The Eccaviyal of Tolkaappiyam


Dear Friends,


By general consensus the Dravidian languages (along with some others) are agglutinative and this feature has been isolated for categorizing them as a distinctive of family of languages and in contrast to the inflexional tonic and so forth. In my studies of Rigkrit, the language of Rig Veda and allied texts and Sanskrit such as that of Bagavath Gita, I am finding that in the formation of Noun Phrases Verb Phrases and Nominal compounds and so forth, the agglutinating tendency is widely present so much so that these languages can be categorized also as agglutinating and combined with identities in core lexicon and grammatical particles , the languages can be categorized also as Dravidian just as much as the languages of Tamil Telugu Kannada Malayalam(perhaps Marathi Gujarati Bengali) and so forth.


I am venturing to translate with commentary the final chapter of the Second Book of Tol. namely the Eccaviyal of Collatikaaram which deals with these and other relevant aspects. I shall follow the commentary of Teyvaccilayaar ( c. 14th cent. AD) in Tamil ( only the most relevant parts), provide an English translation of it, and add some comments if necessary. I shall omit the Tamil texts for IC and akandabaratam but provide the full for Meykandar etc.


The moderators of IC can indicate to me whether I should continue posting to IC or not.


This is by way of educating myself more effectively on the grammatical principles that I see also at work in Sk of Bagavath Gita and so forth. I hope with such a familiarity of the agglutinating features of Tamil language, my comments on linguistic study of Gita will be more refined.







The Eccaviyal of Tolkaappiyar




9: Ţ     9. Eccaviyal


ž â  


The Words that can be used in the composition of Verses




â  ż


iyaRcol tiricol ticaccol vadacolenRu

anaitttee ceyyuL iiddac col


š 򾧾 ɢ, Ţ ; Ţ¡ Ӿġ â ȡ 𧼡 򾡾 ը νоġ .


۾Ģȡ ɢ, Ũ¡ â Ģ о ѾĢ.

-: â ż Ũ, ǡ̾â , -.



If asked what is the title of this chapter, it is Eccaviyal ( that which deals with the remainder); and it deals with all those issues not discussed in the earlier eight chapters beginning with KiLaviyaakkam and ending with Uriccol.

This first Sutra intends to give a kind of categorization of words that are suitable to be used in the (scholarly) compositions of verses.


Meaning: All the words that are deemed suitable to be used in the composition of verses are classified into iyaRcoL, Tiricol, Ticaccol and Vadacol


What these are exactly will be described in the sutras that deal individually with them




즸 Ž




iyaRcol taamee

centamiz nilattu vazakkodu civaNit

tamporuL vazAamai icaikkum collee


-: ġɾ, 즸 ơ -.



‘ż â

¢ ĸ

¢ ӾĢ


mai¡, ĸ Ţ¡, ȡ Ȣžɡ â¢ θ ¡Ƣ ¢ , ⢠ż, ̽Ģ , ̼Ģ .


즸 Ž ơ ġ : - , , , , , ȡ, ɡ, ɡ .


ġ Ģ ġ¢. 了 ɢ




Among these word classes the Natural Words are those

That are in agreement with the usages

in the lands of Centamiz and in active use

without distortions of their natural meanings.


Comments :


Now since it is said in the forward:

(Tollkaappiyar investigated) the phonology

morphology grammar and meanings of both

the language of speech and literature in the lands of Centamiz (the Chaste Tamil)

that lie between Vengadam in the North and Kumari in the South”


the lands meant by the Lands Centamizare in fact the countries lying on the South of Vengkadam, North of Kumari, East of the Western seas and West of the Eastern seas.


Sample of words that are in agreement with thelanguage in use in these lands and which function without deviating in meaning are: cooRu( cooked rice), kuuz ( porridge), malai( hill), maram (tree) ; uNdaan ( he swallowed), tinRaan (he ate), oodinaan ( he ran), paadinaan ( he sang) and so forth.


Such words, because in these lands are in active use without any deviations, are called Natural Words. They can also be termed Cenjcol ( the proper words)


Note: This classification rather different from that available in Sk grammar. What we have here is reference to the LIVING PRESENCE and ACTIVE USE of a language and uniformity in their morphology and semantics throughout the countries demarcated. This comment and definition also points out that at the time of Tol. the whole of South India, the vast stretch of land that now is divided into Tamil Nadu KeraLa Karnadaka and TuLu were known as the land of CenTamiz and hence aland unified with the rule of the SAME language, viz, CenTamiz.


(to continue) 1



The Eccaviyal of Tolkaappiyam




զ Ȣ ڦ ġ

ڦ Ȣ šզ ġ

â Ţ



â о ѾĢ



զ Ȣ ġ, Ȣ ġ Ţ ̾ â, -

ɧ, ġ š âġ šȡ.




oruporuL kuRitta veeRu col laakiyum

veeRuporuL kuRitta voru col laakiyum

irupaaR Renpa tiricol kiLavi


This sutra explicates the meaning of Tiricol




The Tiricol are the semantically ambiguous words where the dual processes hold viz. the same word meaning different things and the same meaning being conveyed by different words.


So such words are those which communicate the same meaning as communicated by the Natural Words but in different ways.


[Notes:( Loga)


Another meaning implicit in the concept of Natural Words is brought out here. While the natural words are semantically unambiguous throughout the whole land of Centamil and hence with univocity in meaning, the Tiricol are different in being polysemantic and polymorphic and thus requiring disambiguation using regional and contextual factors. The morphologically same word may have different meanings in different regions and contexts(polysemantic). Similarly the same MEANING may be communicated by morphologically different words in different regions etc (polymorphic). They are called Tiri Col i.e. deviant words on account of this. Note that this definition presupposes the meaning of the Natural Words.]


