. Ancak Türkçe-İngilizce başlık, özet ve anahtar sözcükler yazıya eklenmelidir. Yabancı dilde yazılmış yazılara derginin hacmine göre % 30 civarında yer verilir. Yazılar, basılı üç kopya halinde ve disketiyle birlikte gönderilmelidir. Özel çeviri yazım işaretlerinin kullanıldığı yazılarda fontlar da (PC uyumlu) diskete yüklenmelidir. Makalenin yazarı, adını, soyadını, görev yaptığı kurumu ve akademik unvanını tam ve açık olarak belirtmeli, kendisiyle doğrudan iletişim kurulabilecek açık adres, telefon numarası ve elektronik posta adresini vermelidir. Yazıların başında kısa birer Türkçe ve İngilizce özet (en çok 100 sözcük) ile Türkçe ve İngilizce anahtar sözcükler (en çok 10 sözcük) bulunmalıdır (İtalik olarak ve Times 9 punto ile yazılmalıdır). Yazılar, Apple Mac Word 5.1 veya MS Word Windows 95 ve üstü programla, Times 10 puntoyla ve 1,5 satır aralığıyla yazılmalıdır. Paragraf başlarında tab tuşu, paragraf aralarında enter tuşu kullanılmamalıdır. Metin içinde göndermeler ad ve tarih ve/veya sayfa olarak parantez içinde belirtilmelidir. Örnek: (Tanpınar 1985) veya (Tanpınar 1985: 316). Üç satırdan az alıntılar satır arasında ve tırnak içinde, üç satırdan uzun alıntılar ise satırın sağından ve solundan birer santimetre içeride, blok halinde, 9 puntoyla, tek satır aralığıyla verilmelidir. Dipnotlar sayfa altında, numaralandırılarak verilmeli ve sadece açıklamalar için kullanılmalıdır. Makalenin sonunda yer alacak kaynakçada kitaplar (koyu ve italik) ve makaleler (dergi adı koyu, cilt Romen rakamıyla, sayı, üst üste iki nokta, sayfa numaraları) alfabetik sırayla ve şu düzenle verilmelidir: ELÇİN, Şükrü, (1998), “Yeşil Abdal’ın Bir Şiiri”, Folkloristik: Prof. Dr. Dursun Yıldırım Armağanı, (Ed. M. Özarslan-Ö.Çobanoğlu), Ankara: Feryal Matbaacılık, 216-231. STOELTJE, J. Beverly, (1983), “Festival in America”, Handbook of American Folklore, (ed R. M. Dorson.), Bloomington: Indiana University Press, 239-246. TANPINAR, Ahmet Hamdi, (1985), XIX. Asır Türk Edebiyatı Tarihi, İstanbul: Çağlayan Basımevi. ÜLKEN, Hilmi Ziya, (1952), “Milli Destan ve Folklor”, Türk Folklor Araştırmaları, II, 33: 513514. WELLEK, R. ve A. WARREN, (1982), Yazın Kuramı, (Çeviren: Y. Salman ve S. Karantay), İstanbul: Altın Kitaplar Yayınevi. WOOLF, Virginia, (1986), “Kitap Nasıl Okunmalı?”, (Çeviren: Kemal Atakay), Adam Sanat, 4: 5-14. Bir yazarın birden fazla yayını kaynak gösterildiği takdirde yayınlar tarih sırasıyla, aynı yazarın aynı yıldaki yayınları ise (1985a), (1985b) şeklinde harf sırasıyla verilmelidir. Tezlerin hangi üniversitede yapıldığı ve hangi akademik dereceye (yüksek lisans/doktora...) yönelik olduğu belirtilmelidir. Yukarıda belirlenen yazım koşullarına uygun olmayan yazılar değerlendirmeye kesinlikle alınmayacaktır. Submission Guidelines for Türkbilig • Türkbilig is published twice a year. • Türkbilig is a refereed publication. After original manuscripts are examined by Executive Board, they are peer-reviewed by three referees. They are published if two of three referees send affirmative report. The authors bear the full responsibility for their articles. • Manuscripts must be originally sent to the board, and no other places for publication or evaluation. The publication language of the issue is Turkish. In condition suited to submission guidelines of Türkbilig, foreign authors’ manuscripts, which can be in English, German, French or Russian, are published in proportion 30% of issue. The manuscripts must have title, abstract and also keywords both Turkish-English. • Manuscripts must be sent as three printed copies along with the disk. If any special fonts (suitable for PC) are used in manuscripts, they must be in the disk. • The authors’ names, last names and academic positions should be written. In addition, the full postal address, fax, telephone numbers and e-mail addresses of the author(s) who will check proofs and receive correspondence and offprint should also be included. • Summaries must be in Turkish and English, should not exceed 100 words and key words may not exceed 5 (5 words for Turkish; 5 words for English) words at the beginning of the manuscripts. Latin words must be in Times 9 point and italic. • Manuscripts must be written in Mac Word 5,1 or Ms Word Windows 95 or further versions (Pc compatible) with Times 10 point with 1,5 line spaced. Tab and enter key must not be used for paragraphs. • Texts must follow in-text footnote system. In parenthesis in the text, author’s name, date of publication, and page number is given. If a source is cited many times, parentheses are given instead of “ibid, idem, op. cit. etc.” For example, (Tanpınar 1985) veya (Tanpınar 1985: 316). Quoted passages under three lines must given by quotation mark. If quoted passage is over three lines, they must be given 1 cm margins from left and right side of line as block with 9 point and odd line spaced. • Additional information must be given on the same page as footnotes enumerated 1, 2, 3. Citations in them must follow the above guidelines. • References must include only the cited sources and be given in alphabetical order. Books names must be written bold and italic. Articles must be written (name of journal is bold, volume in Romen number, issue number, colon and page numbers) as below: ELÇİN, Şükrü, (1998), “Yeşil Abdal’ın Bir Şiiri”, Folkloristik: Prof. Dr. Dursun Yıldırım Armağanı, (Ed. M. Özarslan-Ö.Çobanoğlu), Ankara: Feryal Matbaacılık, 216-231. STOELTJE, J. Beverly, (1983), “Festival in America”, Handbook of American Folklore, (ed R. M. Dorson.), Bloomington: Indiana University Press, 239-246. TANPINAR, Ahmet Hamdi, (1985), XIX. Asır Türk Edebiyatı Tarihi, İstanbul: Çağlayan Basımevi. ÜLKEN, Hilmi Ziya, (1952), “Milli Destan ve Folklor”, Türk Folklor Araştırmaları, II, 33: 513-514. WELLEK, R. ve A. WARREN, (1982), Yazın Kuramı, (Çeviren: Y. Salman ve S. Karantay), İstanbul: Altın Kitaplar Yayınevi. WOOLF, Virginia, (1986), “Kitap Nasıl Okunmalı?”, (Çeviren: Kemal Atakay), Adam Sanat, 4: 5-14. • If more than one source of the same author is cited, they must be put in a chronological order from the oldest to the newest. Sources of the same years must be given letters “1985a, 1985b” • The university and the academic degree (MA. or PH.) of academic thesis must be given. • Manuscripts not prepared based on the directions above will not be taken into consideration for publication in Türkbilig. İÇİNDEKİLER / CONTENTS Türkbilig’den Merhaba / iii Yazılar / Articles Ekrem KALAN İvolga Örneğinde Hun Kent Kültürüne Genel Bir Bakış An Overview of the Hun Urban Culture in the Case of Ivolga 1- 19 Akartürk KARAHAN Kâşgarlı Mahmud’un Kayıtlarında İdil Boyu Türk Lehçeleri Turkish Dialects of the Idil Tribe in the Records of Kashgarlı Mahmud 21-32 Ahmet BENZER Belirtili Ad Tamlamalarının Anlamları Meanings of determinative noun phrase 33-40 José Andrés ALONSO DE LA FUENTE Some Thoughts On Dravidian-Turkic-Sanskrit Lexical Comparisons Dravid Dilleri, Türkçe Ve Sanskritçe Sözlüksel Karşılaştırmalar Üzerine Düşünceler 41-76 Süer EKER ‘Orta Asya’nın Gizemli Halkı’: Soğdlular Soğd Ve Soğdca ‘Mysterious People Of Central Asia’: Sogdians Sogdiana and Sogdian Language 77-92 Erdem UÇAR Uygurca Birgerü Üzerine About “Birgerü” in Uighur 93-100 Ferit YUSUPOV Tatar Türkçesi Ağızları: –DI Ve –GAN Ekli Geçmiş Zaman Şekli Tatar Dialects: Past Tense Forms –DI and –GAN 101-120 Hakan AKCA Ağız Çalışmalarında Gramatikal Dizin Hazırlarken Karşılaşılan Sorunlar The Problems Encountered While Preparing A Grammatical Index in Dialect Studies 121-142 Refika Altıkulaç DEMİRDAĞ Dağlarca’nın Çocuk Ve Allah’ında Dinginlik Ve Mistik Duyuşlar Serenity and Mystic Impressions In Dağlarca’s Novel “ Çocuk Ve Allah” (Child and God) 143-154 Şahika KARACA Aslı Erdoğan’ın Taş Bina Ve Diğerleri’nde Mekânın Hâlleri Different Locations in “The Stone Buildıng And The Others” By Aslı Erdoğan 155-170 Fatih SAKALLI Ahmet Şuayb’in “Musahabe-i Edebiyye” Ve Diğer Yazıları “Musahabe-i Edebiyye” and Other Writers of Ahmet Şuayb 171-186 Kemal ÜÇÜNCÜ Mehmed Esad Serezli’nin Kayıt Ve Tanıklığıyla Serez Yöresi Türk Kültür Geleneği Serez Region Turkish Cultural Tradition As Witnessed And Registered By Mehmet Esad Serezli 187-208 Çeviriler / Translations Sergei Anatolyevich STAROSTIN Çev.: Osman Ali ÇETİN Çince Altayca İlişkisi 209-214 Eleştiri ve Tanıtmalar / Reviews Bülent GÜL Mihály Hoppál (2012), Avrasya’da Şamanlar, (Çev. Bülent Bayram-H.Ş. Çağatay Çapraz), Yapı Kredi Yayınları, İstanbul, 345 s. 215-217 Mutlu ER Berna Moran, Der türkische Roman Eine Literaturgeschichte in Essays-Band 1, Mizan, Harrassowitz: Wiesbaden. 219-222 Esra İLTER Han- Woo Choi, (2010), Türkçe, Korece, Moğolca ve Mançu- Tunguzcanın Karşılaştırmalı Ses Ve Biçim Bilgisi, Türk Dil Kurumu, Ankara. 223-224 Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS * José Andrés ALONSO DE LA FUENTE* Abstract: The main goal of this paper is to analyze a specific group of Turkic lexical items whose historical destiny has been recently tied to the Dravidian languages. Overall, it is unclear how these lexical items should be dealt with, since authors supporting this borrowing route offer no clear picture of the exact nature of the historical setting. In this paper I hope to demonstrate that there is no such thing as Dravidian-Turkic direct contacts. Instead, more conventional borrowing routes, well-known in the specialized literature, might account for the majority of examples brought into discussion. Several other instances, however, must be regarded as cases that arise due to chance similarity. The bulk of evidence relies on both (pre)history and (historical) linguistics. Keywords: historical & comparative linguistics, language contact, historical populations, etymology, Dravidian & Turkic languages. Dravid Dilleri, Türkçe Ve Sanskritçe Sözlüksel Karşılaştırmalar Üzerine Düşünceler Özet: Bu yazının temel amacı yakın zamanda Dravidçe ile ilişkilendirilmiş olan bir grup Türkçe sözcüğü çözümlemektir. Böyle bir ödünçleme yolunu savunan araştırmacılar tarihi ortamın gerçek yapısı ile ilgili yeterince açıklayıcı olmadıkları için bu bağlantıların nasıl kurulduğu belirli değildir. Bu makalede Dravidçe ve Türkçe arasında doğrudan temasın / etkileşimin olmadığını göstermeyi umut ediyorum. Bunun yerine, ilgili yazında iyi bilinen daha geleneksel ödünçleme yolları tartışılan örnekleri açıklayabilir. Ancak diğer örnekler rastlantısal benzerliğe dayalı olarak ortaya çıkan durumlar olarak görülebilir. Kanıtların yada bulguların çoğu hem tarihe (tarihöncesine) hem de dilbilime (tarihsel dilbilime) dayanmaktadır. Anahtar Kelimeler: Tarihsel ve Karşılaştırmalı Dilbilim, Dillerarası Etkileşim, Kökenbilim, Tarihsel topluluklar, Dravid ve Türk Dilleri. §§1. Introduction; 2. Theoretical basis of the Dravidian-Turkic (-Sanskrit lexical) contacts (2.1., 2.2.1., 2.2.2., 2.3.); 3. “New” linguistic evidence; 4. Prevention is better than cure (4.1., 4.2.1., 4.2.2., 4.2.3.); 5. Conclusions. Abbreviations, References. * * I would like to express my most deep gratitude to Marek Stachowski, Tomasz Majtczak, Michał Németh, and Kamil Stachowski (Uniwersytet Jagielloński), Michael Witzel (Harvard University) and Sanford Steever, who read the manuscript of this paper and made many valuable suggestions and corrections. Any remaining errors are, of course, my own responsibility. Universidad del País Vasco–Euskal Herriko Unibertsitatea. Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE 1. Introduction Substratum theories and the discovery of new borrowing routes is a common trait in the study of language contact. A lack of good evidence has become the usual feature of hypotheses dealing with potential genealogical links between (language) families, therefore all of them are systematically rejected. As a result of this, the alternative solution to the genealogical linkage, i.e. areal interaction, arises as the single conceivable (one could also say legitimate) proposal. Thus, every field has its very own substratum theory to account for those lexical and morpho-phonological elements which stand apart from the solid, inherited, defining linguistic core of a given group of languages. Two of the most celebrated works on this matter in the field of Eurasian studies are without a shadow of a doubt Gerhard Doerfer’s monumental treatises Mongolo-Tungusica (1985) and especially Türkische und mongolische Elemente im Neupersischen (1963-1975). These impressive encyclopedic works cover actually most of the loanwords that occurred between components of the so-called “Altaic” family, i.e. Mongolic, Turkic, and Tungusic,1 and the languages from their periphery, basically Indo-Iranian. In spite of such a broad coverage, it is still necessary to explain many other lexical items which potentially involve languages generally not considered to belong to the Eurasian sphere, i.e. Sino-Tibetan, Dravidian, Yeniseian, Chukotko-Kamchatkan, etc. Of course, much has been done in this respect during the last few decades, but we still lack a work in lieu of Doerfer’s magnificent treatises. The utility of such works is immediately tested against certain hypotheses whose very formulation raises problems. The Dravidian-Turkic connection is one such hypotehsis. In spite of self-evident geographical and (pre)historical difficulties, more than one author has dared to take a look at the problem and draw conclusions, always in the fashion of “Dravidian-Turkic linguistic contact is supported by several lexical items which otherwise cannot be assumed to be native”. To the best of my knowledge, the first serious appreciation towards a more elaborated presentation on these “lexical contacts” 2 comes from Robert Caldwell (1814-1891), who in his excellent comparative grammar of the Dravidian languages (19133[1998]), first published in 1854, had already noticed that several Dravidian lexical items very suspiciously resembled Turkic words. Caldwell defended the common origin of all 1 2 42 Japanese and Korean are commonly also included in the equation, thereby forming what Street’s named “Macro-Altaic”, but since these two languages are of no use for the current discussion, I will not mention them any more. No doubts, other authors noted and commented upon similarities between these and other languages before. However, Caldwell was the first to do it with a grain of criticism and common sense. It is opinio communis that Caldwell was ahead of his time in understanding and applying the comparative method. Caldwell’s work was not in vain, as he can presume of having written a comparative treatise, covering all aspects of traditional grammar, is still valid almost in its entirety one-hundred-and-fifty years after its publication. