I. Consonant and vowel orders:
Japanese "Kana" consonant order:
k, s, t2 (dental), [nasal, n], h(p)*[I], [nasal, m], y, r2(also
l2, retroflex?), w, [nasal, n (terminal)]
Tamil consonant order:
k, c, t(retroflex), t2(dental), p, r, [nasal set: ng, ny,
n(retroflex), n (terminal), m, n2], y, r2 (retroflex), l (dental), v, zh
(retroflex), l2 (retroflex)
Japanese and Tamil also share a nearly identical set of vowels (a,
i, u, e, o)*[II]: all of these vowels have both a short and long form. In Tamil
script short and long vowels are represented by different letters, in Japanese
script one repeats the vowel to extend it (sorta).
The letter order similarities could be a classic case of evolution from a common ancestor. It is speculated that the Japanese kana ordering derives from the
Siddham script (based on the Indian Brahmi, an early script for official edicts) c. 9th
centruy AD [6]. Tamil may have incorporated the Brahmi script independently as
well [7]. However, there are few important similarities between the Japanese
and Tamil alphabet that cannot be brushed aside as products of parallel
evolution:
1) Both Tamil and Japanese have two different nasal 'n' sounds,
one of which never ends a word and one of which never begins a words. I am not
aware of a Brahmi parallel.
2) Tamil does not have arbitrary consonant combinations (as do,
for instance, Sanskrit or English). The only valid consonant combinations are
i) the same consonant repeated or ii) the 'n' + consonant combination. The same
is true of Japanese. Is this true of Brahmi?
3) Neither Japanese nor Tamil have a concept of voiced and
unvoiced, or aspirated and unaspirated, consonants (unlike Brahmi which
distinguishes these categories). Whether these were later coincidentally
dropped from Brahmi in both languages remains to be shown.
Hence, it is entirely possible that the phonetic systems in the
two languages shared similarities (which may have predated any script).
Footnotes:
[I] 'h' is the consonant that is modified with a "maru"
to form the 'p' sound in Kana. This is not unique to Japanese. 'p'-sounds in
Tamil, often morph into 'h'-sounds in Kannada (a sister language of Tamil),
e.g. Tamil 'paalu' = Kannada 'haalu' (milk))
[II] Exceptions: The Japanese u sounds more like the German u with
an umlaut, not so in Tamil (but this sound may be a borrowing from Chinese
which has this exact u sound [13]). Tamil has the diphthongs "ai" and
"au" as well, but even though Japanese does not treat these as
vowels, the "au" and "ai" sounds are frequently heard vowel
combinations (e.g. kAU = to buy, hAI = yes), although these are generally not
pronounced as an unitary diphthong.
II. Syntactic similarities
II. A. Word order
Japanese and Tamil have incredibly similar word order. At the very
basic level, they both follow SOV (subject-object-verb) order (unlike English
which follows SVO). While much needs to be said about this, here are some
simple examples to whet your apettite.
Example 1
Consider the following set of sentences:
The cheese was rotten. A rat ate this cheese. A cat chased this
rat. Mike owns this cat.
Try combining these sentences into one in English, and here's what
you might get:
(E1) [[[[Mike owns the cat] that chased the rat] that ate the
cheese] that was rotten].
In this sentence, Mike is the subject of the sentence (each [...]
encodes a complete sentence), so the innermost nested [...] is the core
sentence[I] containing the subject. Now let us try turning this on its head, so
that the cheese is the subject of the sentence.
(E2) [[[The cheese that was rotten was eaten by a rat] that was
killed by a cat] that is owned by Mike].
Again note that here the core sentence (contained in the innermost
nested [...]) is [the cheese ... the rat] in which subject (the cheese) is
present.
Now let's try this in Japanese:
(J1) マイクがかったねこがおいかけたねずみがたべたチーズはくさっていた。
[Mike ga katta [neko ga oikaketa [nezumi ga tabeta [cheesu wa
kusatte ita]]]].