, âġ ɢ ɢ, â ̨ ġ â ġ


Now it may be asked how such words differ from the category of Uriccol. Here we should note that while the Uriccol may come as incomplete words ( i.e. words in adverbial and adjectival constructions) the Tiricol may come as full words ( i.e, capable of functioning as autonomous units like nouns and verbs)


, ɡ 󧾡 Ȣ¢ .


Now because this categorization presupposes the category of Natural Words, we have to take it that, such words are also those that occur in the lands of Centamil


¡: , Ũ , Ƣ, Ƣ , , . Ȣ.


¡Ȣ ̨Ȩ, , Ģ 򾢨 . ¡Ȣ ̨Ȩ, Ţ . Ƣ , , . ڦ Ȣ. š âվĢ â ¢.


The examples areas follows: the meaning of‘malai’ (hill) is also communicated by the words ‘ kunRu’ and ‘varai’; the meaning of ‘kuLam’ (pond) by ‘puuzi’ and ‘paazi’;the meaning of ‘vayal’ by ‘cey’ and ceRu’. These are examples where the SAME meaning is communicated by different morphological units.


The converse holds in the following cases: the word ‘araGkam’meaning the landmass in the center of a river, the stage for dancing, and acorner in the house. The word ‘turutti’ means the land mass in the centre of a river and kind of skin instrument. The word ‘aazi” meansthe sea, the circular disc (of kings) and just simply the circular shape.


Such words are called Tiricol on account of such deviations from the norm (as exemplified by the Natural Words)


(to continue) 2



The Eccaviyal of Tolkaappiyar




Ȣ ɧ  Ţ


ġ о ѾĢ

(-) ɢ ǡ Ȣ ġ -



centamiz ceernta panniru nilattun

taGkuRip pinavee ticaiccoR kiLavi


It is intended to explicate the category of Ticaic Col


Meaning: The Ticaic Col are those words which have only regional prevalent in the 12 countriesthat linguistically belongto the lands of Centamiz.


ɢ : - Ũ¡Ȣ 측 , Ǣ , , 𼿡, ̼ , [ , , , Ƣ , ġ, š , š ż .


(Ȣ: ڧš ý; ȡ ȡ ý)


The TWELVE countries are as follows: -- The PotuGkar Naadu on the South East of the River Vaiyai; oLi Naadu,Ten paaNdi Naadu, karuGkudda Naadu, Kuda Naadu, the PanRiNadu, the KaRkaa Naadu, the Siitai Naadu, the Puuzi Naadu, Malaadu, Aruvaa Nadu and the country on the north of AruvaNadu.


These countries are within the Centamiz Naadu.


Ȩ¡, ȿ Φ ý :



¢ ɼ θ


̼ , ȸ”




“ ̼ Ģ мۨ â

Ԩ âο š¢



ɢ ¢ Ө”




¢ , â ɢ ú ɢ , ġ ŢҨ Ȩ¡ “ ĸ, ¢ӾĢ, ” “ Ž 򦾡 ” žɡ,Ž Ȣ Ħ ξġ, ɢ : â¡Ȣ , , ̼ ȸ, տ Ģ Ģ


, ξĢ , â¡Ȣ ż á Ħ ʧȢɡ .


áŢ żը šġ, Ũ ܼ Խ.

ġ 了Ģ ȡ Ǹ , -.


Now because it is said “ the 12 countries one with the lands of Centamiz” , these lands (of Ticaic col) are different from the above. Some scholars enumerate them as follows:


The lands beyond the boundaries -ancient island in the seas south of Kumari, Kollam, Kuupakam , SingkaLam and Kannadam, teLiGkam koGkaNam TuLuvam Kudakam and kuNRakam.The Sutra from Akattiyam will enumerate them as follows: The countries on both sides of the Western hills where a deviant type of Tamil is prevalent, the words that are pleasant in the lands of the TEN governors along with another TWO who co-exist and who all rule the lands under the control of the Three Rightful Kings. Here the twelve countries are those beyond the boundaries of Tamizakam and where a deviant kind of Tamil is prevalent and where the prevalence of Centamiz was very welcome. This interpretation is further supported by the fact that in the Forword (Paayiram to Tol) , over and above saying “ in the countries where speech in Tamil is prevalent, investigating both the literary and verbal expressions etc” it is also said “ also examining the ancient texts in the lands which are in essence the same as the lands of Centamil’, it has to be taken that the lands meant are those which are related to the boundaries. Thus the 12 countries are: the ancient island in the seas south of Kumari, Kollam, kuupakam, SingkaLam, the countriesWest of the Hills such as KoGaNam, TuLuvam Kudakam and KuNRakam; the countries on the East Karnaadakam, Vaduku TeliGku andKaligkam

Among these it may be possible that because Kuupakam and Kollam were swallowed by the sea, people migrated to the northern shores of Kumari and called the lands with the same name. Now the people in the North also call these lands the Panjc Dravida , and hence these five cannot be countries on the South of Vengkadam


The words in these countries, even if different from the Natural Words of Centamil, but on account of they occurring in the literary compositions of great scholars, they will not be disallowed.


Notes( Loga)


It appears that at the time of Tol (c. 3rd cent BC) the whole of the South , the lands on the South of the Vindhyas, were essentially Tamil with the languages beyond the borders of CentamiL Naadu, considered the lands of Kodun Tamil, a kind of deviant Tamil and hence only dialectically different and where the Tamil language was most welcome. It is interesting that in this list we have Sri Lanka, KaliGka, Andhra, Karnadaka and perhaps also Maharastra and Gujarat ( as they were also known as lands of Pancha Dravida)


There is clear recognition of DIALECTICAL variations of Tamil which was a living language in the whole of the South with the rule that even words specific to a dialectic variety will be accepted by scholars as part of Centamil language.