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS of them, under the paradigm of what was called, at that time, the Scythian language family after Rasmus Rask’s Den skytiske Sproget (1934). Later, individual treatments of the question do not abound, the topic having been subsumed under discussions on larger proposals of “Nostraticistic” nature. As far as I am aware of, the most recent attempt at a discussion can be found in a paper addressed to the Festschrift volume to honour L. Johanson, where the Russian Turcologist Kenesbay Musaev3 touched upon the topic, introducing, as one would rather expect, Sanskrit into the equation (1996). 4 Naturally enough, Sanskrit is the key language which could actually account for such “Turkic-Dravidian” correspondences. However, in Musaev’s opinion, there are many instances where Sanskrit’s influence does not apply, and therefore Dravidian-Turkic contacts should be considered to explain those rare cases.5 2. Theoretical basis of the Dravidian-Turkic(-Sanskrit lexical) contacts 2.1. A significant part of the Dravidian scholar community seems to have always been open up not only to substrata theories, but also to long-range comparisons.6 In fact, Kamil Zvelebil (1927-2009) once commented on A. Dolgopolsky’s Nostratic work saying that “[a] Dravidianist can find only a small number of Proto-Dravidian reconstructions which would be either unacceptable or rather doubtful. Dolgopolsky has clearly worked very carefully with the Dravidian Etymological Dictionary. [...] Another very positive feature of the Dolgopolsky approach (I am again speaking from the Dravidian perspective) seems to be his caution and a great deal of common sense” (1999: 360-1). Thomas Burrow (1909-1986), the co-author along with Murray Emeneau of the monumental (comparative-)etymological dictionary of the Dravidian languages mentioned by Zvelebil in the previous quote, or the Russian specialist Mikhail Andronov, expressed positive thoughts on the matter in press 3 4 5 6 As usually, the transliteration of the name (Cyrillic М а ) appears differently rendered, e.g. “Musayev”. In this paper I will homogenise all of them by using only “Musaev”. This paper is actually a translation of Musaev (1984: 147-53). Readers will not be aware of this fact unless they know Musaev’s book, that had been published 12 years earlier. What is more notorious is the fact that Musaev did not make use of this opportunity to reconsider his ideas, the only changes he introduced being seemingly the data on population census in ft. 1. At the same time, it is worth noting that Musaev was engaged in discussions about the lexical origins of Turkic vocabulary as soon as (1975), when his ideas about language contacts were much more restricted. Thus, regarding potential IndoIranian loanwords, Musaev comments: “И а , х а а а c pe а х а ” (1975: 337). Thus, one is forced to conclude that Musaev changed his mind in favor of a much more open attitude. Alternative and conventional routes and/or chance similarity are after all the solution to many proposed “exotic” lexical contacts, e.g. Gramkrelidze & Ivanov’s IndoEuropeanisms in “Altaic” (Róna-Tas 1988, also see Erdal 1993) or Dul’son & Werner’s Germanisms in Yeniseian (vid. i.a. Georg 2008: 154). For genealogical links involving Dravidian languages with a brief characterization and prospects see in general Krishnamurti (2003: 43-7). 43 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE (1944 and 1971, respectively). The above-mentioned Robert Caldwell, 7 was somewhere in the middle, for although he always seemed to be somewhat optimistic about long-range relationships, at the very same time was also well aware of the fact that Eurasia could be after all an immense Sprachbund (or a cluster of Sprachbünde). 8 On the contrary, Bhadriraju Krishnamurti, well known for his skepticism on these enterprises, 9 has recently recognized, talking this time on Tyler’s Dravidian-Uralic connections (1968), that: “From the English-speaking world Tylor has a convincing paper [...] showing lexical comparisons between Dravidian and Uralian based on 153 etymologies. He has also given a convincing set of correspondences between Proto-Uralic and Proto-Dravidian [...]. The phonetic-semantic similarities are striking, although one does not know how to interpret these data genetically” (2003: 258). 10 It is legitimate to assume that Krishnamurti’s opinion on Nostratic, although never stated per se in press, would be formulated in similar terms, if not more pessimistic ones. Murray Emeneau (19042005), Krishnamurti’s teacher, as well as other Dravidian scholars like Jules Bloch (1880-1953), never supported or rejected these kinds of comparisons, but as happened in the case of Krishnamurti, one can conclude with a high degree of confidence that the fact the topic was passed over in silence in their publications is a sign of skepticism, if not plain rejection. With this brief bibliographical account what is shown is, that in spite of such academic openness (even towards Nostratic!), no one has attempted to offer any serious “down-step” research to find out whether these similarities could be the result of different borrowing layers or that they are actually evidence of genealogical linking. By now the most logical and reasonable conclusion is that of Krishnamurti, which I repeat again here: “[...] one does not know how interpret 7 8 9 10 44 However, Caldwell did not discover the relationship between the Dravidian languages, or, more properly, they displayed no-relation with Sanskrit and other Indo-Iranian languages, a merit belonging to Francis Whyte Ellis (1777-1819). See now Trautmann (2006) for an excellent treatment of the question. See for example Caldwell’s plain rejection of any Semitic connections (19133[1998]: 605-6) in contrast to what later happened with Nostratic vel sim. proposals, according to which it seems that Semitic languages play a central role (in spite of several internal “dissidents” who instead prefer to see Nostratic and Afro-Asiatic as two equal taxonomic levels or the sort). See his reaction to McAlpin’s Elamo-Dravidian hypothesis, in theory “closer” in every possible sense than the Dravidian-Uralic one (2003: 43-7). Interestingly enough, the very same Zvelebil hailed MacAlpin’s “achievements”: “[D. W. McAlpin] has removed Dravidian languages from their isolation by positing what he calls ‘Proto-ElamoDravidian’” (apud Sorrentino 1988: 245). Tyler does not address the questions where, when and how. It means, he does not elaborate farther about time and space. Southworth (2005: 49, 89, 255), while recognizing the value and interest of Tyler conclusions (then theoretically Southworth seems to agree with Krishnamurti and Burrow), he neither addresses these questions. One wonders what would be his opinion regarding Dravidian-Turkic comparisons. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS these data [...]”. 11 However, it is my understanding that if we concentrate for a moment on the Turkic-Dravidian aspect of the question,12 we will soon realize that there is only one way to interpret the data, since most of it can be accounted for by assuming chance similarities and areal contact involving not-so-new borrowing routes. 2.2.1. According to Musaev’s exposition, Turkic-Dravidian contacts belong to the parental language phenomena, i.e. contacts between one and another took place in Proto-Turkic and Proto-Dravidian times (1996: 173-4). A hypothesis can be put forward that the Dravidian languages, still preserved as linguistic enclaves in Afghanistan, and the Turkic languages and their ancestors had direct contacts and possibly a common base, not less than 4000 years ago, i.e. not later than the 2nd millennium B.C. This means that before the Indo-Iranian conquest 3500 years ago the Turks and the Dravidians were direct neighbours and that it was only the Indo-Aryan wave that severed their relations. Their contact could possibly have taken place in the territories of Middle Asia, of the Ural-Volga region, and of the Caucasus. The previous paragraph contains several statements and ideas which deserve careful attention and some extensive commentaries. Fortunately enough, where and when those proto-languages could have been spoken is a matter of not much dispute, at least in comparison to that in other fields like Indo-European or Tungusic linguistics.13 It is necessary to say from the very beginning that actually there are no 11 12 13 The situation gets even worse if one prefers to give some credit to Ramanathan’s ProNostratic hypothesis (2002, 2003), according to which Dravidians moved into South Asia over a land bridge, probably a series of islands, from Africa when the seas were lower on account of glaciations. Ramanathan considers that Dravidians are related to Australians, and that since the former entered Australia from South India at least 40,000 years ago, then it is easy to imagine that Dravidians, who moved then from South to North, and into Iran and further (Uralic, Altaic, and even Indo-European break off from a main Dravidian stock), are the native population of India from long time ago. For a balanced presentation of Ramanathan’s ideas, see recently Levitt (2009: 140-4). Although these new theories on migratory Dravidians combine quite nicely data from genetics, archaeology and prehistory, their supporters still miss the “linguistic point”: all these data coming from non-linguistic fields may say nothing on linguistic genealogical relationships until hard evidence, only linguistic in nature, is put forward, and that has not been done so far. To take into account pro-Nostratic literature, with a balanced critique, requires a knowledge of languages and philologies that unfortunately I am unable to respond for. However, I offer a hint in Alonso de la Fuente (2007) as to how Dravidian materials are very often used in Nostratic studies by analyzing the well known comparison between Indo-European */médh-u-/ ‘honey’ and (South) Dravidian */maṭṭu/ ‘honey, toddy’ (plus additional materials from Finno-Ugrian and Etruscan). In the end it turns out that the most likely solution involves, as expected, areal contact and chance similarities (Levitt 2003: 178 prefers to connect the Dravidian word to PIE */méli-t/ id., both motivations and argumentation remaining unclear to me). At this point it would be interesting to note that both Dravidian and Turkic historical linguistics are well established fields with more than a century of tradition. In his general 45 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE recognized pre-historical connections enabling us to set up a realistic scenario where such exchange of words between speakers of PDr. and PTr. could have taken place. As is well known, Dravidians may have been the native population of India. Although it is gaining more supporters every day, i.e. the view that (Proto)Dravidian could have been spoken in the Northwestern regions of India (current Gulbarga, Raichur, Bellary, the district of Karnataka and Kurnool, one of the departments of Andhra Pradesh) in correlation with the Southern Neolithic Complex culture, the traditional position locates those same speakers much more deeper in the Southern part of the continent (vid. i.a. Krishnamurti 2003: 2-5, Southworth 2005: 245-55). Brahui and Malto, members of the North Dravidian branch, are two languages which clearly branched off later and initiated a long process of migration to their current localizations, even though Brahui does not seem to retain remarkable archaisms. Whatever option one chooses, the dating for Proto-Dravidian in both hypotheses still revolves around the second half of the 3rd millennium BC, with the entrance of Dravidians in India around 4th millennium BC. This simple fact already diminishes almost totally the possibility of potential contacts with (Proto)Turkic populations, which in this epoch were located much more into East Central Asia. What is more important, between one group and another, is that several civilizations were finding their place in the continent: Indus, later Indo-Iranians and Indo-Aryans, Sino-Tibetans, etc. 14 Dravidian and Indo-Aryan (not Indo-Iranian!) contacts did not begin till the Ṛg-Vedic period, i.e. 1200 B.C., the moment in which it is possible to talk about the first (Old) Indo-Aryan and Dravidian contacts. 15 These contacts affected Old Indo-Aryan in different ways. One of the most interesting was studied by the Dutch scholar Kuiper who in an excellent paper (1967) demonstrated that the use of OIA iti ‘thus’ was syntactically inspired (and subsequently copied!) by the usage of the past participle of PDR. */an/ ~ */en/ ‘to say’ (DEDR [868]). The reverse situation, i.e. Indo-Aryan influence on Dravidian, is uncommon and some of the not so many alleged borrowings are contested or polemic. 16 Well consolidated and generally accepted instances however can be 14 15 16 46 appreciation, Campbell (1998: 164-5) estimated that the state of the PDr. and PTr. reconstruction is “moderate”. See map of substrate languages of pre-Indo-Iranian in Southworth (2005: 65). Whether Indo-Aryans were immigrants into the Indian subcontinent, or indigenous to it, is a very complex question marginal to the present issue. Bryant & Patton (2005) offers a balanced discussion of the pros and contras of each position (see “Introduction”, pp. 1-18, and “Concluding Remarks”, pp. 468-506). Further detailed discussions may be found in different contributions to that volume. Those cases of lexical contact between Dravidian and Indo-Iranian deserve almost the same qualification. The contact between these two parental languages would have taken place somewhere around Badakhshun and comprises cases like PII */māyā-/ ‘supernatural’: PDr. */mac-/ ~ */may-/ ‘to mystify, confuse’ (or */māy-/ ‘to disappear, be lost’), etc. For further examples and extensive comments see Southworth (2005: 88-90). Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS quoted, e.g. OIA akṣa ‘axle’ → PDR. */accu/ ‘axle’ (DBIA 7) or OIA s ci ‘needle’ (already in Ṛg-Veda, from s v- ‘to sew’) → PDR. */cūci/ ‘needle’ (DBIA 171).17 These general facts are not mentioned for nothing because: 1) if the earliest lexical contacts between Indo-Aryan and Dravidian did not start before 1200 B.C. and 2) any possibility of contact between Turkic and Dravidian populations had already happened before the third millennium B.C., then those items which are allegedly shared by Turkic, Dravidian and Sanskritic materials cannot be explained as result of Turkic-Dravidian contacts. So what about Turkic? The parental language from which all the historical Turkic languages descend,18 has been for a long time thought to be spoken somewhere in west and central Siberia, although maybe also in southern regions, but not further into it. The earliest news concerning Turkic peoples seems to support this assumption, for Chinese sources say that Turkic tribes were living in the Northeastern part of the Central Asiatic steppe zone, i.e. roughly speaking in present-day Northern Mongolia (basic facts in Róna-Tas 1998: 67-80, Golden 1998: 16-21, 2005, esp. 138-40, generally Golden 1992). The current distribution of Turkic people is a secondary migratory phenomenon, and it can be accounted for by accepting the idea of a Northeastern Urheimat. As it is selfevident, this information does not help at all to see how Dravidian and Turkic populations could have come into contact. 