Corresponding English words in Japanese word-order:
[Mike kept/raised [cat chased [mouse eaten [cheese bad was]]]]
In the case of the Japanese sentence, the cheese is clearly the
subject (nested, inner-most sentence is [cheese was bad]). Clearly, a direct
correspondence seems impossible between Japanese and English syntax: while
sentence J1 appears to follow the word order of E1, however does not have the
same subject/core sentence as E1. Again, while J1 has the same core sentence as
E2, the word order is completely different. Let us now try this in Tamil.
(T1) Mike valarttha poonai thuratthiya eli pusittha cheesu oosi
irundadu.
Corresponding English words in Tamil word-order:
[Mike kept/raised [cat chased [mouse eaten [cheese bad was]]]]
T1 and J1 are identical!
If you give it some thought, you will realize that this is a direct consequence
of the SOV word order [16]. Ok, are there any parallels that are not a direct
consequence of the SOV order? Here is another example.
Example 2
Consider the sentence:
(J2) 猫に食べ物をやりました。
[Neko ni tabemono wo yarimashita]
Corresponding English words in Japanese word-order:
[cat {indirect object marker} edible-item {direct object marker}
gave]
In English this would be:
(E3) I gave the cat food.
In Tamil this would be:
(T2) Poonai-irkku unavupporul-ai kodutthen.
Corresponding English words in Tamil word-order:
[cat {indirect object marker} edible-item {direct object marker}
gave]
Again, note that while T2 and J2 are in direct correspondence, E2
differs significantly in terms of word order. Also interesting is the fact that
dropping the subject (I) is completely acceptable in both Japanese and Tamil.
Except that the final verb conjugates according to person and number in Tamil
(but not in Japanese), so that the subject of the sentence T2 can only be
"I (singular)". Whereas in Japanese, the subject could be any person
(first, second or third, singular or plural). [II]
Footnotes:
[I] What I keep calling "core sentence" here probably
has a (bettter!) name in linguistics jargon I am not aware of. I understand
that the other clauses that are tacked on to the "core sentence" are
called "pre/postnominal relative clauses" [16].
[II] This feature (or lack of it) has been source of incredible
amusement to me and many others. I quote from [19]: "...Because of this,
the sentence "He just killed her!" and "I just killed her!"
sound exactly the same, meaning that most people in Japan have no idea what is
going on around them at any given moment. You are supposed to figure these
things out from the "context", which is a German word meaning
"you're screwed"....". For the rip roaringly funny article (to
be read in a spirit of good fun), see [19].
II. B. Particle correspondences
I am currently working through Chevray and Kuwahira's Japanese
Grammar (Schaum's Outlines), as well as the two volumes of げんき (Banno,
Ohno, Sakane, Shinagawa, Tokashiki; Japan Times), both introductory texts of
conversational Japanese. Based on these, I have found that particle
correspondences between Tamil and Japanese can be built in a consistent way.
More on this will appear soon!
II. C. 1. Words derived from the same root with related meanings
A remarkable feature I have observed in both Tamil and Japanese,
is the derivation of a variety of loosely related words from the same
root.
For instance, consider the set of words:
kan (eye) -- kaan-pathu (see) -- kaanu-vathu
(visible) -- kaan-pippadhu (show) -- kandu-pidipathu (find/discover)
These are all derived from the same Tamil root “kan
In Japanese, an identical set of words can derive from the same
root: “me” (pronounced may) also
meaning eye.
me (eye) -- mi-ru (see) -- mi-eru
(visible) -- mi-seru (show) -- mi-tsukeru (find/discover)
Of course, this is how simple and compound words are built in
every language. So what is unique about Tamil and Japanese?
It is my opinion that such root-relationships for words with
related meanings share close correspondences in Tamil and Japanese, whereas in
other languages such relationships never existed, or have diverged to such an
extent as to be unrecognizable.
Note that the verb “discover” does have something to with being
able to “see” a hidden object. However the English terms “find” or “discover” have
little direct etymological relationship to “eye”, whereas their Japanese and
Tamil counterparts do.