The Eccaviyal of Tolkaappiyar




żĢâ 츽



ż Ţ ż ã

򦾡 ҽ ġ



żŢ ¡ о ѾĢ


-:żġ ġ żƢ째 âš 츢 Ƣâ 򦾡 ҽ ġ, -


ġ š¢, ż ¢ ̾Ģ ż ¢. ż Ⱦɡ 측 ġ .


żƢ¡ : - šâ, , . , , .


The Grammar of Northern Words




VadacoR kiLavi vadavezuttu orIi

ezuttodu puNarnta collaa kummee


Meaning: The words designated as Vadacol ( Northern Words) are those which are incorporated into Tamil by deleting those phonemes specific to the Northern languages and replacing them with equivalent phonemes available in Tamil

Even though common to all the nations, because current in the Northern countries, they are termed as Northern Words. Now because it said Vadacol ( Northern Words), the living languages such as Prakrit are also included. ( prakiritam> paakatam)

The example of words of Northern Language: vaari , maNi, kungkumam and so forth. The Prakrit words are : vaddam, naddam, paddinam and so forth. There are many such words (in Tamil)


ľ ɨ




â Ũ¡


ľ ɨ о ѾĢ


- : 򦾡 Ȣ â, 츢 ɨ , -

¡: - , , , , , . š

Ģ, , Ţ  â . â , Ţ . ż á šá.


Additional Explanatory Notes for the Sutra Above




citaintana varinum iyainatana varaiyaar


Meaning: Even if some words come with gross distortions and where systematic phonemic mapping does not hold, even such words will not be rejected as long as they are overall consonant with the phonemes of Tamil

The examples are : kantam, tasanaangku, saakaramm sattiram pavaLam and so forth.

Among these FOUR different of words, it may be observed that the Natural Words may occur as Nouns Verbs, the Enclitics, the Prepositionsand so forth; the TriCol and Ticaic Col occur mainly as Nouns and in a limited way as Verbs. The Northern Words do not occur except as Nouns


Notes ( Loga)


We shouldn’t fail to notice that Tolkaappiyar offers here a linguistic survey of a kind of the WHOLE of INDIA, where the Tamil lands are designated as the lands of SenTamil, the adjacent lands, including Maharastra Gujarastra ( the lands of Pancha Dravida) are designated as lands of Kodun Tamil, which is perhaps a native term for what is called now as dialectical variations. The countries outside these are simply termed Northern lands and where Sanskrit Prakrit and such other languages prevailed with their distinctive phonemic clusters.


We should also note that Tol. does not speak of lexical borrowings and hence Sanskritization/ Aryanization of the Tamils and Tamil language ( something that came in vogue only at the beginning of the 17th cent.) . Though there may be words with peculiar phonetic realization in Sanskrit and Prakrit, but they are taken as COMMON WORDS (potu coL) which shows of there cognition that throughout India there was a common repertoire of lexemes but with different kinds ofphonological realizations, some even undergoing gross distortions (citaivu).


It also follows that Tol. notices that there can be differences in what can be called STYLES of Language use where while one term may be actively in use in one linguistic community, this may not be so in another ( even neighboring) community. Thus the Vadacol are such words, words which are also native to Tamil but NOT in current activeuse., and hence words that have become obsolete.



The Eccaviyal of Tolkaappiyar







ŢâƢ Ţâ Ƣ


Ģ ɡ .



Ũ Ǹ Ţ о ѾĢ.


Phonological Modifications in Literary Compositions


annaaR collum todukkuG kaalai

valikkumvazi valittalum melikkumvazi melittalum

virikkumvazi virittalum tokkumvazi tokuttalum

niiddumvazi niiddaluG kuRukkumvazi kuRukkalum

naaddal valiya enmanaar pulavar



It is intended to explicate how the above FOUR kinds of words suffermorphological changes when used in literary compositions




:-- , â, , ż , -. , ȡ ǡ -.


The FOUR kinds of words are : - the Natural Words ( iyaR col) , the Non-natural Words ( Tiricol), the Dialectical Words ( Ticaic col) and Northern Words( Vadacol) . The phrase ‘todukkuG kaalai” means ‘when composing a connected piece of verse or literary piece i.e. discourse)


ĢƢ Ģ 򾢨 򾡸 Ƣ 򾡸 . ĢƢ Ģ 򾢨 򾡸 Ƣ 򾡸 .


Hardening the word when needs to be hardened ( i.e. introduce the plosives) means replacing the soft phonemes ( the nasals)with hard consonants when composing.Softening when needed to soften means the transforming the plosives into nasals when composing.


ŢâƢ Ţâ ٽ Ģ Ţâ츧 Ƣ Ţâ . Ƣ զǪ Ģ .


When needed toexpand ( i.e. geminate ) then expanding means within words that communicate meanings introducing one or two additional phonemes. When needed to delete then deleting means with words that communicate meaning, deleting one or two phonemes.


Ƣ 򾡸 򾡸 Ƣ 򾡸 . Ƣ ġ 򾡸 򾡸 츧 Ƣ 򾡸 .


When needed to lengthen then lengthening means, lengthening the short phonemes of meaningful words when needed to do so when composing, if necessary. When needed to shorten then shortening means, replacing the long phonemes with short phonemes in meaningful words when composing, if it is necessary.


Ƣ ɡ šȡ ĢԨ â, -. , -.


Ţ ġ 󾾡


The meaning of “ the scholars will say, this is the way” means that introducing such changes in literary compositions , will contribute to the strength ( of the language or compositions); that it is not grammatically unacceptable.