2.2.2. Does all this means that Turkic and Dravidian, or for that matter, Indian populations, were never in contact? Only in recent times have contacts between Turkic and mainly Iranian populations (vid. i.a. Aalto 1971, some contributions in Johanson & Bulut 2006) shown it is possible that some terms traveled from Turkic to Dravidian languages, using as main intermediaries Urdu or Persian. The first contact between Turkic and Indian populations—by Indian I mean not only 17 18 Other authors claim that the convergence was much more multidirectional and involved many other languages, e.g. Tibeto-Burman, Nahali or Burushaski (vid. i.a. Hock 1975). However, this is of no relevance here, for in Musaev’s hypothesis Turkic-Dravidian contacts had to take place in even earlier times than those discussed by Hock and other authors supporting his views. See Rybatzki (2010) for Turkish loanwords in Burushaski. Witzel (2006) addresses some of the problems involved in the question, but unfortunately not so extensively as to cover also the res Turcologicæ. However, it may be used as a methodological framework to work out other issues, as we are attempting in the present paper. Witzel (2005) and Southworth (2005) are excellent presentations of many points touched upon in discussions on Indo-Aryan substrata. The exact position of Chuvash is still a matter of discussion, but generally speaking, Chuvash and Turkic (= Common Turkic) are actually descendants of an older entity (Proto-Turkic). Many of the Chuvash features are best accounted for by assuming that this language was in contact with both Common Turkic and Common or ProtoMongolian at some point in the past. For a clear presentation of the problem, vid. i.a. Poppe (1965: 33-8) and Schönig (2003). As is immediately obvious, Chuvash material is of paramount importance in evaluating the nature of Turkic etymologies, especially from a chronological viewpoint. 47 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE Dravidian groups, but rather any of those populations inhabiting India in more recent times—of which we are aware has been recently described by KuczkiewiczFra (2001), who is also interested in the linguistic aspect of population contact between Turkic nations and India. As she informs us, “[t]he first significant contact between Indian and Turkic people and languages took place in the period of the conquest of Northern India by the army of Mahmud of Ghazna in the beginning of the 11th c. A Turkic dynasty founded by him and called ‘Ghaznawids’ ruled in Punjab and parts of Sind to the end of the 12th c. At the same time the neighboring area of Khwarazm and Khorasan (which were provinces of Eastern Iran) was seized by the Seljuqs, a branch of the Oghuz Turkic people belonging to the steppes north of the Caspian and Aral seas. The Punjab, and particularly Lahore, under Ghaznawids become a great market-place for goods and merchants from West (Iran) and East (India). It was also a place where different languages and cultures met and mixed. [...] soldiers of their armies were recruited in the greater part from the peoples inhabiting the large region of Central Asia, from the borders of China in the East to the Mediterranean in the West, but the official and cultural language of the army was Persian (in that time also strongly influenced by Turkic)” (2001: 44). It goes without saying that this scenario does not suffice to account for the lexical items that Musaev thinks are an evidence in support of his Dravidian-Turkic lexical contacts. So far, Iranian-Turkic contacts is the earliest layer of lexical borrowing we can posit between Turkic and Indian continent populations.19 2.3. From what can be inferred in Musaev’s exposition, his confidence about the solidity of Dravidian-Turkic contacts relies on three lexical comparisons (1996: 171), all of them taken from a very obscure publication. 20 I will quote here Musaev’s passage verbatim in order to avoid possible misunderstandings: Concerning the connections of the Dravidian languages with Turkic, Indian scholars state in particular the following: The Tamil word denoting “Muslim” —Tulukkan— came from the North Indian, and the Sanscrit correspondence is Turuṣka ‘the Turk’. 19 20 48 It must be noted that Musaev seems to be aware of the fact that his hypothesis is highly incompatible with current views. Thus, he says: «One widespread theory is of the late arrival of Turkic tribes to Middle Asia and Europe from Central Asia. If we accept this theory the problem of ancient ties between the Turkic languages (and their ancestors) and the Dravidian languages loses its foundation entirely, though it is hardly a proper scientific solution.» (1996: 170), and after presenting his own hypothesis, affirms that «[t]his supposition differs from those fixed notions imposed upon the world for a century by the Indo-Europeans, and as a result the existing theories of the relationship between the Indo-European languages and other languages of the world will need reconsideration.» (ibid., p. 174). Why Musaev should blame “Indo-Europeans” in this context is beyond my comprehension, for the current, whole picture of the archaeological and linguistic situation in prehistorical India (and for extension Musaev’s Middle Asia) has been elaborated upon with the help and collaboration of specialists from many different fields, including those of the Sino-Tibetan, Munda, and Dravidian ones. Musaev’s exact reference is: M. Satyanarayana et al., Affinity of Indian Languages, Delhi 1959. Unfortunately, I have been unable to locate this work anywhere. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS In Dravidian there are some borrowings from ancient Chinese, as well as from ancient Turkic. For instance, the “common North Indian” word h kur looks like a borrowing of ancient Turkic tegin, tegir ‘a host, the owner’. The Turks were the first Muslims to conquer North Indian [...] and what is most astonishing, a comparatively small group of words of Turkic origin penetrated into the Indian languages [...]. And even today you can come cross in Tamil [with] the Turkic word tanga meaning ‘gold’. This is by far the most revealing passage in Musaev’s paper, as it makes very clear what his methodology looks like. All the statements contained in the previous passage can be contested with no great efforts and should be taken as the starting point to understanding what will be exposed in the rest of the present contribution. The three words chosen by Musaev (and the Indian scholars quoted by him) belong to a very exclusive group of words which are scattered across the entire Eurasian continent (Wanderwörter), so to use them in order to demonstrate any kind of direct contact between two given populations far from each other is a vain exercise, because the potential borrowing routes could have covered the intermediate territories (and its languages). It is only a matter of time and patience to find out the borrowing routes they followed to end up in each language. Fortunately for us, the history of those words involved in Musaev’s statement are more or less well understood, so it can be shown immediately what the real scenario, or at least the most likely one, may have looked like. (1) The TL has tulukkan ‘Musulman; Turk’ & turukkan ‘native of Turkey; Muhammadan’ (IV: 1989a & 1982a), and in both entries is clearly stated that the etymology of these words goes back to Skt. Turuṣk - ‘Turks’ (pl.), cf. turuṣka‘Olibanun’ (sg.), after folk-etymology. The Sanskrit word arrived in Dravidian likely via Prakrit forms like turukka- (KEWA I: 515). If one takes into account that the name of the Turks traveled as far as Greece and it is logically attested in to Mongolian, then there cannot be any doubt about the direction of the loanword: [Turkish / Mongolian →] IndoAryan → Dravidian. (2) Sanskrit h kkura- ‘chief, man of rank, object of reverence, deity’ has been preserved in modern languages like Marathi h k r or Hindi & Nepali h kur ‘id.’. It corresponds to the modern Th k r = Tagore added to names to convey respectability and honorifics. The etymology of this word is unclear, and the best solution so far is that proposed by Mayrhofer: “[...] das Wort ursprünglich der Name eines mächtigen und darum geachteten Stammes war; dieser Name wäre dann wohl nichtarischer Herkunft, wie früh vermutet wurde” (KEWA I: 458). Any comparison with the Turkic title tegin is, to say the least, far-fetched, for the word has traveled considerably across Central Asia maintaining almost intact its original phonetic shape (TMEN 2922, Rybatzki 2006: 380-4 s.v. digin).21 It is very hard to believe that IA languages, perfectly capable of adapting the word without a change given the compatibility of the phonetic systems, 21 As far as tegir is regarded, to the best of my knowledge this is a verb form, derivate of teg- ‘to approach, come about, get (near)’ (vid. i.a. DTS 548a-b). Regardless of its etymological relation to the title tegin, it should not be quoted as a synonym (cfr. EDT 485 s.v. tegi:r ‘price, value’ ← aorist form of teg-). 49 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE would modify it so wildly. The sound changes one would have to contemplate to accept such an equation as Turkic → Sanskrit are too drastic. Thus, this is most likely a good instance of chance similarity. (3) Tam. & Mal. are the only languages where the word ta kam (neu.) ‘(pure) gold’ is attested to within the Dravidian family (TL III: 1710). It must be remembered that Mal. is the historical result of the independent evolution of the northern Tamil dialects branching off around 12th c. It is commonly accepted that this word comes from Sanskrit a ka- ‘gold’ (DEDR 3013, KEWA I: 456). However, it is more likely that it came via Persian tanga ‘golden coin’ ← Chag. tä kä, cf. Uyg. täñgä, Tat. täñkä, Uz. tanga ‘gold’, the last origin of which is controversial, since there are many possibilities, oscillating from Arabic to Tibetan (TMEN 2946). Despite the complexity of the whole picture, it is obvious that Tam. & Mal. words arrived via Sanskrit or Persian intermediaries. As is clear from these three cases, many words shared by Dravidian, Sanscrit and Turkic usually made their way through following the route Turkic → Sanskrit/Persian → (Southern) Dravidian. Other instances, however, may be easily explained as chance similarities. In the next section I will discuss the rest of the linguistic data provided by Musaev, this time his own etymologies. 3. “New” linguistic evidence22 It is necessary to underline from the outset that Dravidian serious etymological research is a rather young discipline in comparison to what the specialist faces in other fields. In the ‘50s of the twentieth century Emeneau denounced the situation in Sanskrit etymological treatises where any glimpse of Dravidian influence was dismissed always in favour of Sanskrit influence. Mayrhofer’s KEWA is actually the first non-Dravidian treatise to take up seriously Dravidian materials. In order to improve the philological approach to these languages, Emeneau proposed a list of criteria (diagnostic features) in order to recognize original Dravidian words in the Indo-Aryan lexicon: (1) a Sanskrit word without Indo-European etymology, (2) a wide currency of the etymon in Dravidian making a basic item, (3) a root derivation, (4) the earliest attestation in Tamil, i.e. antiquity, (5) the comparative lateness of appearance of the word in Sanskrit, (6) phonetic criteria, (7) semantic criteria (1967a[1954]: 164-70). They all are perfectly applicable to other fields by changing accordingly “Sanskrit” and “Dravidian”. In fact, Emeneau’s criteria are born out of common sense and they actually form the basis for responsible etymological research. Bearing this in mind, now I will analyze Musaev’s etymological proposals. I have tabulated the alleged cases of lexical contacts put forward by Musaev (vid. Table I). Numbers correspond to those in the table, and for the sake of clarity a brief cognate quotation will open every section. As far as the table is concerned, 22 50 In order to simplify typography, some graphic conventions typically used in Turkological studies, e.g. allophones of /k/ written as <q> after low vowels, <k> after high vowels, <γ δ> instead of <g d>, etc., will not be used in here unless they are strictly necessary. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS instead of reproducing verbatim his materials, I have just substituted some of the forms by those listed in authoritative dictionaries with not aim of completeness.23 As for the Dravidian languages is regarded, it seems that Musaev experienced some troubles in noting retroflex consonants, long vowels, etc., so I provide a much more comprehensive list of materials. A natural consequence of checking informed dictionaries is the possibility of taking a look at the whole picture, and not only to the restricted view Musaev decided to offer to the reader. Musaev’s surprising decision of using the old-fashioned comparative vocabulary of the so-called pañca dr viḍa b śegaḷu, lit. ‘the five Dravidian languages’, i.e. the five most important Dravidian languages: Tamil, Kanna a, Malayāḷam, Telugu and marginally Tuḷu, literally ripped off any possibility to observe what happened with other components of the Dravidian family (no less than 20 languages, maybe not so geopolitically notorious, but linguistically as important as the so-called “the five Dravidian languages”). There is no philological justification in Musaev’s decision, but only negative considerations, as we will see immediately. (1) 23 24 TAMIL ANAL ‘FIRE’ VS. TURKISH ALEV ‘FLAME’ VS. SANSKRIT ANALA ‘FIRE’. The page Musaev quotes from ÈSTJa actually draws the reader’s attention to the root */al-/ (more properly quoted in its extended, complete form, i.e. */āla/), an element used in the formation of colour names, and to the best of my knowledg never related to the other group of words meaning actually ‘flame’. Although Doerfer states that “Die Etymologie des Wortes ist schwierig” (TMEN, p. 358), it is commonly accepted that the most likely explanation for this word is the Persian borrowing option, which takes us to Sanskrit al tam ‘firebrand, coal’, Latin altar ‘altar’ & adoleō ‘burn a sacrifice’, or even Swedish ala ‘blaze, burn’ (Mallory & Adams 1997: 87). Thus, it makes little sense from the outset to establish any comparison with Dravidian. Obviousness aside, alaw could not be considered a word of Dravidian or Sanskrit pedigree, since sound correspondences like /-n-/ : /-l-/ and /-w/ : Ø are very serious obstacles. To account for them would require a entirely new battery of lexical comparisons, which as far as I know is not available.24 The word is not attested to in Old or Middle Turkic, so it must be regarded as a secondary, recent derivate. Actually, the word has been traditionally considered a loanword from Persian. On the other hand, Mayrhofer openly admits that this is likely a SDr. loanword in Sanskrit. Although it is well known that the suffix *-al is very productive in Dravidian, this fact does not help much, for we still do not know what is the root. The suffix *-al is attested to in every language except in Brahui (Rao 1971: 84-90). It is used mainly to form nouns from verbs, i.e. it is deverbal noun suffix, e.g. */āṭu/ ÈSTJa+EDT (occasionally TMEN), VEWT, DEDR and KEWA for Turkic, Dravidian, and Sanskrit, respectively. As we learn from Musaev, he made use only of ÈSTJa as well as some papers to elaborate his comparanda. The lack of contrasting sources is very surprising and without any doubt the reason for the failing of his proposals. Of course, the change * /l/ > /w/ is fairly common (see the very well studied case in Slavic historical linguistics, e.g. Flier 1983 with bibliography), but in this case there seems to be no scenario to apply it, at least without going into wild speculation. 