Here’s another example: In Japanese, the noun pre/post fix (居), “i” (pronounced “i” as in hit) can mean
either “being” or “sitting”; the commonly encountered Japanese verb “i-ru”
(to-be) derives from the same root. In Tamil, the same root (in fact the same
sound) “i-ru” has similar connotations: “iru-ppu” means “status/being”; “iru-kkai”,
a related term means “seat/sitting”
In Tamil, certain roots “kel” can
assume a variety of different meanings based on context.
a) kel (hear), b) kel (ask/inquire), and c) kel (to be effective)
a) is the most common. b) is most familiar in its noun-conjugated
form “kel-vi” (questions). e.g “nee kel” can mean “you ask” or “you hear”.
c) is encountered less frequently, and is more of a colloquialism,
but is commonly used to assess the effectiveness of, for instance a medicine: “indha
marundu ket-kum” (this medicine is effective).
These usages are already highly disparate. Imagine my surprise
then, when I came to know that the root “kiku” in Japanese is used in the exact
same three contexts!
a) kiku (hear), b) kiku (inquire), c) kiku (to be effective)
In Japanese “kare ga kiku” (彼が聞く)
can mean either that “he listens” or
that “he asks/inquires”, just as in Tamil. The Kanji for these two usages is
also the same.
Amazingly the same word “kiku” can also be used in the third
context: 薬がきく(kusuri ga kiku), means the “medicine is
effective”. Notice that the Kanji for this usage of kiku is different from the
other two.
How all three parallel usages could have come about by convergent evolution defies simple explanation, but if you do have one, feel free to leave a comment.
II. D. The ubiquitous verb-pairs: Transitive and Intransitive
Transitive/intransitive verb-pairs from the same root are
ubiquitous in both Tamil and Japanese. For example: The Japanese verb-pair “hajimaru/hajimeru” and the Tamil-verb pair “thodanguvadhu/thodakkuvadhu”.
While such verb-pairs are common in many languages, the
relationships among these verbs share certain unique features in Tamil and
Japanese. More on this soon.
This is an incredible finding! I had not dwelt on the technical aspects that draws a parallel between the two languages, however, my personal experience had made me understand that both Japanese and Tamil 'SOUND' the same. If you are able to conduct an experiment on two persons speaking Japanese and two other persons speaking Tamil, for a person who understands both languages as a 3rd or 4th Language, they sound confusingly the same!
ReplyDeleteVery interesting Article! Thank you!
Japanese clearly distinguishes voiced and unvoiced consonants... that's what the diacritical marks in kana indicate.
ReplyDeleteHarrisonGlen,
ReplyDeleteThank you for pointing this out.
It is indeed true that modern Japanese does distinguish between voiced and unvoiced consonants. However, it is speculated that the diacritical mark or 'dakuten' (濁点) used for distinguishing voiced from unvoiced consonants was not native to Japanese, and may in fact, have been borrowed from China.
I quote: "The kun-yomi pronunciation of the character 濁 is 'nigori'; hence the daku-ten may also be called the nigori-ten. The meaning of this character—muddy or turbid—is a hint that the theory behind this kind of spelling device originally came from China, where consonant sounds were traditionally described by terms such as clear, half-clear, muddy, and so on."
Source: http://en.wikipedia.org/wiki/Dakuten
Unlike the Devanagari, Greek or Cyrillic alphabets that have distinct letter forms for voiced and unvoiced consonants, both Japanese and Tamil letters use the same letters for either consonant type. As a native speaker of Tamil, I can say with fair confidence that words employing voiced consonants in Tamil almost exclusively derive from Sanskrit. I wonder if the same is true for Japanese.
Perhaps someone with knowledge of ancient Japanese could comment on the use and/or prevalence of voiced consonants in classical Japanese speech or texts?