The context where the compositions take place without any such changes is already implicated though unsaid explicitly


Notes (Loga)


Teyvacilaiyaar also gives many examples from various verses where such transformations are evident. It must be noted that these phonemic mappings are quite often occasioned by the needs of the CeyyuL, i.e. verses which have strict rules about ciir todai and so forth. A verse as opposed to prose compositions are governed by strict rules, very complex in kind, that accommodate metrical matters as well ‘isai nayam” , the pleasantness to hear. A verse though terse and rule bound, should also sound pleasing to the ears to qualify as something good. There is extensive discussion of these matters in the Chapter on CeyyuL, the penultimate chapter in PoruLatikaaram, the massive third and final book of Tol. and which must have taken millenniums of careful and extensive studies of literary compositions.



The Eccaviyal of Tolkaappiyar


ҽ :




ɢ Ȣ Ƣ

ɿ ҽ



Ǹ ȡ о ѾĢ




ɢ, , Ȣ, Ƣ , Ƣ ҽ -.


The Four Different Ways Meanings are Communicated in Verses.




niraniRai cuNNam adimaRi mozimaaRRu

enanaan kenpa poruLpuNar iyalbee


It is intended to explicate how words are combined literary compositions so that appropriate meanings get communicated.

Meaning:The scholars will say that the different patterns of word combinations in literary compositions that communicate meanings are the four: niraniRai, cuNNam,. AdimaRi and MozimaaRRu.




Tol. isolates here only the non normal ways in which meanings are communicated in literary compositions, it being unnecessary to explicate the natural ways in which meanings are communicated in both existential and literary uses of the language where the meanings are immediately grasped. These will be explained at the appropriate sutras.


Notes (Loga)


Tolkaappiyam, a massive book of three divisions, stands as one of the most significant cultural accomplishment of mankind ( as noted by M.B. Emeaneu). Even Panini’s Ashtattiyaayi, though certainly a great accomplishment in itself, does not have anything equivalent to the PoruLatikaaram of Tolkaappiyam. This difference is very significant for it shows fundamentally different approaches to the grammatical study of languages. In Tol. the central place is given to PoruLatikaaram and where the earlier Collatikaaram and Ezuttatikaram are dealt with as components that would contribute towards the essence of PoruLatikaaram. Thus this third book that deals essentially with Speech Acts ( Ta. kuuRRu) both in literature and existence, provides that for which both grammatical analysis of sentences and the phonological aspect of words and so forth are undertaken. And since speech acts deal with INTENTIONAL meanings that come through linguistic meanings, the whole of the grammatical analysis is executed without ever forgetting that language in ordinary speech contributes towards communicating intentional meanings through word meanings and sentential meanings. In ezuttatikaaram, the phonemes ( ezuttu) are identified only in terms of their significance for meaning constituting, other sounds like grunts, snorts etc rejected as sounds of language i.e. phonemes. In Col. there is the beginning of something like dictionary meaning in the chapter on Uriccol which led to the tradition of NiGandu, explaining rare words in terms of the words that are more prevalent and easily understood. This led later to the tradition of Dictionaries proper as the alphabetically ordered Catur Akarati of Viramaa Munivar ( c. 17th cent AD) In PorLatikaraam, there is an immense complexity in the Theory of Meanings and which includes body language, the meypaadu etc. the CeyyuLial deals with the prosodies and related matters including semiology, the meanings as communicated by analogies metaphors, similes and such other figures of speech. Hinted at but not elaborated sufficiently are meanings of Mantras.

In this sutra and those that follow, what we have is the study of meanings which are ambiguous or uncertain and which require various of kinds TRANSFORMATIONS to disambiguate. The natural order of words are disturbed and transformed and in that specific meanings are communicated. These sutras deal with the various ways in which these transformational processes must be overcome to get at the meanings.


(to continue)6




The Eccaviyal of Tolkaappiyar






Ţ¢ â Ȣ



ɢ ¡ о ѾĢ


- . , ɢ¡ Ţ¡ á á , Ȣ -




The Way of getting at the Meaning by the Method of NiranniRai



niraniRai taanee

vinaiyinum peyarinum ninaiyat toonRi

colveeRu nilai.i poruLveeRu nilaiayal



Meaning: Among the ways (of getting at the intended meaning) enumerated above, the NiraniRai method consists in reflecting on the scattered NPs and VPs collectively and reorganizing them so that each NP is followed with the most appropriate VP.


“ 츢 Ӹ




, , , Ţ, ȡ , էǡ, , š Φ.


“Ũ ɢ” Ƣ, Ũ ɢ ¦á , Ũ, 𨼧ɢ š .

Ţ Ţ ǧ,Ƣ Φ.


adal veel amar nookkki ninmukang kaNdee

udalum irintoodum uuzmalarum paarkkum

kadalumkanaiyiruLum aampalaum paambum

tadamatiyam aamenRu taam


Here the VP’s udalaum ( will suffer) oodum ( will run away) , uuzmalarum (will blossom) paarkkum (will see) must be connected with the NP’skadal ( the seas)iruL ( darkness) aambal ( the lotus) paambu (the snake) in that order and reorganize all as the conjunction of thesentences : kadal udalum ( the seas will suffer ), iruL oodum ( the darkness will run away) ,aambal malarum (the lotus will blossom) and paambu paarrkum ( the snake will see)


When it is said : kodikuvaLai koddai nucupuNkaN meeni”, the NP’s kodi ( creeper), kuavaLai (the flower), koddai (the seeds) go with the NP’s nucuppu ( the waist), uNkaN (the devouring eyes),and meeni (the body) where we have to derive the sentences ‘ kodu nucuppu ( the creeper-like waist), kauvaLaik kaN ( the lotus-like eyes), koddai meeni ( the seed colored (i.e. black) body

Now when you come across such constructions only with verbal nouns and VP’s, then apply the same rules.