51 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE ‘to move’ → */āṭ(ṭ)-al/ ‘moving’ > Ka.&Ta. -al, Tu. ḍ-el-e id. No verbal stems of the shape */an-/ are known in Dravidian fitting the requirements to be the base for such derivate as Ta. anala, leaving aside the suspicious nature of the final -a. Sanskrit anala has been traditionally analyzed as an-ala, with al- < IE */al-/ ‘to grow, bear’ after lateral dissimilation (Pokorny 1959.I: 26-7, SED 26b related it to IE */an-/ ‘to breath’). (2) (3) (4) (5) 52 TAMIL ALACU ‘TO BE DISTRESSED’ VS. TURKISH ALÏK ‘STUPID’ VS. SANSKRIT ALASA ‘LAZY’. Musaev must have made a terrible mistake here, because his Turkic reference is ÈSTJa I: 128, after which we are supposed to connect the meaning of his Turk[ish] alasa ‘shortish, weak’ (I have been unable to locate such form) with the root */al-/ ‘to take, grasp’. This is most shocking if we take into account that in p. 132 we are told the solution to the relationship of the materials quoted in Table under (2). As Sevortjan explains too, the most likely etymology for alas, and maybe too for alïg, could be */al-/ ‘below, down’, as in OTr altïn or Y alïn id. All in all, Musaev’s original intentions regarding this item remain unclear to me, for this is a clear case of Dravidian → Sanskrit loanword. As for the Sanskrit word, main etymological tools agree that this is a Dravidianism. Lubotsky (2001: 304) comments on the rarity of -sa- in the Indo-Iranian inherited lexicon. TAMIL KA U ‘SEVERE’ VS. TURKISH (DIAL.) KAT ‘HARD’ VS. SANSKRIT KA U ‘SHARP’. Semantically there is no way to connect Dr. and Turkic. It has been assumed that the Dr. words are Indo-Iranian loanwords, although the details still wait to be worked out. TAMIL K R ‘BLACK’ VS. TURKISH KARA ‘BLACK’ VS. SANSKRIT K LA ‘BLACK’. This case is by far the most interesting of all the proposed by Musaev due to its historical and cultural implications. From a linguistic point of view, however, there is not much left to discussion. The word is widely attested, both historically and synchronically, in Turkic (already common in the Kül Tegin, Bilga Kagan, and Toñukuk inscriptions, i.e. Runiform Turkic, ca. 6th-9th cc.) and Dravidian (Old Tamil monuments also register it frequently). It appears also in Mongolian and Tungusic. The easiest and more simple solution to account historically for all the attested forms is to assume a borrowing path as shown in the following chain: Dravidian → Sanskrit → Turkic → Mongolian → Tungusic. This sequence also respect all the well known borrowing routes already described in the literature: Dravidian borrowing into IA (the other way around is improvable, given the lack of internal explanation for the IA materials, e.g. a competing IE etymology; /l/ in the Sanskrit form is highly unusual, and it is normally taken as to be the nearest and best fitting equivalent of Ka. kaḍu ‘’blackness, black), Sanskrit borrowing into Turkic during pre-Buddhist times (the word is attested in Chuvash!), and Turkic borrowing into Mongolian due to the same reason. Mongolian and Tungusic loanwords reflect the most recent layer. TAMIL KU AM ‘CUP’ VS. TURKISH KUTU ‘BOX’ VS. SANSKRIT K AM ‘POT’. There is no dispute about the external (Dravidian) origin of the Sanskrit word. The Turkic forms, which ended up in Persian, Urdu and even Georgian, have not been etymologized so far, and I am unable to provide one proposal on this respect. I must admit that the similarities are tempting, but I have no answer for this. Given the cumulative evidence I am presenting in this paper about the fact that there Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS (6) 25 were no Turkic-Dravidian contacts, then I would assume that this is another case of chance similarity. Additionally, Musaev considers that the very same root present in Sanskrit and Dravidian also appears in Tam. kuti ‘hut, house, temple’ and Sanskrit ku i id. In this case, although the relationship of both words is well known in specialist literature (see EWA I: 362 with references), one is forced to assume here a Sanskrit loanword in Dravidian, for there is a fairly reasonable internal etymology for the Sanskrit term (from IE */kert-/ ‘to turn, roll, wind’, vid. Pokorny 1959.I: 584-5), while Dravidian only has Southern forms and no way to be account for, etymologically speaking, 25 both potential pieces of evidence to claim borrowing. In spite of Musaev’s confidence (“[e]vidently these two words have one and the same root element”, p. 172), it remains still unclear to me in what grammatical terms we would have to relate both set of forms beyond surface root similarities. I have no clue where root vowel length in Sanskrit k am comes from. TAMIL KARU ‘EMBRYO’ VS. KAZAKH QARÏ ‘SEED’. In quoting the materials for this item it is made obvious that Musaev lacks even the most basic acquaintance with the Dravidian languages, for he says that “[t]he Turkic word differs from the Dravidian counterparts through the absence of the final vowel. It seems that there exists a general phonetic law of presence/absence of final vowels within the groups of related languages, [...]” (p. 172-3). The “final” vowel in the Dravidian forms is an automatic, so-called “enunciative” vowel, which appears in some consonant-ending bases (for further details vid. i.a. Zvelebil 1970: 53-5, Bright 1975[1990], Krishnamurti 2003: 90-98). Dravidologists note this after a slashed bar, i.e. </u> or by superscript <u>, as I do in the present paper (its continuation in the historical languages comprises different realizations, like [ï] in Tamil, etc., see Bright 1975 for further details). Be that as it may, the only Turkic language involved in this comparison is Kazakh which is supposed to preserve the word qarï meaning ‘clan, tribe’. Musaev’s source is most likely Ïskakow, Sïzdïkowa & Sarïbaew’s brief etymological dictionary where under the entrance “ а I” they discuss several sentences, most of them proverbs and other popular expressions, concluding that the word under scrutiny may mean ‘clan, tribe’ indeed (1966: 122). Musaev’s remark on the historical value of folk-lore language is legitimate and from time to time there are contributions which remind us about the obligation of linguists and philologists concerning the necessity of paying more attention to it (see recently Shagdarsürüng 2005: 184-5 on Mongolic). Latin adagios like unus testis, nullus testis aside, this case reaches here a death point, because both semantics and phonetics are reasonable. However, the fact that the word is theoretically just attested in Kazakh should be enough to reject this comparison, at least as evidence of Turkic-Dravidian contacts. However, this word is nothing else but a semantic extension of Common Turkic */karï/ ‘old (woman or man)’ (see EDT 644, TMEN 31452, ÈSTJa V: 311-2, 314-6, DW(S) Unless we are open to accepting a relationship with verbal bases like Tam. k t i- ‘to come together, join, meet; gather’ (DEDR1882) with application of Krishnamurti’s Law (vid. i.a. Zvelebil 1970: 184), i.e. PDR. verbal */kūṭ-/ ‘to come together’ → nominal */kuṭ-i/ ‘place to come together’ > ‘hut, house’. 53 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE (7) (8) 54 169, 170). Indo-Iranian parallels may be brought into discussion: Pashto kaṛwasáy ‘great-grandchild’ is segmented *kar- + nwasáy ‘grandson’ (← Persian nab sa id) by Morgenstierne (2003: 40) who recognizes that the segment kar- is so unknown. I wonder what is the relationship of *kar- to Pashto kor- ‘house’ (related to Old Persian k ra- ‘people, army’ or Kurd bē (kas ) k r ‘without relatives’, note the semantic change ‘family’ > ‘house’, vid. Morgenstierne 2003: 39). Other Indo-Iranian forms, less problematic from a semantic point of view, would include those related to Persian k r- ‘to sow, till’ (< ‘to scatter’, cf. Avestic k raya- id.), from IE */(s)ker-/ ‘to spring, turn’ → Present Indicative 3rd SG */kr8-á-ti/ > Sanskrit kiráti ‘he strews’ (Pokorny 1959.II: 933-4, Bailey 1979: 53a, Morgenstierne 2003: 39). What relationship links all these forms remain unclear to me. However, note that the lack of cognates in Brahui (Dravidian) or Chuvash (Turkic) rather points out that those words are not very old. Dravidian and Turkic certainly are not to be related to each other, but Turkic and some Iranian forms may be actually, explaining for example the origin of Pashto *kar and related. TAMIL MEI ‘TRUTH; BODY’ VS. OLD TURKIC BOD ‘BODY’. The earliest Turkic records show clearly that the word contained originally /-d/ ~ /-δ/, a phoneme which, as is well known, only in later periods (ca. 10th-11th cc.) would have changed into /-y/. Consequently, this fact dismisses from the very outset any comparison with Dravidian forms, already in PDR. with */-y/. Musaev is aware of forms like Tuva or Khalaj, both with final obstruent. However, he seems not to be interested in explaining them out in relation to the rest of the forms showing /y/. For the change b- into m-, and viceversa, see Schönig (2002). TAMIL MA AI ‘A K. OF BOWL’ VS. CHAGATAY MANDAI ‘FOREHEAD’. Since this word is attested to in Turkic languages only from Chagatay onwards (namely, 15th c. as earliest dating), it has been traditionally assumed that it is a Mongolian loanword. Dravidian forms are restricted to South-Central languages, what makes already a good case for loanword. As is well known, the Moghul empire ruled the north of India for sixth centuries and for three centuries over the Brahmin Sultans of Deccan, so in that time many Persian and Arabic loanwords come into Dravidian. From the 15th c. onwards many words found their way into South Dravidian through Dakkhini Urdu (Krishnamurti 2003: 478). Musaev did not mention Skt. ma ḍa- ‘head’, already quoted in DEDR, or má ḍala- ‘shield, circle’. Although semantically all these forms fit very well, Sanskrit lacks a solid Indo-European etymology (but cfr. Lat. mundus ‘world’, etc.), so there is no way to reject the Dr. loanword option. As far as I know, no one has ever proposed any internal etymology for the Dravidian words farther than recognizing in -ai the very productive deverbal nominal suffix which incidentally we have seen already in this paper. We are left with *ma -, which could be actually a causative stem (< *ma -tt-), however there are no roots with fitting semantics. It is hard to conclude anything concerning these words without an in-depth monographic study. In any case, something is clear: Dr. and Sanskrit are related somehow, in the same way Turkic and Mongolic are, but there is no manner to link Dr.Sanskrit and Turkic-Mongolic. For one thing, the consonant clusters */-nd-/ against */-ŋl-/ (Turkic clusters are clearly secondary) cannot be derived one from Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS (9) (10) (11) (12) the other (in spite of Musaev’s comment that “[t]he phonetic correspondences are beyond reproach”, p. 173). TAMIL K NTAL ‘WOMAN’S HAIR’ VS. TATAR KÜLTE ‘BUN OF HAIR’. This is a very illustrative case of irresponsible comparative linguistics. The Dr. word is only attested in Tam. & Mal. and it does not appear in DEDR. These two facts already should raise some doubts as for its Dr. pedigree. The confirmation of such suspicion is given in TL, where additionally a third word is added to the comparison, namely Ka. k dal showing regular cluster reduction */-nt-/ > /-t-/. According to TL’s authors, the three Dr. words are lastly to be derived from Sanskrit kuntala, a derivate of kunta ‘lance, spear’, in its turn tentatively linked to other IE words like Lat. contus, Gk. ο τό , Latvian s ts ‘hunting spear’ (< IE */k@ent-/ ‘to stick, pierce’ → nominal */k@ont-/, vid. Pokorny 1959.I: 567, with irregular outcome of */k@/). The Turkic forms are usually regarded as Mongolisms, although it is not clear at all the way these forms came to be, for there are questions concerning details waiting for an answer. Be that as it may, it is notorious the presence of the Trk. consonant cluster */-lt-/ against Dr. & Sanskrit */-nt-/. TAMIL UCCI ‘CROWD OF HEAD’ VS. TURKISH UÇ-MAK ‘TO FLY’. I cannot help but wonder how Musaev expects us to understand the relationship between ‘crown of the head’ and ‘to fly’. Common Turkic */ūč/ ‘end, edge’, cf. Turkish uç id. (EDT 17-8, ÈSTJa I: 611-2, DW(S) 241) would be much more suitable, but yet, it is necessary to emply a good deal of semantic elaboration. I understand that the former is high with respect to the soil, and that something flying is by definition over the soil, but that does not mean to be “high” (otherwise to say “flying high” would be a non-sense, something which any native speaker of English would immediately deny). Thus, although there is an obvious formal similarity between the Dr. noun and the Trk. verb, the semantics are insurmountable. TAMIL NAMU U ‘LOWER LIP’ VS. OLD TURKIC AGÏZ ‘LIP’. Although the situation in Dr. seems to be highly complex, this is not the case. The presence of the initial /n/ in Tam. & Mal. against Ø in the rest