Your footnotes for the "Consonant and Vowels orders" says that the Japanese 'u' sounds more like the German 'ü'. The German 'ü' is a CLOSE FRONT ROUNDED vowel, but the Japanese 'u' is a near to CLOSE BACK VOWEL (although a bit compressed with protruded lips). This vowel is strikingly close to the Tamil's "Kuttriyalukaram" which is the allophone of the vowel 'u' usually at the end of words (though in the middle is not uncommon) as in the ending 'u'in அது (adhu)... further strengthening the Hypothesis :)
ReplyDeleteTwo things I found strikingly similar between Japanese and Tamil
ReplyDeleteka at the end of a sentence as in unakku arivu irukka (tamil) and the
na (ending adjective) as in azhagana poo (kireina hana in japanese)
திருக்குறளில் ஔகாரமும், ஔகார வரிசை எழுத்துக்களும் பயன்படுத்தப்படவில்லை. ஆனால் ஐகாரமும் ஐகார வரிசை எழுத்துக்களும் பயன்படுத்தப்பட்டுள்ளன.
ReplyDelete(ஔகாரத்தின் இணையெழுத்து உ, ஐகாரத்தின் இணையெழுத்து இ.)
அதே போல சங்க இலக்கியங்களில் ஐகாரம் , அகரமும் இகரமும் கூடிய ஒலியாகக் கருதப்பட்டு ஐ என்று மட்டுமே எழுதப்பட்டது ; அய் என்று எந்த ஓர் இடத்திலும் எழுதப்படவில்லை. அதாவது சங்க இலக்கியங்களில் ஐவர், ஐந்து, ஐம்பது போன்ற சொற்கள் அய்வர், அய்ந்து, அய்ம்பது என்றாற்போல எந்த ஓரிடத்திலும் எழுதப்படவில்லை. ஐவர், ஐந்து, ஐம்பது என்றாற்போல மட்டுமே எழுதப்பட்டன.
ஆனால் இதற்கு மாறாக, ஒளகாரமோ சங்க கால இலக்கியங்களில் ஒள என்றும் அவ் என்றும் இரு வகையாக எழுதப்பட்டது. பௌவம் (கடல்) என்ற சொல், பௌவம் என்றும் பவ்வம் என்றும் எழுதப்பட்டது. அதே போலக் கௌவை (அலர் அல்லது பழிச்சொல்) என்ற சொல், கௌவை என்றும், கவ்வை என்றும் எழுதப்பட்டது.
பெரும்பாலும் பிறமொழிச்சொற்களே ஓகாரவரிசையைக் கொண்டுபிறக்கின்றன. மௌளி, கௌரி.
இவற்றை நோக்கும் போது ஔ என்பது தமிழில் கடையாகச் சேர்ந்த சொல்லாக இருக்க வாய்ப்பிருக்கிறது.
ஜப்பானும் தமிழும் ஒரு மொழிக்குடும்பத்தில் பிறந்தது என்று வைத்துக்கொண்டால் ஜப்பானில் அதற்கு தனியொருயெழுத்து இன்னும் தோன்றவில்லை. தமிழ் பிரிந்துவந்தபின் அது தமிழில் தோன்றியிருக்கலாம்
ithey pol "padi" enra thaizh vaarthai pala porul tharukirathu. 1. to study/ to read; 2. to obey-"keezhpadi"; 3. step; 4. measuring tool.
ReplyDeletehowever, i congratualate you for this wonderful findigns! i am inspired by you. i also come across such similarities between tamil and other languages. for example, tamil and greek. there are numerous words that are similar in both pronunciation as well as meaning. thank you!
I saw the same examples 1 and 2 in Professor Ono's article.
ReplyDeleteOK I've heard this theory before & thought it was kooky, but yes, eerily similar syntax & morphology. And yes, sir, you are correct. Voiced "muddy" sounds were borrowed from China.
ReplyDeleteThe only thing about the organisation of consonants/ vowels -- that is probably a later adaption from Sanskrit via Buddhist studies, which included śabdavidya (linguistics). And that Sanskritised organisation of sounds also influenced the organisation of Dravidian languages.
It doesn't seem farfetched, because Korean -- spoken just across the sea -- also shares all that grammatical similarity with Japanese. Yet the vocabulary is completely different. So Dravidian languages like Tamil & Malayalam across the ocean share the same syntax, but also with completely different set of lexicon.
Question is, did this proto-language travel via land or sea...?
good article and very interesting, excellent research.
ReplyDeleteAnother feather in the cap of Tamil Language.
Cats are used as an example and japanese are crazy for cats and love cats. (pun intended).
Hats off