Ȣ Ȩ¡ Ţ ġ, ȡ . “Ǣ Ǣ, ” ( ¢- -282) Ƣ, ǢȢ , Ũ ¢, Ǣ , š .


On account ofit being said, only by reflection that the proper meaning emerges, we can consider all constructions where unless properly reorganized the meanings do not emerge clearly.When it is said “ kaLiruG kantum poola naLikadaR kuumbuG kalanum toonRum( Mayilai- pp 282), because thekuumbu( the beams) is not a possible analogy for kaLiRu (the elephant) andthe Kantu (the pole) for the kalan ( the boat) , it is best to reorganize as :the boat-likeelephant,the beams like the wooden pole(for tying down the elephants)


Notes ( Loga)


It looks as though we have here some elementary principles of Transformational Grammar as aspects of disambiguation whereby the HIDDEN and CONCEALED proper form and order of a number of sentences juxtaposed are RECOVERED through appropriate reordering and reorganizing maneuvers. The initial sentence has the surface form of : {{NP1, NP2, NP3, NP4 }& {VP1, VP2, VP3, VP4}} This surface form is semantically ambiguous or confusing and in order to disambiguate and attain an understanding of the intended, the whole sentence has to be CAPTURED semantically the SAME as { NP1&VP1, NP2&VP2, NP3&VP3, NP4&VP4} Here the recovered set is the INTENDED and which has the SAME MEANING as the presented. The order of words in the presented form or the Surface Structure, does not make any sense immediately but which contains within itself the Deep Structure on the recovery of which the meaning becomes quite obvious.


Using this and other truth-functional notions, I have developed Process Grammar as an alternative to the TG Grammar of Chomsky. For example the subject-predicate sentence “The dog barking” when used for assertion, meaning a certain proposition in which the referred object the dog is presupposed as real and assertion being an attribution of something to this and which can agreed upon or not so.


This sentence and hence the assertion can be considered to have been generated out TWO more primitive kinds of assertions : ( That1 is dog ) O (That1 is barking) ------The dog is barking ( the number refers to the identity of the object). As Teyvaciliaiyaar has noted in the chapter on KiLaviyaakkam ( the genesis of speech), unless it is presupposed that the first is true ( as indicated by the symbol “o” ) the assertion “ The dark is barking” will be impossible. In making truth presuppositons central in the generation of subject-predicate and other kinds of complex sentences, Process Grammar differs from not only the TG Grammar of Chomsky but also the predicate Calculus of Russel-Whitehead.




The Eccaviyal of Tolkaappiyar





Ƣ Ȣ н


о ѾĢ



о ¡ áš á н Ƣ Ȣ , -.


. ̾ ̨Ⱦġ Ǧ ¢




The Way of disambiguating called CuNNam


CuNNan taanee

PaddaaG kamainta iiradi yeNciir

Odduvazi yaRintu tuNittanar iyaRRal



This seeks to explicate the attainment of Meaning (of verses) called CuNNam




The CuNNam way of getting at the meaning occurs when in the context of verses with two and eight prosodies( ciir), they are isolated and reorganized properly till a sensible meaning emerges.

The meaning of ‘paddaangku” is as it occurs naturally i.e. without forced distortions to comply with the rules governing the prosodic structure.


( )


ͨ¡ Ũ¨

¡ Ħ



ͨ, , ¡ , н ; Ţš . н .


The following is an example of this cuNNam


Curai aaza ammi mitappa varaiyanaiya

Yaanaikku niittu muyaRku nilaiyenpa

Kaanka naadan cunai


These lines have to be reorganized as ‘ammi aaza’ ( the grinding stone drowning in the waters) curai mitappa ( the dried guard floating in the waters) , yaanaikku nilai ( a station for the elephants),muyaRku niittu(the departure for the rabbits) in order to derive the sensible meaning hidden in these lines.


Notes: (Loga)


Each one of these words: curai, aaza, ammi, mitappa, varaiyanaiya and so forth is called a ‘ciir’ and since each line has FOUR of such ciirs, the adi or feet is said to be of FOUR such ciirs. The lines as given are either confusing or ambiguous and each feet has to be cut and re-assembled with such ciirs so that a sensible meaning is attained.


(to continue) 8




The Eccaviyal of Tolkaappiyar




Ȣ ʿ â



Ȣ¡ о ѾĢ




Ȣ â¡ ʿĨ â , -

ڸ ¡

Ǣ ý ɧ

š ɢ ¡ ŧ

ɡ ġ


, ! ո Ȣ ڸ¡, Ǣ š, šáƢ ɫﺡ Ȣ ٽš Φ.


Ƣ â.




adimaRic ceyti yadinilai tirintu

ciirnilai tiryaatu tadumaa Rummee


It is intended to explain the method of AdiimaRi ( in getting at the meanings)


Meaning: When the ciirs or prosodies remain in proper order but not the feet, where they may be distorted or improperly ordered, we have the method of AdimaRi ( re-ordering the feet) as a way of getting at the meanings


An example:

cuural pambiyac ciRukaan yaaRee

cuurara makaLir aaraNaG kinaree

caara Lenin yaananj cuvanee

caara naada niivara laaRee;


Here the final line or feet, Caara naada! Nii varal aaRee!( O you from the country bordering the hills, the course you would take) has to be brought in as the first line in order to make sense of the whole verse: the course you would take is full small desert streams where thrive intense heat and filled with celestial beautifies who would oppress you and of which I am afraid each you mention of your arrival)


Here there is re-ordering the lines and which is the meaning of adimaRi, so that a collective and agreeable meaning can be derived for the whole of the verse. Notice also in these four lines, there is no disturbing the prosodic order within each line - they all follow the same pattern and because it is distinguished from CuNNam, that was explained above where the ciir itself has to be re-ordered.