of the languages is not awkward (Zvelebil 1970: 132-3) and must be described for several items. The same holds true for the alternation /m/ ~ /v/ attested both in Auslaut and Inlaut (Zvelebil 1970: 125-7). Mal. geminated /mm/ may be very well secondary, for this sequence is usually the result of assimilatory processes after derivation. Tam. uta u belongs to another set of unrelated words. Although both forms deserve still some comments (second “base” vowel /u/ ~ /i/, etc.), I think it is unnecessary to go deeper, since by now it must be clear that PDR. */namuṭu/ cannot be related to PTR. */agïŕ/ on any reasonable ground. It is my understanding that Musaev felt a sort of attraction after comparing Tel. avudu with, let’s say, Tat. awïz, but both forms are nothing else but the historical regular, systematic continuations of the proto-forms mentioned above. TAMIL RU ~ EYIRU ‘TOOTH’ VS. “TURKISH EZU ‘A CORNER OF THE MOUTH’”. There must be a misprint in Musaev’s Tel. eiguru, as it only exists as (c)iguru. The same goes for Mal. eyirŭ ‘mouth’. There is not much sense in marking the final vowel as short, since no long vowels can occur in a final position. In addition, this is the so-called automatic, enunciative vowel already commented upon. As 55 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE (13) (14) 56 for actual Mal. ekiru, it just means ‘tooth’. It is my understanding that Musaev’s “Turk. ezu ‘a corner of the mouth’” must be somehow related to the items enumerated in (11). Be that as it may, I have been unable to locate in any known dictionary the form quoted by Musaev. Räsänen & Radlov’s forms listed in Table 1 are equally problematic, as usually happens with the Finnish and German-born scholars’ quotations. Notwithstanding, I found much more compelling the comparison between these Dr. forms and those Turkic analyzed in (11). Common phonological processes aside (i.e. */VγV/ > /VV/), although the semantics are not enough narrow, the formal similarity is striking. However, lacking any motivation to assume direct contact (and answers to simple questions like the direction of the borrowing), I prefer to see this comparison as another (unfortunate) case of chance similarity. TAMIL TO AI ‘THROAT’ VS. KIRGIZ TANDAY ‘PALATE’. This is a self-evident Mongolian loanword in Turkic. The Mongolian internal etymology is transparent, namely the nominal base */taŋ-/, as in /taŋri/ ‘heaven, sky’ plus derivate suffix +lai (of unclear meaning, but certainly Mongolic, Volker Rybatzki, p.c.), whereas Turkic is opaque in derivate terms. Whatever the last origin of Mongolian */taŋ/ is (see recently Georg 2001), this word is Mongolic in its structure and thus should be regarded as a Mongolism in Turkic. Dr. presents /o/, and Turkic-Mongolian /a/, there is no trace of /ŋ/ in Dr. (in spite of being a very popular phoneme in those languages) and the semantics are far from being acceptable. TAMIL ARA KAI ‘PALM OF HAND’ VS. KIRGIZ ALAQAN ‘PALM (OF HAND)’. This item involves a body term. The lexical field has been extensively discussed in the literature, including the Turkic-Mongolian-Tungusic words, so the reader is entitled to consult those references to avoid here any useless repetition. As for Dravidian, the group of words under scrutiny turn out to be rather interesting. The etymology of Tam. a -kai and uḷḷa -kai ‘palm of hand’, the only two forms quoted by Musaev, is according to Tamil specialists akam+kai (TL I 23a s.v. a kai), where akam means ‘inside, mind, heart’. DEDR’s form ara kai is considered to be also the result of altering the very same compound (TL I 8b s.v. aka kai & I 173b s.v. ara kai), although the reasons and ways this has been achieved are not specified. The variant uḷḷa -kai (cf. Mal. uḷḷa -gai) seems to originate in the same semantic compound, i.e. uḷḷa(ka)m ‘mind, heart, inside’+kai. Same parallels exist for the sole of the foot: Tam. uḷḷa -k l or Tel. ara-k lu. It follows that the element containing the idea of hand or foot is the second member of the compound. The fact that Dravidian words have a very clear internal etymology should be enough to stop looking for “external” etymological explanations of their origin. The same holds true for the usually quoted form from Tocharian B alyiye ‘id’. This form may be shown to derivate regularly, with Tocharian A le-m ‘palms of hand (dual)’, from */h3elVn-/ ‘id’ > Old Irish uilen ‘corner’, Old Norse o7ln & Gothic aleina ‘ell’, Greek ὠ ‘forearm’ & Hesychius ὠ ό ‘elbow’ (vid. i.a. Mallory & Adams 1997: 176b, cf. ibid. also */h3elek-/ ‘id.’ > Old Church Slavonic lakъtь, Lithuanian úolektis, Armenian olok ‘shin, leg’). Of course, one just can compare all these forms and propose Nostratic common heritage, but that option is at the moment less than Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS (15) recommendable. I do not see any problem why we should not stick to the traditional etymology proposed already by A. Ščerbak et alii according to which this is a pretty regular, transparent derivate of */al-/ ‘to take, grasp’ (vid. i.a. ÈSTJA I: 127-8, DW(S) 259). Note additionally that the most common word for ‘palm (of hand)’ in Turkic is */āya/ (vid. i.a. ÈSTJa I: 100-1) and that the historical continuations of the derivate of */al-/ involve languages where */āya/ disappeared or were just replaced. KANNA A PO E ‘BELLY, PAUNCH’ VS. UZBEK BUT ‘GROIN’. There is no Tam. po -ai ‘belly’, but pocc-ai (DEDR 4478), actually a child variant of the expected **po -ai (cfr. Tel. bojja by side of po a) which however has survived in Mal. and other languages (there is a Tam. po -ai, but it means ‘blindness, blearsight’). The meanings of Kan. and Tul. in Musaev’s list are incorrect. One wonders if the expected **po -ai was altered to avoid potential confusions with forms like Ta. po a am, po a i, po alam, po ali ‘small bundle, parcel’ (+Ka., Te., G.), in its turns a well-known Indo-Aryan loanword, cf. Skt. po ala‘bundle, pocket’ (DBIA 277). The relationship of these Dravidian forms with Turkic has to be discarded on phonetic (retroflex consonant, different root vowel qualities and quantity) and semantic grounds. Summing up the previous etymological comments (see: [IN] = inherited, [LW] = loanword, [CS] = chance similarity, >→ = Dravidian loanword in Sanskrit, ←< = Sanskrit (Indo-Iranian/Indo-Aryan) loanword in Dravidian, Ø = no-contiguous relation, x = no-available materials for comparison): Ø ← ←< ← Ø? →? Ø Ø LN LN IN LN LN LN? x IN? ←< IN LN + CS Ø LN (← Persian) LN IN (→ Mong.) IN? IN? IN LN (← Mong.) IN or LN (← Mong.) IN Nature of the relationship LN + CS LN LN + CS LN LN (+ CS?) CS + LN? CS CS (+ LN?) x x CS Ø IN x x CS Dravidian Turkic (1) (2) (3) (4) (5) (6) (7) (8) IN IN LN IN IN IN IN IN? >→ >→ Ø >→ >→ Ø Ø Ø (9) LN Ø (10) IN IN (but IN? if compared to [12]) (12) IN Ø (13) IN (14) (15) (11) Sanskrit x x CS? Ø IN (but IN? if compared to [11]) LN (← Mong.) x x IN Ø IN x x IN Ø IN x x CS CS (+ Tocharian) CS 57 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE None of Musaev’s comparisons resists even a cursory philological examination. All of them can be arguably explained out according to more traditional borrowing routes or just as chance similarities. 4. Prevention is better than cure 4.1. There are a few cases when establishing borrowing routes with a certain degree of safety is a matter of great relevancy, as the items under study may be of great importance in describing the cultural history of the given region, in this case Eurasia. Thus, in English philology the word for rice (Oryza sativa) is a well-known instance of a Dravidian loanword, i.e. see for instance PSDr */(v)ar-i/ ‘rice’ or */(v)ari-ki/ ‘paddy, rice without husk’ (DEDR [215] > Tam. ari & arici, Tu. ari & akki, also Te. arise ‘a sweetmeat made of rice’, SDr+CDr */variñci/ ‘rice’ (DEDR [5265] > Gō. wanj , Ga. vasil.26 All etymological dictionaries provide the basics on it (whether the word came directly from Dravidian into English is a much more delicate and debatable question addressed to specialists). Generally, the vicissitudes concerning the spread of the word ‘rice’ across the world’s expanse have been more or less described, at least in specialist literature. The existence of different terms scattered across the Turkic linguistic map to name this grain includes Old Turkic & Uygur göröìč, Middle Turkic (Kāšγarī) tuturkan, Turkmen bürinč, Turkish pirinč, Tatar aryš, Yakut ir s ~ ris or Chuvash yraš and ris(ь), the (dialectal) multiplicity of forms displayed by those few languages is due to internal borrowings within the Turkic family, as well as other non-Turkic borrowings, e.g. Persian (e.g. š l → i.a. Uyg. šal) or even Indo-Aryan languages (Sanskrit vr hí or Afghan vriže → Persian birinǯ ~ gurinǯ → i.a. Khalaj birinǯ), folk-etymology, etc. However, the commonest term is ris and variants, documented in both (geographic) extremes of the Turkic world (Yakut & Chuvash). There is little doubt that the origin of these forms is the Russian word rožь ‘rye’ (see Räsänen in VEWT (26a), to be more precise some Old Eastern Slavic (*)rъžь) and ris ‘rice’ (both borrowed in Chuvash), then pushing the borrowing date as back as possible allows a more suitable scenario for some variants, e.g. Chuvash *araš > yraš, to be proposed27 The repetition of all these well known facts aims to make clear that the existence of properly understood loanword routes enables us to avoid ridiculous proposals as to say, for the sake of matter, that those Turkic words are actually remnants of direct contacts between PTR. and PDR. populations, instead of assuming that the very same Turkic words are recent loanwords from adjacent languages, all in spite of the fact of the phonetic and semantic similarities between Turkic and Dravidian words. In addition, this concrete case allows us to illustrate a very important point: in order to propose direct contact, we should first be able to respond to all the alternative scenarios involving non-direct contact, even more when, as in this case, the “alternative scenarios” are actually the most likeable. In the (pre)history of Dravidian languages there is a very well known sound change by which */ki/ yields 26 27 58 The bibliography on this specific item is very rich. For recent and/or extensive accounts see Southworth (1988: 659-60) and generally Witzel (2006). For further details on the etymology of particular items, see Stachowski (2008: 63-9). Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS after palatalization segments like /c/ ~ /s/ in some Southern & Central languages (Zvelebil 1970: 117-9), e.g. PDR. */kī(-)/ ‘pus, snot’ > SDr.: Ta. & Mal. c , CDr.: Kui s -v-enḍi, h -v-enji , Kol. s -m id (DEDR [1606]), etc. Thus, if Turkic words are actually the result of contacts with Dravidian populations, and those contacts took place presumably in the proto-language stage, then one would expect to see somewhere the preservation of P(S)Dr. */ki/. Since the Turkic words contain palatal segments, direct reflections of the Dravidian ones after solving PDr. */ki/ (for the sake of argument, cfr. SDr. */variñci/ and Turkmen bürinč), then it follows naturally that the Turkic forms are the result of much more later expansion of the word as a Kultuwort (for there is no way to directly compare SDr. and Turkmen!). Another piece of evidence which stands against direct contact between Turkic and Dravidian populations is the very same diversity of terms found among historical Turkic languages. Had the contact indeed taken place, one would expect to find a sort of historical continuity in the form of cognates descending from a common ProtoTurkic root. However, what we can describe is just the contrary: many different terms, sometimes shared with other Turkic languages, sometimes with non-Turkic languages. This, again, seems to point to the fact that all these words are just of recent distribution. In conclusion, we observe the journey of words through historically documented languages whose locus and momentum are very different one from another, i.e. fragmented. They never point to proto-languages, and when they actually do, the geographical and chronological problems are so insuperable that the whole proposal does not hold water. 4.2. Maybe not so important, from the global culture point of view as it may be the rice, the word for mule could be listed also as a promising case of TurkicDravidian contact. Spoiled by the proposals done by scholars like Musaev, someone could see a potential relation between Old Turkic katïr ‘mule’ and Old Tamil kutirai & Modern Tamil kutira ‘horse’. 4.2.1. The word katïr ‘mule’ (DTS 435b, ETD 604b, ÈSTJa V: 339-40, Eren 1999: 218a) is preserved among modern languages in Turkish & Nogay katïr, Azeri & Turkmen gatïr. It appears for the first time in the Karakhanid period, i.e. Middle Turkic period (10-11th cc. onwards). The first mention is recorded, as expected, in Kāšγarī’s dictionary (I 364, III 302). The second source comprises two passages in fully developed literary works. On one hand, the famous poem Kutadgu Bilig (11th c.). The text runs as follows (text after Rahmeti Arat 1947: line 5370 [38512], p. 534, English translation according to Dankoff 1983: 214): tir-il-d-i tümen mi talu köp titir yazï-da kalïn yond akur-da katïr gather-PASS-PST-3 ten.thousand thousand selected many (she-)camel steppe-LOC many horse stable-LOC mule-SG ‘thousands of choice she-camels are gathered [for you], herds of horses in the steppes and mules in the stables’ 59 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE On the other hand, kagatïr ‘mule’ (DTS 405b),28 a hapax legomenon appearing in the legend of the Oguz Kagan (according to the Uyghur manuscript preserved in Paris, 13th c.?, although the contents could be easily much earlier). The fragment to which the word belongs is the following one (text after Bang & Rachmati 1932: 18 [272-274]): [...] andag ulug ölük bargu düš-d-i, kim, yüklä-mäk-kä, kil-dür-mäk-kä ad, kagatïr, ud as-lïk bol-d-ï [...] PTC big dead property find-PST-3 that load-INF-DAT bring{come-CAU}-INF-DAT horse mule ox small become-PST-3 ‘[...] they found so many non-movable properties that horses, mules and oxen were not enough to carry and load.’ Bang and Rachmati explains kagatïr, which they correctly related to Kāsγarī’s katïr, as follows: “Das Wort ist weder türk. noch iranisch”, furthermore, they ask: “Woher stammt es?” (1932: 33). In answering that question, their etymological appreciation–not of Iranian origin–turns out to be partially incorrect. The state-ofart of the Turkic etymology is curiously offered by Bailey, a non-Turkologist but an Iranist, in his monumental dictionary of the Khotan Saka language (1979: 70-1 s.v. khaḍara ‘mule’, see also Eker 2009: 260-1): “From *xara-taka- ‘on one side ass’. Sogd[ian]. [...] γrt’rk *xarataraka- ‘mule’; it passed to Turkish qatïr, whence it was brought back to Oss[etic]. D[igoron]. qadir, I[ron]. qadyr, beside I[ron]. xärg-äfs ‘ass-horse’. The form is like Zor[oroastrian].P[ahlavī]. ’stl *astar, N[ew].Pers[ian]. astar from *assa-tara ‘on one side horse’, O[ld].Ind[ian]. aśvatara-.” One could add to the otherwise very appropriate resumé by Bailey that, as happened in Ossetian where the Iron dialect preserved both forms qadyr and xärg-äfs, New Persian also preserves both forms, namely astar and q ṭïr (TMEN 31395). Clauson (EDT 604a) argues for a semi-parallel semantic derivative involving Turkic kat‘side, layer’. The other feature of the mule is its endurance and stubbornness, so I wonder whether Old Turkic kat- ‘to become hard; dry’, katïg ‘hard’ and related forms could be actually the origin of katïr (note that this word also means ‘stubborn’ in some Turkic languages, e.g. in Gagauz and Crimean Karaim!). Unfortunately, such etymology is as fanciful as Clauson’s proposal, so I will not elaborate further on this. Last but not least, it is worth noting that the word appears in The secret history of Mongols in a couple of passages: as the name of a river in [§194] Qačir usun, lit. ‘water-mule’, and as a common plural word in [§274] qačidud, in this case a double plural, like in [§55] lausasut ← lausa ‘mule’ < 28 60 In the Oguz Kagan text as well as other documents written with Uyghur script the sequence <VγV> appears in several cases where historical languages have /VV/, e.g. [249] has <Dagam> ‘roof’ (cf. Trkm t m id, TMEN 2834), etc. Thus, in the case of katïr we would have to emend †kātïr. However, as Doerfer explains (TMEN 31395, p. 393), the historical continuation of this word into modern languages does not corroborate the emendation, cf. Trkm gatïr id. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS Chinese. It is indisputable that this is a Turkic loanword reflecting the regular sound change /ti/ > /či/ (Poppe 1955: 40), at a later stage re-introduced once again in some Turkic languages, e.g. Kazakh kašïr, Crimean Tatar kačïr, etc. De Rachewiltz follows Mostaert and translates qačir as ‘riding mule’ (2004.II: 1009). The Mongolian form would return later to some Turkic languages (TMEN 31395, pp. 392-3). 4.2.2. The etymology of Southern Dravidian */kutir-a/ ‘horse’ > OTa. kutirai, Ta. kudire, Ma. kutira (DEDR 1171, TL II: 991b s.v. kutirai (dial. kudira, kudure, kutira) ‘horse; twisting stick for making rope, timber frame for twisting cable; bridge of a stringed instrument; cock of a gun; crate for casks; gimnastic horse-bar’) has been discussed in several occasions. The most convincing explanation comes from Thomas Burrow’s pen (1972), where he explains that the word is a derivate from PDr *kuti- ‘to jump, to leap’ plus a very productive suffix *-(V)r- that like in Turkic languages, fulfills the function of nomina agentis (Rao 1971: 91-6). The semantics taking as the starting point one feature of the animal appears to be natural, if not more convincing, when compared with other well-known names of this animal. The most common Proto-Indo-European word to name this animal is *ék@u9os > Lat. equus, Skr. aśvas, Toch. B yakwe, Old Irish ech, Old English eoh ‘id.’, an adjectival form derived by means of the suffix *-u9o- from the verbal root *h1ek@- ‘to run’ (Mallory & Adams 1997: 273a-279b), then *h1ek@u9o- means ‘one which runs’ (see recently De Vaan 2009).29 As a semantic parallel for Dravidian, one can quote Old Norse hestr ‘horse’ as being related to Greek ί or Lithuanian šókti ‘to jump’, šóku ‘jump’ (< Proto-Baltic *š k-), all these forms in turn derived from PIE *k¤eh2k- ‘id.’ (Rix 20012: 319). As a matter of fact, the running horse is a very recursive topic in Tamil proverbs and riddles and the term appearing in all of them is always kutirai, e.g. kutirai ōṭ-a ōṭ-a, v 1 kurai-kir-atu. atu enna? atu ciyum n lum. horse run-INF run-INF tail shorten-PRE-3SG.NTR that what that needle thread While a horse runs and runs, its tail shortens. What is it? A needle and thread. (Kapp 1994: 143). Burrow considers that kutirai is the only Dravidian native word for horse, despite the fact that it is attested in Southern Dravidian languages alone; the rest of words for horse are well known or easily demonstrable to be loanwords.30 On the 29 30 The bibliography related to the topic is immense, but it is worthy to noting Hamp (1990), Rix (1994: 9-10) and the general (but not orthodox) views adopted in Mallory & Adams (1997: 273-9). Some other cases deserve still much more attention due to their unclear origins. For instance, (Old) Tamil ivuḷi ‘horse’ (TL I: 349b) is an isolated form, mysterious to some authors. TL’s authors relate ivuḷi to the verb ivar- ‘to rise in high, ascend’. Although one could argue that phonologically this etymology may raise some problems, as it semantically seems to fit perfectly not only the first meaning of ivuḷi, but also its second, 61 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE other hand, Parpola & Janhunen (2011) recently concluded that *kutirai may have originally been used for ‘wild ass’, since this is the only equid indigenous to the subcontinent. These authors consider that *kutirai was a general descriptive appellation of equids, but ‘horse’ is likely to have been a secondary reference.31 An alternative etymology is that of Levitt (2003: 9, 19, 2007: 20-1) who considers that Skt. kilkin, kilvin, kindhin (v.l. for kilkin), kuṇḍin (v.l. for kindhin) are related to Ta. kutira, Ma. kutira, Te. gurramu, and Kol. gurram: “[...] the Sanskrit forms are variants for one another. In Dravidian -k- and -v- are regular derivative suffixes, as well as -nt-. Here we have the common derivative suffixes -k-, -v-, -ntalternating in the Sanskrit forms in question, the forms with -nt- being a euphonic combination in force” (Levitt 2007: 20-1).32 It seems to be rather obvious that the Sanskrit forms, if related at all, should be regarded as late loanwords, for they lack any Indo-Iranian background, let alone Indo-European, diagnostics of that seems to be the fact that they are attested to very late in the huge Sanskrit corpus. Yet Levitt adds: “Also, we might note here that Kalki, Kalk , Kalkin, “The White Horse”, the name of the 10th incarnation of Viṣṇu yet to come, which incarnation is seated on a white horse with a drawn blazing sword for the final destruction of the wicked, the 31 32 62 less quoted, one: ‘Mango tree’. Of course, it is necessary to make an effort and to understand that the tree elevates, ascends, get higher as it grows, while about the horse it may be said that this animal jumps or plays by elevating its legs. Burrow thinks that Tamil ivuḷi (already present in old Tamil literature, e.g. Puranaanuuru or Sangam) and Brahui hull hides the truly word for horse in Dravidian. Burrow thinks that the coming of Tamil via the northwest part of India fits with the possibility of having one word for horse. For those who prefer to defend the nativeness of the Dravidians, located from the very beginning in the southern part of India, then they would have to explain the origin of such words, keeping in mind the fact that horse is an animal non native to India. It is too bad that Janhunen does not mention Turkish katır and related forms in his discussion on names for wild ass in Turkic, Mongolic, and Tibetan (Parpola & Janhunen 2011: 90-100, 111-114). Levitt (2007) proposes that PDra. */mā(v)/ ‘deer, elk’ (DEDR4780) is related (genetically!) to Chinese mǎ ‘horse’. Levitt does not quote any modern etymological tool for the latter (Schuessler 2007: 373, Matisoff 2003: 82, 249 [incidentally Proto-Lolo-Burmese]), otherwise he could have easily checked that: (a) is opinio communis to consider it a loanword from a Central Asian language, with Middle Korean mol and Japanese uma, because both the horse and the chariot were introduced into the Shang period in China around 1200 BC from the West, and (b) the word is univocally related to Proto-TibetoBurman */mraŋ/, cf. Written Burmese mra B and Old Tibetan rma (with metathesis) id. Therefore, the Dravidian forms could be easily considered also very old loanwords from the same Central Asian language. In fact, due to the main goal of this topic, Turkic and Dravidian languages are linked somehow because of the horse. In a rather imprudent way, Levitt decides to trust Winter’s archaeological and linguistic arguments about the antiquity and presence of horses in different parts of Africa and East-Central Asia (2005). As far as his linguistic skills are concerned, Winter (1989) explains in great detail the deficiency of his method and evidences. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS renovation of creation, and the restoration of purity, is the same Sanskrit word, of Dravidian origin as I have argued” (Levitt 2007: 21). However, Levitt does not explain what has happened with the different quality of root vowels in Dravidian and in Sanskrit. In the latter case, we have even two different vowels! Although Levitt seems to be aware of Burrow’s (1972) paper on the same topic, he decided to ignore some of the British Indologist’s proposals regarding some forms. For instance, Burrow considered that Telugu gurra-mu comes from a Prakrit ghoḍa(a)- ga-,33 and at the same time that neither this nor Sanskrit ghoṭa-ka- are related to SDr. */kuti(r)ai/. The lack to response for these and other problems render Levitt’s proposals very untenable. 4.2.3. All in all, it would be rather naïve to propose just a direct borrowing between Dravidian and Turkic, even though formally (they show identical consonantism, i.e. k-t-r :: k-t-r) and semantically (‘mule’ :: ‘wild ass > horse’) they seem to fit pretty well. On the contrary, one should not be surprised to come across with such a comparison, especially since it is an almost perfect match from a phonological and semantic viewpoint, much more salient as that of the great majority of Musaev’s items. The only author I am aware of is Chatterji (1965: 37), who thinks that Old Tamil kutirai must be linked somehow with Old Egyptian ḥ-t-r, Modern Greek γαδα ο ‘donkey’ and Turkish katır ‘mule’, as well as with “Hindustani” khaccar, xaccar. Such megalocomparisons must be understood in the framework of the Ancient Eastern Mediterranean theory defended by Chatterji, according to which there is a substratum or adstratum along the Mediterranean zone.34 However, consulting informed etymological dictionaries dealing with any of the languages alluded to by Chatterji should be enough to see that such comparisons does not hold water. Most of these words have clear internal etymologies. For example, Vycichl (1983: 315a-b) explains clearly that Old Egyptian ḥ-t-r ‘horse’ derives from the homonym verb base meaning ‘to join, to put together’. This word has survived as such in Coptic and there may have connections in Beja hat y ‘id’ and even Arabic ḥatar ‘to squeeze’. Then, he reconstructs */ḥt-3/ and/or */ḥt-y/, namely plural and feminine respectively, for the given languages. As for the 33 See DBIA [121] with additional Dravidian forms: Gadba (Oll.) gōṛa ‘horse’, Gonda kōḍa, Kui gōṛa, Kuwi gōḍa, Kurux ghoṛō, Malto goṛo id. (for further details regarding Kuṛux-Malto and the implications of this and other Indo-Aryan loanwords, see 34 Kobayashi 2009: 114). Chatterji (1965: 52) thinks that Skt ghōṭa / ghōṭaka “[...] is the word which unquestionably is also the last source of Tamil kutirai, Kannada kudure and Telugu gurra-mu (from an earlier *gudra-m)” and at least recognizes that Skt is the result of Prakrit developments. K. Menges offered several potential connections between Mesopotamia and the Altaic world. Along the same line, McAlpin linked Achaemenid Elamite forms to the Dravidian material which I am going to comment upon immediately (McAlpin 1981: 147-8, item D2). 63 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE Modern Greek form Chatterji quotes, it must be firstly emended for γ δα ο , and secondly, it is a well known derivate from the verb γ α ί ‘to scream’ (< *ὀγ α ί , cf. Classic Greek ὀγ ῶ α ), via *γ( ) δα ο (Andriotis 1951: 40b, 45a). The noun derivate γ δα ο is of recent coinage, being unattested to in Classical and Medieval texts. Last but not least, Hindustani khaccar ‘mule’ (Turner 1962-1966: 197 [3765]). The inescapable conclusion is that, once more, chance similarity must account for the similarity in the shapes of those words.35 As for the Turkic and Dravidian words, the potential link between them is not exempt from (insurmountable) problems. First of all, none of the forms agrees either in shape or in meaning altogether: Dravidian languages shows */u/ in the first syllable, whereas Turkic has */a/, and nothing in the material points out that the vowel quality would have been somewhat different in earlier stages in either group. Moreover, Dravidian means ‘horse’, whereas Turkic means ‘mule’,36 a non-trivial distinction with serious historic and cultural issues related to them. 37 Biological differences between the former and the latter are more than obvious and it would be rather hard to believe that those details were ignored in the process of borrowing. Most important from a linguistic viewpoint, both forms may be fairly explained by internal means, i.e. they have Turkic and Dravidian proper etymologies. Now, bringing into discussion the mass of lexical “evidence” quoted i.a. by Chatterji, I think that a responsible linguist cannot deny the obvious formal similarity between all these words, from Greek to Dravidian, showing consistently a shortness of the basic scheme */kVtVr/. This situation resembles powerfully that already described in many papers and books about one of the most famous Euroasiatic words for horse which can only be reconstructed, in similar fashion, as a consonant base, namely 35 36 37 64 Just out of curiosity, Chatterji is the one who addresses a message to Dravidian linguists in the pages of the same number of the International Journal of Dravidian Linguistics where Burrow published his paper on horse, asking for intense and good philological research in Dravidian languages (Chatterji 1972). As far as the Turkic languages are concerned, at is doubtless the common word for ‘horse’. Ščerbak (1961) dealt long time ago with horse names in Turkic languages. The most thorough study regarding the history of the horse taking into account Eurasian data is Kelekna (2009). Archaeological evidence for horses in South Asia before the advent of the Indo-Aryans is very scant, so it is very likely that the horse may have been introduced with those groups. Even so, early Dravidians may have used their own lexical resources to form a word for ‘horse’ rather than borrow it. The most complex question regarding this problem deals with history and archaeology, i.e. when was the horse introduced into India? It is indeed possible that the Dravidian constituted a first wave of central Asian tribes that came to Iran before the IA. In that case they could have known about the horse. One can even assume that the early testimony of the introduction of horse and camel from the Iranian plateau into Sindh (Pirak and Kachi plain in western Sindh) is due to the Dravida (c. 1700 BCE). Be that as it may, this is of no relevance to our study, for the Dravidian word means ‘mule’. It is a very typical mistake of long-range studies to mess animal species and deal carelessly with botanic and zoological data. Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS */mVrV/, allegedly attested even among Indo-European words.38 In a state-of-theart paper about the linguistic distribution of this word across Eurasia, Janhunen comments that “[o]f course, we do not know whether there was a direct linguistic contact between Pre-Proto-Mongolic and Pre-Proto-Koreanic. Perhaps more probable, there was a chain of borrowings involving an unknown number of unidentifiable intervening languages. It is however, important to note that Mongolic *morï and Koreanic *morV clearly represent a single original word shape from which Tungusic *murin and its presumable Para-Mongolic source are separated by a secondary vowel development” (1998: 418). I think that “Euroasiatic” */kVtVr/ must be approached in identical terms. Both are Kulturwörter and Wanderwörter whose last origin and more recent fate is lost for ever due to the complexity of the material documentation as well as our incapability to recover earlier–much earlier– stages of the languages and peoples involved, making it impossible to set up a more or less plausible scenario. Notwithstanding, what we can do of course is to establish a working hypothesis based on the reliable date we have at disposal. Thus, from the historical account provided in §2.2.2, we can make several observations which actually seem to favor the borrowing of the Turkic and Dravidian words: 1) 2) 3) In Turkic the word appears attested only in 11th c., just when Turkic-Indian contacts seem to flourish; The Dravidian word is only attested in Southern languages, thus it is legitimate not to assume that the word is old. This fact makes much more credible the option of a recent loanword; The languages to be considered as the main source of loanwords are those belonging to the common soldiers of Muslim armies. It is not too difficult to locate in the lexical stock of those soldiers a word for ‘horse’ or even ‘mule’. Unfortunately, I do not see how this situation may be corrected for good, so for now I think that the most reasonable solution is to conclude, with Janhunen who already considered PIE */márkos/ ‘horse, mare’ and “Euroasiatic” */morV/ the result of just chance similarity (see generally Janhunen 2007, 2010), that our “Euroasiatic” */kVtVr/ may be a mixture of inheritance, on one side, and partially chance similarity, on the other, blurred by the course of years. It is my understanding that this is the most sincere answer to the etymological inquiries on these words.39 38 39 Mallory & Douglas (2006: 141) illustrates that for Indo-Europeanist the option of the Euroasiatic loanword is also well-known. Mikhailova (2007: 4-9) has recently proposed that Celtic mark-os ‘saddle horse’ could be actually of Scythian origin, this being in its own turn an “Altaic” borrowing (everything inserted in a rather naïve Nostratic framework). Unfortunately, the philological argumentation does not exist out of Celtic, and the lack of the most remarkable works on the question should suffice to keep a skeptical position towards Mikhailova’s conclusion. Curiously enough, */mVrV/ has been quoted as a typical Nostratic lexical item from IlličSvityč’s earliest works on the matter, while to the best of my knowledge */kVtVr/ has been passed in silence. 65 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE A potential line of research, though dangerous given its nature, would insert both horse-like terms into sound symbolism and/or onomatopoeia scenarios. This phenomenon is behind the origin of many zoonyms to be identified crosslinguistically, e.g. sequences of the sort /kr xr gr/ are usually and legitimately related to bird-names (see an enlightening general discussion in Liberman 2010). However, the nature of expressivity needs clarification, otherwise “sound symbolism” becomes a lame excuse from the researcher’s excuse not to admit the unknown origin of a given term. In the case of bird-names it seems obvious that those sound sequences are an attempt at rendering the singing or screaming made by the bird. I cannot help but wonder what would be the reasoning for */k-t-r/ or */mr(-C)/, which actually do not evoke noises made by horses or mules. Are the crosslinguistic distribution and the presence of sounds like /r m k/ misleading us? 5. Conclusions th after 13 c. ← (→) → Turkic 8th-12th cc. → (←) Indo-Iranian before 7th c. Dravidian The main goal of this paper was to discuss some alleged evidences for the existence of Dravidian-Turkic-Sanskrit lexical contacts offered by the Turkologist K. Musaev. In doing so, I have tried to bring into the picture all the necessary materials and mention the most important considerations, from the Turkic, Dravidian and in smaller degree, Sanskritic fields in equal conditions. Musaev argued that Dravidian-Turkic contacts took place in proto-language times. However, as we have seen, there is no way to set up a historical scenario where both communities could have come into contact, if not altering a so far rather reasonable picture of archeological and linguistic prehistory of Central Asia (not just India!). Moreover, the linguistic evidence presented by the Russian scholar is not convincing at all. Most of the comparisons can be easily explained as recent borrowings from Indo-Iranian languages into Sanskrit and from there into Turkic, or just as chance similarities. The philological analysis of other potential comparisons also forces us to conclude that the nature of the similarities of those lexical items is fortuitous or the result of borrowing. Put it in other words: Dravidian and Turkic could eventually share similar vocabulary, but always via Indo-Aryan or IndoIranian languages as main donors. From a Turkic perspective, the chronology and direction of such borrowings must resemble what is described briefly in the following chart: Thus, Turkic alasa is as much Dravidian as English rice is. Even if the conclusion of this paper could have been already clear and/or obvious in the mind of most Turkologists and Dravidologists, surprisingly the scientific discussion of it did 66 Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS not find its way into the press. Some ideas deserve to be condemned in the same degree as the lack of its discussion. To conclude I would like to add that this paper has demonstrated, or at least I hope so, that Musaev’s proposal is unconvincing. However, the question of the existence of real Dravidian-Turkic contacts cannot be answered, basically because we lack the necessary data. Based on what we have seen in this paper as well as on what we already knew after decades of studies in the different fields, such contacts are most likely unrealistic. Of course, it goes without saying that the research of these proposals usually helps us to understand the historical (and linguistic) nature of the entities we are dealing with, but they do not allow us to reach more or less definite answers. Quite the contrary, it settles new, more complex questions. Emeneau, in a response to a paper by Prince Peter the Great on probable SumerianToda cultural and linguistic contacts, said that “[...] historical contact [...] between Sumerian culture and the remote ancestors of the Todas is something that cannot be either proved or disproved without vastly more evidence than we are likely ever to have [...]” (1967a[1953]: 61). In the case of the Dravidian-Turkic-Sanskrit situation it is much more positively bearable, and we can conclude with certain sureness that no known Dravidian-Turkic contacts are to be found, and those which are suspicious of being an evidence, can be easily accounted for as typical cases of Dravidian-Sanskrit-Turkic borrowings. 67 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE TABLE* Turkic-Dravidian-Sanskrit materials (1) (2) Dravidian SDr: Tam. anal ‘fire, heat (as of fever)’, Mal. anal, Kan. DEDR327 analu ‘id.’ [PDr */ala-/ ‘to get tired’ >] SDr: Tam. ala-c/u ‘to suffer, be distressed’, Mal. ala-sal ‘agitation’, Ko. ala- ‘to crave’; CDr: Kol. ala-y- ‘to DEDR236 become tired’, Pa. ala-c ‘illness’; NDr: Ku. al-g ‘without energy’ Turkic SW: Trk. alev, Trkm. (dial.) & SE: Uz. & NW: Nogay alav ‘flame, heat’ [PT */al-/ ‘(to become) weak, bad, short’ (→ */al-ïg/ ‘weak, inferior; fool’) >] SW: Trk alïk, NW: Tat. alama, Kaz. ala ‘unattentive’, SE: Uyg. alaq id., NE: Tuva ala ‘doubt’, etc. Cfr. Trk (dial.) alas ~ alïs ‘fool’ Sanskrit TMEN 31354, ÈSTJa I: 126 ÈSTJa I: 145-6, TMEN 2116, VEWT 16-7. ÈSTJa I: 132 [PDr */kaṭu/ ‘bitter, sharp’ >] (3) SDr: Tam.+Mal. kaṭ/u ‘severe, tense, cruel’; CDr: DEDR1135 Kol. keḍed ‘severe’, Nai. kaṛ/u ‘bitter’; ODr: Br. xar-ēn ‘bitter’. (4) [PDr */kār/ ‘blackness’ >] SDr: Tam.+Mal.+Kan. k r ‘blackness, darkness’; CDr: Nai. kar-en ‘black’. DEDR1278 [PSDr */kār/ ‘black’ >] SDr: DEDR 1494 Tam. k r, Kan. k r, k ḍ/u ‘blackness, black’, Tu. k ri ~ k ḷi ‘blackish’. [PSDr */kuṭ-am/ ‘(water)pot’ (5) (6) (7) (8) >] SDr: Tam. kuṭam, Te. kuḍ- DEDR1651 aka ‘cup, bowl’. [PSCDr */karu/ ‘foetus, embryo, seed’ >] SDr: Tam.&Mal. karu, Tel. karuvu, DEDR1279 Ko. karv, To. kef id., Go. garba ‘egg’; CDr: Pa. kerba, Ga. karba id. [PSCDr */may/ ‘body, person’ > ] SDr: Tam. mei DEDR5073 ‘truth, reality, body’, Mal. mai ~ mei ‘body, person; truth’, Kod.&Tul. mai ‘id.’; CDr: Tel. meyi ~ mē ‘body, side, DEDR4704 manner, method’. Cfr. Pengo mai ‘breast, teat’. [PSDr */manḍ-ai/ ‘bowl, jar; head, skull’ >] Tam. maṇṭai ‘mendicant’s begging bowl, DEDR4682 earthen vessel, head, skull, cranium, brainpan, top portion 68 [PT */kat-ï/ > ] OTr qat-ïg, SW: Trk (dial.) kat, NW: Kaz kattï, NE: Y&Dol. kït -nak; Chv. xïdŭ ‘hard, firm’ [PT */qara/ >] OTr kara ‘black’, SW: Trk & NW: Kaz., Nogai kara, SE: Uz. qore, Uyg. qara, NE: Y xara, Dol. kara; Chv xora ~ xura id. Mongolian: WMon. & MMon. qara, K & B xara, D xar(a) ‘black’ Tungusic: [*/karā/ >]: Ewenki & Ewen kar (-), Nanay kar- in karkaj, Orok kar w, Manchu qara ‘black’ SW: Trk. qutu ‘box’, (dial.) ‘pot’, Trkm gutï, SE: NUyg. quta ‘can’, Uz. qutï id., NW: Nogay qutïq (+Dim.) ESTJa V: 334-5, DW(S) 170 anala ‘fire’ KEWA I:33 alasá ~ lasa ‘lazy, tired’ KEWA I:55 kaṭú ‘sharp, harsh’ < */kr8t-ú-/, see kṛntáti, cf. Lith. kartùs ‘bitter’ KEWA I:203 k la ‘darkblue, black’ KEWA I:203 k ṭam ‘pot’, perhaps related to kuṭa ‘unhorned’ KEWA I:251 EDT 643-4, DTS 422-4, TMEN 31440, ÈSTJa V: 286-9 KW 168 SSTM I:379-80, MT 99 TMEN 31569, cf. EDT 596 NW: Kaz qarï ‘genus, seed’ Ïskakow, Sïzdïkowa & Sarïbaew (1966: 122) [kiráti ‘strew’] [PTr */bod/ ‘body; tribe, clan’ >] OTr bod (Orkhon), NE: Tuva bot; NW: Nog.&Kaz. & SW: Az.&Trk boy, Tat. buy id.; Chv pü; Khalaj bod id. ÈSTJa II: 176-8, EDT 296-7, DTS 108 s.v. boδ, 110 s.v. boj, TMEN 2812 — Chag. mandai ‘forehead’, NW: Kaz. & SE: Uz. ma dai, NE: Y ma nai id. ← MM ma lai id. VWET 327, TMEN 1369. KW 257, TMEN maṇḍa‘head’ ~ máṇḍala ‘shield, circle’ (?) Pokorny 1959.II: 933-4 KEWA 294 Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS as of palms, a standard of measure’, Ma. maṇṭa ‘skull’, Ka.&Ko. maṇḍe ‘head’. (9) (10) Tam. & Mal. k ntal ‘woman’s hair’ TL II: 933 [PSCDr */uc(c)-i/ ‘crown of head’ >] SDr: DEDR579 Tam.&Mal.&Kan. ucci ‘id.’; CDr: Tel. ucci ‘id.’ [PDr */namuṭu/ ‘lip’ >] SDr: SW: Trk küleş, Trkm kültem ‘sheaf, bun of hair’; NW: Tat. külte ‘bun of hair’. Cfr. MM küli- ‘to bind’, Klm küle-, etc. [PT */uč-/ >] OTr. & SW: Trk & SE: Uz. uč, NE: Y&Dol. uhun- ‘to float’, NW: Kaz.&Nogai uš- ‘to fly’. ÈSTJa V: 139-40. KW 245 EDT 19, ÈSTJa I:612-3, DW(S) 241. kuntala ‘the hair of the head’ ← kunta ‘spear, lance’ SED 291a — Tam. namuṭu ~ namiṭu ‘lower lip’, Mal. ammiṭṭam, Kan. avuḍu ‘id.’; CDr: Tel. avuḍu [PT */agïŕ/ ‘mouth; lip’ >] OTr agïz ‘lip’, NW: Tat.&Kaz. awïz, Bashkir awïδ, SE: Uz. ogiz, NE: Tuva s, Y&Dol. uos ‘lip’; Chv s’ăvar ‘mouth’. odaḍu ‘lip’. [PSCDr */ekiru/ = [eγiru?] ‘gums’ >] SDr: Tam. ru ‘id.’ ← eyiṛu ‘gums, tooth, tusk of elephant’, Mal. ēri ‘id.’ ← DEDR554 ekiru ‘tooth’; CDr: Tel. (c)iguru ‘the gums, the tender part under the nail’. [PDr */tonḍ-ai/ ‘throat’ >] NW: Kaz. izəü, Balkar ezǖ ‘Brustschlitz, Aufschlag ( , p а)’. Cf. Kazakh izǖ ‘ein Bustlatz’ avu(ṇ)ḍu ‘lower lip’, Tul. (11) DEDR3596 ‘under lip’; NDr: Kur. lauṭō DEDR606 ‘lip’. SDr: Tam.&Mal. utaṭu, Kan. (12) SCDr: (13) (14) Tam. Mal. NDr: Mal. tōnṛe ‘id.’. [PCSDr */araŋ-kai/? >] Tam. ara -kai, Tel. ara-cēyi, Kol. DEDR310 ran-kei ‘palm of hand’. [PDr */poṭ(ṭ)-a/ ‘belly’ >] SDr: Kan. paunch’, (15) toṇṭai, toṇṭa, Ka. dçṇḍe, Kod. toṇḍ-e DEDR3498 ‘id.’, Kūi tōtro ‘throat (lower)’, Pengo ṭoṭra ‘neck’; poṭṭe Tul. ‘belly, poṭṭŒ ‘pregnancy’, CDr: Tel. poṭṭa DEDR4494 ‘id.’, NDr: Kur. poṭṭ ‘bowels, entrails’, Mal. puṛa ‘id.’ EDT 91, ÈSTJa I: 813, DW(S) 245. — VEWT 53 Radlov (1893: 15423) — NW: Krg. tanday, SW: Trk. ta lay ‘palate’. Cfr. Mongolian ta lai > ta nai, Klm ta nää ‘id’, etc., see (8) VEWT 4612. KW 379 — NW: Krg., Kaz. & SE: Uyg. & SW: Trk (dialectal) alaqan ‘palm (of hand)’ ÈSTJa I: 133, VEWT 15. — [PT */būt/ ‘thigh; leg, foot’ >] OTr. (Uyghur) but ‘thigh’, SE: Uz. but ‘groin’, Uyg. put, NE: Tuva but ‘leg’, Y&Dol. b t, NW: Kaz.&Nogay but ‘thigh’ DTS 129, EDT 297, ÈSTJa II: 280-2, DW(S) 67. — * I do not offer exhaustive quoting neither of languages nor of sources (some items are well-known), but only those which may prove more useful and self-evident. In presenting the materials I follow the traditional classification of both families (vid. i.a. Johanson 1998: 82-3, Krishnamurti 2001: 381, 2003: 19). 