ľ ɨ





*¢ں £ Ȣâ

Ũ¡ Ȣ ¡


*¢պ (ɡ )


ľ ɨ о ѾĢ




á Ƣ ھ â 측 -


Additional Comments for the Above




poruderi maruGkin

iiRRadi *yiRuciir eruttuvayirR Ririyun

tooRRamum varaiyaar adimaRi yaana




Within the method of getting at the meaning called AdimaRi, when meaning obtains even with the final ciirin the final line gets distorted, it will not be excluded from this method.


Notes: (Loga)


Within prosodic compositions that come along with definite structural regulations as the ordaining principles of the order in which the LINES or feet occur, there comes to DEVIATIONS from the norm that makes the sense quite obvious and immediate. However the requirements of rhythms, cadence and so forth force deviations from the normal and which have to RESTORED to get the meaning. Thus like in modern art, the combinations of prosodies and lines can occur as distorted and hence meaningless or confusing at first glance but by which they also demand the re-ordering so that the communicated meaning is attained.


( to continue) 9






The Eccaviyal of Tolkaappiyam









ɢ Ȣ

Ƣ ǡ


Ƣ о ѾĢ




Ƣ Ȣ ġ Ȣ, Ƣ ׾, -.




, Ȣ Ƣ ɨ¡, Ƣ 񺣦 оɡ, ɢ Ȣ ɨ¡ ¢ ¢ Ƣ , -


ը Ȣ



Ƣ, ¢ 󿡨 ƢȢ . “ Ũ šĢ ” Ƣ, Ч , ƢȢ . .


MozimaaRRup poruLkooL ( Getting at the Meaning through Reordering Words)


MozimaaR RiyaRkai

Connilai maaRRip PoruL etir iyaiya

Munnum pinnum koLvazik koLAal


It is intended to Explain Getting Meaning Through Reordering Words.




The technique of MozimaaRRu involves reordering the words so that an agreeable meaning emerges.




Since above the reordering the whole feet as well as the prosodic elements (cuNNa MozimaaRRu)  involving the eitght ciirs were explained, and since it is mentioned here only generally word reordering, it should be taken that a reordering of words involve both  within the feet and across the feet.


In the example:


Panmaiyum orumaiyum paalaRi vanta

Annaa laintum muunRutalai yidda

MuunuRak kiLanta uyartiNai yavvee


We have reorder the second feet within  as “muunRu taliyidda annaa laintum’ .

However in


 “aalattu meel kuvaLai kuLattuLa vaali nediya kuraGku”


we have reorder across feet as “aalattu meel kuraGku”.


There are many other such cases.


Notes ( Loga)


When the meaning is not clear collectively and confusing because of inappropriate word order, then the reordering takes place till a suitable sense emerges. This can be within the Feet or Adi or across the Feet. Such a restructuring is different from earlier forms of such exercises where either the whole sentence or the  prosodic elements( ciir) within and across the feet were involved. We have pointed that the first type Niran-niRai is transformational:




It looks as though we have here some elementary principles of Transformational Grammar as aspects of disambiguation whereby the HIDDEN and CONCEALED proper form and order of a number of sentences juxtaposed are RECOVERED through appropriate reordering and reorganizing maneuvers. The initial sentence has the surface form of : {{NP1, NP2, NP3, NP4 }& {VP1, VP2, VP3, VP4}} This surface form is semantically ambiguous or confusing and in order to disambiguate and attain an understanding of the intended, the whole sentence has to be CAPTURED semantically the SAME as { NP1&VP1, NP2&VP2, NP3&VP3, NP4&VP4} Here the recovered set is the INTENDED and which has the SAME MEANING as the presented. The order of words in the presented form or the Surface Structure, does not make any sense immediately but which contains within itself the Deep Structure on the recovery of which the meaning becomes quite obvious.



The disambiguation through just simply word reordering as above does not seem to involve transformational grammar. It is simply recovering the proper order from distortions introduced perhaps because of rules of grammar pertaining to the compositions of verses.



The Eccaviyal of Tolkaappiyam-11





Ѿ â â¡


⧠о ѾĢ




, , ,   Ӿġ, Ţɨ ѾĢ ê ¡ â 츢 â , , -. ԫ .




ta, n_a, n_u, e  enum avai mutalaakiya

kiLainutaR peyarum pirippap piriyaa


It is intended to explain the grammatical essence of glued words (odduppeyar)


Meaning:  The glued nouns and verbal nouns indicative of kinship meanings that begin with ‘ta’, ‘n_a’, ‘n_u’ and ‘e’ do not allow meaningful separation into the basic words - they always function as glued together into a single word.


Ţ Ȣ- (¡)


, , , Ӿġš:


, , ¡, ¡, , , ¢ Ȣ ȡڨ ƨ Ȣ, ɢ ڸ â , , , , , 쨸¢; , , , , , , ¢ .


Notes( Teyvaccilaiyaar)


The pronouns taan, naam, yaam, n_ii, n_iir and n_iiyir, in the context of implicating possessive meaning, that of the sixth case marker, gets transformed by deletion and change in the case of third person pronouns into taman, tamaL, tamar, tamatu and tam while in the context of second person n_uman, n_umaL, emaL, eman, emar, ematu and em.


Ѿ á:


Ȣ , ȡ ڨ Ȣ Ģšȡ , , ; , , ¡ ; , , ; , , ; ӾŨ, ȡ, 󾡨, 󾡨 ġ Ӹ Ȩ  ġ ɨ â, -.


šá ɡ, ý á Ƣ, , Ӿġ¢ â â¡ . , , šɢ, â


 The kinship terms are as follows:


In view of what has been said above, in the three kinds of pronominal senses, the glued words, to indicate the possessive sense of the Sixth Case marker assume the forms: tan_tai, n_n_tai,  en_tai; taay, njaay, and  yaay, for describing the father and mother. Now to indicate the siblings - the elder brother and younger brother,  we have: tammun, n_mmun, and emmun; tambi, n_umbi and embi.  All these words have a plural sense and refer to the interlocutor and the Other. They also do not allow being analyzed into the component words.


Now in view of the Utti - one can also legitimately mention what has remained unsaid - we can also include the Verbal Nouns (tozil peyar) that are causative, the genitive nouns and the quality nouns (paNpup peyar)


Now if asked about the words for Others such piRaL piRan piRar and so forth, it must be noted that these are not glued nouns



Notes (Loga)


There is an important grammatical distinction between Glued Words (odduppeyar) and Compound Words (tokaippeyar) both indicative of the agglutination processes so characteristic of the Dravidian family of languages such as Tamil. Tol clarifies first the Glued Words just before discussing the more widespread Compound Words (Tokaippeyar). The next several sutras are concerned with Compound Words where there are also extensive commentaries. We should note here that Compound Words are quite widely prevalent also in Sk.


The Eccaviyal of Tolkaappiyam-12


Ǣ  Ũ







š Ƣ ħ


ġ о ѾĢ




ڨ Ȣ , Ȣ , Ţ򦾡Ƣ Ȣ , Ȣ , , Ȣ Ũ Ĩ -


The Varieties of Compound Words




veeRRumait tokaiyee yuvamat tokaiyee

vinaiyin tokaiyee paNbin tokaiyee

ummait tokaiyee yanmozit tokaiyen

Ravvaa Renba tokaimozi n_ilaiayee


It is intended to explicate the genesis of the different Compound Words.




The Compound Words are six types and they are as follows: there are compounds implicating the sense of the case markers  (veeRRumai tokai), the sense of analogy or metaphorical meanings (uvamat tokai), the sense of actions and activities (vinai tokai), the sense of conjunctive relationships (ummai tokai) and those implicating as the meaning an  absent object (anmozi tokai)


Notes (Loga)


The Compound Words (CWs), the Tokai Col, are different from the Odduc Col, the Glued Words (GWs) described above. Both are products of the agglutinating processes of word formations quite typical of the Dravidian family of languages. The difference between these two is that while the GWs disallow meaningful analysis into component words, the CWs allow for that and hence can be taken products of transformational processes (mozi maaRRu) rather than something generated just by gluing together.


It is said that a distinguishing mark of CWs is that they function as if a single word (ellaat tokaiyum oru col nadaiya). For example, while the GW ‘entai’ cannot be further analyzed into constituent words as en-tai etc, the VeeRRumai Tokai like ‘padaikkai’ can be. It is the NP “padaiyaip piditta kai’ (The hand that held the sword) that becomes “padaikai” (The sword-hand). Here not only we have the possibility of the analyzing ‘padaikkai’ into the meaningful units of Padai (sword) and Kai (hand) but also see the word as a product of the TRANSFORM of a statement “padaiyaip pidittak kai”. Thus CWs are TRANSFORMS, the MozimaaRRu when that statement happens to be presupposed as true and something further is sought to be asserted about it as for e.g. Padaikkai vedduppaddatu (the sword-hand was cut off)


The transformational notions in the genesis of CWs and their ABSENCE in GWs have NOT been brought out sufficiently well by the traditional commentators and I shall fill up these lacunae in my notes.




The Eccaviyal of Tolkaappiyam-13


Ǣ  Ũ




ڨ ڨ


ڨ ¡ о ѾĢ


(-)  ڨ Ҩ ڨ -





veeRRumait tokaiyee veeRRumai iyala


It is intended to explain the grammatical essence of Case CWs




The Case Compound Words are those which carry the meaning of cases (the markers of which get deleted)




ڨ¢ Ⱦɡ, ڨ Φ ȡ ڨ ¡ . Ӿ ڨ ɢ Ƣ Ţʨ ġ, Ţâ ġ ¢ ¡Ƣ ̦ . ȸ.


On account of being said “in the essences of case meanings”, we have to take compound words that agglutinate but carry the meanings of cases from the accusative to locative as the Case CWs.  Here we have to exclude the Nominative case as in the subject-predicate constructions, they remain enunciated separately.  The Eighth Case of the Vocative stands always expanded. So those that allow for agglutination must be taken as the remaining cases.


The eekaaram here is teeRRa eekaaram, the particle that isolates and emphasizes


Notes( Loga)


The technical term “toku” has to two senses - that of deletion and that of agglutinating. The ‘toku’ in Tokaic col has the second meaning of agglutinating, coming together to form a compound that functions grammatically as a single word. The same sense is communicated by Sk term ‘samaajam” which has the Sumerian ‘sum, sam’ as its root and which means ‘to unite, to come together “ etc. Here it is interesting the Nominative and the Vocative cases are excluded in the sense where we have sentences in these cases they do not allow for the generation of Case CWs. This is quite clear in the case of the Vocative: Kantaa!  Varuka ( Kantaa! Come) This is Direct Speech and does not allow relativization but only reporting such as Kantan was called etc. In the case of the Nominative “naay kuraittatu” (The dog barked) we can say only ‘kuraitta naay’ when relativizing and where it remains free of agglutination.