69 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE Abbreviations 1,2,3 Az. Br. CAU Chag. Chv DAT dial. Dol. G IA IE INF Ka. Kaz. Klm Ko. Kol. Krg. Kur. LOC Mal. MM Nai. NE NOM person Azerbaijanian Brahui causative Chagatay Chuvash dative dialectal Dolgan Gōṇḍi Indo-Aryan Indo-European infinitive Kannaḍa Kazakh Kalmuck Koṇḍa Kōlāmī Kirgiz Kuṛux locative Malayāḷam Middle Mongolian Nāiki Northeastern nominative NTR NUyg. NW OIA OTr Pa. PCDr. PDr. PNDr. PRE PRT PSDr. PST PTr. SE SG Skt. SW Tam. Tat. Te. Trk Trkm Tu. Uyg. Uz. Y neutrum New Uygur Northwestern Old Indo-Aryan Old Turkic Parji Proto-Central Dravidian Proto-Dravidian Proto-Northern Dravidian present particle Proto-Southern Dravidian past Proto-Turkic Southeastern singular Sanskrit Southwestern Tamil Tatar Telugu Turkish Turkmen Tuḷu Uyghur Uzbek Yakut References DBIA = Emeneau, M. & Th. Burrow (1962): Dravidian Borrowings from IndoIranian. Berkeley & Los Angeles: University of California Press. DEDR = Burrow, Th. & M.B. Emeneau (19842): A Dravidian Etymological Dictionary (Revised edition). Oxford: Clarendon Press. DTS = Nadeljaev, V.M., Nasilov, D.M., Tenišev, È.R. & A.M. Ščerbak (1969): Drevnetjurkskij slovar’. Leningrad: Nauka. DW(S) = Stachowski, M. (1993): Dolganischer Wortschatz. Kraków: Księgarnia Akademicka. (+ Dolganischer Wortschatz. Supplementband. Kraków: Księgarnia Akademicka, 1998). EDT = Clauson, G., Sir (1972): An Etymological Dictionary of Pre-13th-Century Turkish. Oxford: Clarendon Press. 70 Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS ÈSTJa = Sevortjan, È.V., ed. (1974-2003-): Ètimologičeskij slovar’ tjurkskix jazykov, vols. 1-7-. Moskva: Nauka. EWA = Mayrhofer, M. (1986-2000): Etymologisches Wörterbuch des Altindoarischen, 3 vols. Heidelberg: Carl Winter. KEWA = Mayrhofer, M. (1956-1976): Kurzgefasstes etymologisches Wörterbuch des Altindischen. Heidelberg: Carl Winter. KW = Ramstedt, G. J. (1935): Kalmückisches Wörterbuch. Helsinki: Lexica Societatis Fenno-Ugricae. MK = Dankoff R. & J. Kelly, eds. (1982-1985): Maḥmūd al-Kāšγarī’s Compendium of the Turkic Dialects. Dīwān Luγāt at-Turk, 3 vols. Cambridge: Harvard UPress. MT = Doerfer, G. (1985): Mongolo-Tungusica. Wiesbaden: Otto Harrassowitz. SED = Monier-Williams, M. (1899): A Sanskrit-English Dictionary. Oxford: Clarendon Press. TL = Vaiyapuri Pilla, S., ed. (1982[1924-36]): Tamil Lexicon, 6 vols. Madras: University of Madras. TMEN = Doerfer, G. (1963-1975): Türkische und mongolische Elemente im Neupersischen, 4 vols. Wiesbaden: Franz Steiner. SSTM = Cincius, V. I., ed. (1975-77): Sravnitel’nyj slovar’ tunguso-man’čžurskix jazykov, 2 vols. Leningrad: Nauka. VEWT = Räsänen, M. (1969): Versuch eines etymologischen Wörterbuchs der Türksprachen. Helsinki: Suomalais-Ugrilainen Seura. AALTO, P. (1971): «Iranian contacts of the Turks in Pre-Islamic Times». In: L. Ligeti (ed.), Studia Turcica, 29-37. Budapest: Akadémicai Kiadó. ALONSO DE LA FUENTE, J. A. (2007): «Some thoughts on Etruscan *maθ‘(honeyed) wine’». Historische Sprachforschung 120(1/2), 305-10. ANDRIOTIS, N. P. (1951): Dictionaire étymologique du grec moderne. Athènes: Institut Français d’Athènes. ANDRONOV, M. S. (1971): “Comparative studies on the nature of DravidoUralian parallels: A peep into the prehistory of language families”, In: R.E. Asher (ed.), Proceedings of the Second International Conference-Seminar of Tamil Studies, vol. 1, 267-77. Madras: Christian Literature Society. BAILEY, H. (1979): Dictionary of Khotan Saka. Cambridge: Cambridge UPress. BANG, W. & G. R. RACHMATI (1932): Die Legende von Oghuz Qaghan. Berlin: Akademie der Wissenschaft (SPAW). BRIGHT, W. (1975[1990]): « The Dravidian enunciative vowel». In H. F. Schiffman & C. M. Eastman (eds.), Dravidian Phonological Systems, 11-46. Washington: Institute for Comparative and Foreign Area Studies, University of Washington. [Revised in W. Bright (1990): Language Variation in South Asia, 86-117. New York and Oxford: Oxford UPress.] BRYANT, E. & L. PATTON, eds. (2005): The Indo-Aryan Controversy. Evidence and inference in Indian history. London & New York: Routledge. 71 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE BURROW, T. (1944). «Dravidian Studies (IV): The body in Dravidian and Uralian», Bulletin of the School of Oriental and African Studies 11(2), 32856. —. (1972): «The Primitive Dravidian Word for the Horse», International Journal of Dravidian Linguistics 1(1), 18-25. CALDWELL, R. (19133[1998]): A Comparative Grammar of the Dravidian or South Indian Family of Langauges. New Delhi and Madras: Asian Educational Services. CAMPBELL, L. (1998): Historical Linguistics. An Introduction. Massachusetts: The MIT Press. CHATTERJI, S. K. (1965): Dravidian. Annamalainagar: Annanmalai University. —. (1972): «Address to the Dravidian linguists», International Journal of Dravidian Linguistics 1(1), 1-17. DANKOFF, R. (1983): Yūsuf Khāṣṣ Ḥājib’s Kutadgu Bilig (Wisdom of Royal Glory). Chicago & London: The University of Chicago Press. DE RACHEWILTZ, I. (2004): The Secret History of the Mongols, 2 vols. Leiden: Brill. DE VAAN, M. (2009): «The derivational history of Greek híppos and hippeús», Journal of Indo-European Studies 37, 198-213. EKER, S. (2009): «Divanü Lugâti’t-Türk ve İran Dillerinden Kopyalar Üzerine I», International Journal of Central Asian Studies (Festschrift für Prof. Tâlat Tekin) 13, 233-84. EMENEAU, M. (1967a[1953]): “The Todas and Sumeria-A hypothesis rejected”. In: M. Emeneau, Collected Papers, 61-3. Annamalainagar: The Annamalai University. —. (1967a[1954]): «Linguistic prehistory in India». In: M. Emeneau, Collected Papers, 155-71. Annamalainagar: The Annamalai University. ERDAL, M. (1993): «Around the Turkic ‘Apple’», The Journal of Indo-European Studies 21(1/2), 27-36. EREN, H. (1999): Türk Dilinin Etimolojik Sözlüğü. Ankara: Bizim Büro Basım Evi. FLIER, M. S. (1983): «The Alternation l ~ v in East Slavic». In: M. S. Flier (ed.), American Contributions to the Ninth International Congress of Slavists (Kiev, September 1983), vol. 1 Linguistics, 99-118. Columbus (OH): Slavica. GEORG, S. (2001): “Türkisch-mongolisch tengri ‘Himmel, Gott’ und seine Herkunft”, Studia Etymologica Cracoviensia 6, 83-100. —. (2008): «Yeniseic languages and the Siberian linguistic area». In: A. Lubotsky, J. Schaeken & J. Wiedenhof (eds.), Evidence and Counter Evidence. Essays in Honour of Frederik Kortlandt, vol. 2, 151-68. Amsterdam & New York. GOLDEN, P. (1992): An introduction to the history of the Turkic peoples. Wiesbaden: Harrassowitz. —. (1998): «The Turkic Peoples: A Historical Sketch». In: L. Johanson & É.Á. Csató (eds.), The Turkic Languages, 16-29. London & New York: Routledge. 72 Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS —. (2005): «Some Thoughts on the Origins of the Turks and the Shaping of the Turkic People». In: V. Mair (ed.), Contact and Exchange in the Ancient World, 136-57. Honolulu: University of Hawai’i Press. HAMP, E. P. (1990): «The Indo-European Horse». In: T. L. Markey & J. A. C. Greppin (eds.), When Worlds Collide. Indo-Europeans and Pre-IndoEuropeans, 211-26. Ann Arbor: Karoma. HOCK, H. (1975): «Substratum influence on (Rig-Vedic) Sanskrit?», Studies in Linguistic Sciences 5(2), 76-125. ÏSKAKOW, A., Sïzdïkowa, S. & Š. Sarïbaew (1966): Kazak tiliniŋ kïskaša etimologijalïk sözdigi. Almatï: Gïlïm. JANHUNEN, J. (1998): «The horse in East Asia: reviewing the linguistic evidence». In: V. H. Mair (ed.), The Bronze Age and Early Iron Age peoples of Eastern Central Asia, vol. 1, 415-430. Washington, D.C.: The Institute for the Study of Man. —. (2007): «Typological expansion in the Ural-Altaic belt», Incontri linguistici 30, 71-83. —. (2010): «Reconstructing the linguistic map of Prehistorical Northeast Asia», Studia Orientalia 108, 283-305. JOHANSON, L. & Ch. BULUT, eds. (2006): Turkic Iranian Contact Areas. Historical and linguistic aspects. Wiesbaden: Harrassowitz. JOHANSON, L. (1998): «The History of Turkic». In: L. Johanson & É.Á. Csató (eds.), The Turkic Languages, 81-125. London & New York: Routledge. KAPP, D. (1994): «A Collection of Jaffna Tamil Riddles from Oral Tradition», Asian Folklore Studies 53, 125-49. KELEKNA, P. (2009): The Horse in Human History. Cambridge: Cambridge UPress. KOBAYASHI, M. (2009): «Indo-Aryan Loanwords and the Prehistory of Kurux and Malto». In: K. Yoshida & B. Vine (eds.), East and West. Papers in IndoEuropean Studies, 111-131. Bremen: Hempen Verlag. KRISHNAMURTI, Bh. (2001): Comparative Dravidian Linguistics. Current Perspectives. Oxford: Oxford UPress. —. (2003): The Dravidian Languages. Cambridge: Cambridge UPress. KUCZKIEWICZ-FRA , A. (2001): «Turkic in India and its elements in Hindi», Studia Turcologica Cracoviensia 8, 43-8. KUIPERS, F. B. J. (1967): «The genesis of the linguistic area», Indo-Iranian Journal 10, 81-125. LEVITT, S. H. (2003): «New Dravidian Etymologies for Sanskrit Words», International Journal of Dravidian Linguistics 32(2), 7-22. —. (2007): «A word for “horse” in Chinese and Dravidian», International Journal of Dravidian Linguistics 36(2), 19-44. —. (2009): «The Ancient Mesopotamian Place Name “Meluhha”», Studia Orientalia 107, 135-76. 73 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE LIBERMAN, A. (2010): «Iconicity and etymology». In: J. Conradie, R. Johl, M. Beukes, O. Fischer & Ch. Ljungberg (eds.), Signergy, 243-58. Amsterdam: J. Benjamins. LUBOTSKY, A. (2001): «The Indo-Iranian Substratum». In: Ch. Carpelan, A. Parpola & P. Koskikallio (eds.), Early Contacts between Uralic and IndoEuropean. Linguistic and Archaeological Considerations, 301-17. Helsinki: Mémoires de la Société Finno-Ougrienne. MALLORY, J. & D. ADAMS, eds. (1997): Encyclopedia of Indo-European Culture. London: Fitzroy Dearborn. MALLORY, J. & D. ADAMS (2006): The Oxford Introduction to Proto-IndoEuropean and the Proto-Indo-European World. Oxford: Oxford UPress. MATISOFF, J. (2003): Handbook of Proto-Tibeto-Burman. Berkeley: University of California Press. MCALPIN, D. (1981): Proto-Elamo-Dravidian: The Evidence and its Implications. Philadelphia: The American Philological Society. MIKHAILOVA, T. A. (2007): «Macc, Cailín and Céile – an Altaic Element in Celtic?». In: H. L. C. Tristram (ed.), The Celtic languages in Contact, 4-24. Potsdam: Potsdam UPress. MORGENSTIERNE, G. (2003): A New Etymological Vocabulary of Pashto. Wiesbaden: Ludwig Reichert. MUSA[Y]EV, K. M. (1975): Leksika tjurkskix jazykov v sravnitel’nom osveščenii. Moskva: Nauka. —. (1984): Leksikologija tjurkskix jazykov. Moskva: Nauka. —. (1996): «Dravidian-Turkic-Sanskrit Lexical Comparisons». In: Á. Berta, B. Brendemoen & C. Schönig (eds.), Symbolae Turcologicae. Studies in honour of Lars Johanson on his Sixtieth birthday (8th March 1996), 169-74. Stockholm: Swedish Research Institute in İstambul. PARPOLA, A. & J. JANHUNEN (2011): «On the Asiatic wild asses and their vernacular names». In: T. Osada & H. Endo (eds.), Linguistics, Archaeology and the Human Past (Occasional Papers, vol. 12), 59-124. Kyoto: Indus Project (Research Institute for Humanity and Nature). POKORNY, J. (1959): Indogermanisches Etymologisches Wörterbuch, 2 vols. Bern: A. Francke. POPPE, N. (1955): «The Turkic loan words in Middle Mongolian», Central Asiatic Journal 1, 36-42. —. (1965): Introduction to Altaic linguistics. Wiesbaden: Harrassowitz. RADLOV, V. (1893): Opyt” slovarja tjurkskix” narečij, tom I, čast’ 2. Sanktperterburg”: Imperatorskaja Akademija Nauk”. RAHMETI ARAT, R. (1947): Kutadgu bilig, vol. I Metin. İstanbul: Millî eğitim basımevi. RAMANATHAN, P. (2002): «Situating Indus Valley Civilization in the Protohistory of India», PILC Journal of Dravidic Studies 12(1/2), 79-106. 74 Türkbilig, 2012/24: 41-76. SOME THOUGHTS ON DRAVIDIAN-TURKIC-SANSKRIT LEXICAL COMPARISONS —. (2003): «Direction of Movement of Dravidian Speakers in Prehistoric Times– Descent from the North or Ascent from the South?», Dravidian Studies 1(3), 90-107. RAO, G. Sambasiva (1971): A comparative study of Dravidian noun derivatives, Ph.D. (mimeo), Cornell University. RIX, H., ed. (20012): Lexikon der indogermanischen Verben. Wiesbaden: Ludwig Reichert. RIX, H. (1994): Die Termini der Unfreiheit in den Sprachen Alt-Italiens. Stuttgart: Steiner. RÓNA-TAS, A. (1988): «Altaic and Indo-European: Marginal Remarks on the books of Gamkrelidze and Ivanov», Acta Orientalia Academiae Scientiarum Hungaricae 42(2/3), 391-404. —. (1998): «The Reconstruction of Proto-Turkic and the Genetic Question». In: L. Johanson & É.Á. Csató (eds.), The Turkic Languages, 67-80. London & New York: Routledge. RYBATZKI, V. (2006): Die Personennamen und Titel der mittelmongolischen Dokumente. Helsinki: Institute for Asian and African Studies. —. (2010): «Türkische Lehnwörter im Burushaski». Studia Orientalia 108, 14979. SCHÖNIG, C. (2002): «Anmerkungen zum B- ~ M-Wandel in den modernen Türksprachen». In: I. Hauenschild, C. Schönig & P. Zieme (eds.), Scripta ottomanica et Res Altaicae. Festschrift für Barbara Kellner-Heinkele zu ihrem 60. Geburstag, 257-75. Wiesbaden: Harrassowitz. —. (2003): «Turko-Mongolic relations». In: J. Janhunen (eds.), The Mongolic languages, 403-19. London & New York: Routledge. SCHUESSLER, A. (2007): Etymological Dictionary of Old Chinese. Honolulu: University of Hawai’i Press. SHAGDARSÜRÜNG, Ts. (2005): «Problems of methodology of Altaic comparative linguistics». Rocznik orientalistyczny 58(1), 179-85. SORRENTINO, A. (1988): «Corrispondenze lessicali nell’ipotesi di parentela elamo-dravidica», AION 10, 239-54. SOUTHWORTH, F. (2005): Linguistic Archaeology of South Asia. London: Routledge Curzon. STACHOWSKI, K. (2008): Names of Cereals in the Turkic Languages. Kraków: Księgarnia Akademicka. ŠČERBAK, A. M. (1961): «Nazvanija domašnix i dikix životnyx v tjurkskix jazykax», Sbornik “Istoričeskoe razvitie leksiki tjurkskix jazykov”, 82-172. TRAUTMANN, Th. (2006): Languages & Nations. The Dravidian Proof in Colonial Madras. Berkeley: University of California Press. TURNER, R. L., Sir (1962-1966): A Comparative Dictionary of Indo-Aryan Languages. London: Oxford UPress. TYLER, S. (1968): «Dravidian and Uralian: The Lexical Evidence», Language 44, 798-812. 75 Türkbilig, 2012/24: 41-76. José Andrés ALONSO DE LA FUENTE VYCICHL, W. (1983): Dictionnaire étymologique de la langue copte. Leuven: Éditions Peeters. WINTER, C. A. (2005): «The Dravidian origin of the word for horse», International Journal of Dravidian Linguistics 34(2), 193-202. WINTER, W. (1989): «On a new claim concerning substratum influence upon Tocharian», Central Asiatic Journal 33(1/2), 126-32. WITZEL, M. (2005): «Loan words in western Central Asia. Indicators of substrate populations, migrations, and trade relations». In: V. Mair (ed.), Contact and Exchange in the Ancient World, 158-90. Honolulu: University of Hawai’i Press. —. (2006): «South Asian Agricultural Vocabulary». In: T. Osada (ed.), Proceedings of the Pre-Symposium of RHIN and 7th ESCA Harvard-Kyoto Round Table, 96-120. Kyoto: Research Institute for Humanity and Nature (RHIN). ZVELEBIL, K. (1970): Comparative Dravidian Phonology. The Hague & Paris: Mouton. —. (1991): «Long-range language comparison in new models of language development: the case of Dravidian», PILC Journal of Dravidian Studies 1, 21-31. —. (1999): «The Dravidian perspective». In: C. Renfrew & D. Nettle (eds.), Nostratic: Examining a Linguistic Macrofamily, 359-65. Cambridge: The McDonald Institute for Archaeological Research. 76