Now let us consider the sentences: a. teer kutiraikaLaal   puuddap paddna. (The horses were harnessed to the chariot) b. atteeril  avan cenRaan (he traveled in that chariot). Combining both by presupposing the truth of the first we have: avan kutiraikaLaal puuddappaddat teeril cenRaan (He traveled in the chariot harnessed with the horses). This same sentence can be said as: Avan kutirait teeril cenRaan (He traveled in the horsed chariot). Here we have the Case CW ‘kutiraitteer’ and which can be expressed only by the bizarre English expression ’horsed chariot’ and which means ‘horse fitted chariot”. Here we also see the deletion of several particles for the purpose of generating the CW. This process of agglutinating is a characteristic mark of Dravidian languages as is also of Sanskrit and because of which it can also be taken as cognate with Tamil.


In SumeroTamil we see the beginning of such processes and strangely enough in the names of people: utuhegal (utu hegaL: the rising sun), Urnammu  (uur-nambu: the lord of the city) Enuduanna (eeN-udu-anna: the Lord of the heavenly stars) etc


(To continue) 13




The Eccaviyal of Tolkaappiyam-14


Ǣ  Ũ


Notes( Tev. Cont)


  :- . ̨ƨԨ ; ̾¡ ;  ¡ ; ; Ũ¢ɢ Ţơ Ţ; ¡ ; šơ ܨ (Өȧ) 쨸; ̨측; ̾ ; ; ; ŨŢ; ¡째; ܨ . , šǡ ɡ, , Ũ¢ɢ , Ƣ, ھ Ƣ á Ƣ , , š ɡ, Լ, Ũ, ġ .  , šǡ , ̼ξ, Ũ¢ɢ , 츽 , , š, Լξ, Ũ, .


 This is how the Case Compound Words(CCWs) are produced by the process of agglutination:  the phrases padaiyaiyip piditta kai ( the hand holding the sword), kuzaiyaiyudaiya kaatu ( the ears having ornaments);  kutiraiyaaR puddappadda teer ( the chariot harnessed with the horses);  taayodu kuudi muuvar ( the three along with the mother); kadicuuttirattiRku vaitta pon. (the gold set aside for a new waist-chain) , varaiyininRu vizaaninRa aruvi ( the waterfall falling from the hill top); yaanaiyatu koodu ( the tusks of the elephant); kunRattuk kaN vaaza ninRa kuukai ( the owl living in the hills) become respectively : padaikkai, kuzaikkaatu; kutirait teer; taaymuuvar; kadicuuttirappon; varaiyaruvi; yaanaikkoodu and kunRakkuukai.


Now such phrases as ‘nilattaik kadantaan’ ( he who crossed the land), ‘vaaLaal veddinaan’ ( he who cut with a sword), ‘kolaiiku udanpaddan’ (he who succumbed to murder), ‘variyinRu paayntaan’ ( he who jumped from hill top), ‘kunRattukkaN iruntaan’ (he who stayed at the hilltop) and so forth, if they do not implicate the actions but serve as part of the names for those who effected such actions, can become glued words such as ‘nilang kadantaan’, ‘vaaL veddinaan’, ‘kolaiyudanpaddaan’, ‘varaipaayntaan’, kunRattiruntaan where two nouns stand glued to generate a single word. Now such verbal nouns as ‘kaLLaiyundal” ( the drinking of beer), ‘vaaLaal veddal ( the cutting with a sword), ‘kolaikkudampadal’ ( being subject to murder), ‘varaiyiniRu paaytal” (the jumping from the hilltop), ‘maadattukaNiruttal’ ( being in the upper floor) become the verbal nouns : kaLLaiyuNdal, vaaLveddal, kolaiyudampadal, varaipaaytal, maadattiruttal and so forth.


Notes( Loga)


Teyvacciliyaar is at pains to draw a distinction between the Odduppeyar, the Glued Words and VeeRRumai Tokai, the Case Compound words. Just to take a convenient example the genitive construction ‘yaanaiyatu koodu’ , the tusks of the elephant, becomes ‘yaanaikkoodu” where the meaning remains an invariant with changes only  in the morphology. The agglutination process underlying such linguistic habits, deletes the possessive case marker ”-atu” and with that glues ‘yaanaai’ the possessor with koodu, the possessed.  This is a linguistic TRANFORM in which the deep structure is always recoverable as the MEANING remains invariant in such transforms.


Now such phrases as ‘nilaittaik kadantaan” can serve as a NAME and where the seemingly verbal expression is actually a verbal noun that names a person. In other words it is NOT the normal subject predicate expression even though the surface structure may lead us to think so. But the DS is the meaning and it means a NAME - the one who crossed the land and NOT the proposition ‘He crossed the land” . In such cases too, there is agglutination whereby the phrase becomes ’nilaG kadantaan’ ( the land-crosser), but here there is NO transformation as above but rather only deletion operations for the sake of generating Glued Words and not Case compound words.


More to follow


(to continue)14




The Eccaviyal of Tolkaappiyam-15


Ǣ  Ũ


Notes( Tev. Cont)



, š   š ɢ, ڨ š, Ƣġ, §á Ƣ, ġ ; ġ . ȡž , 𦼡 ȡ ɢ, θ . 째 Ƣ 𦼡¡ . Ƣ, ¡ ġ . 측 . “ ڨ ڨ ¢” Ⱦɡ ڨ ƢȢ Ģ, 򦾡 ƢȢ ξ . ¾, ¾ , . .


Now it can be asked: since such CWs come with the deletion of the case markers, can’t they be said to be simply CWs with case deletions?


It cannot be for in accordance with essence enunciated for such CWs, when they end with verbs and nouns, they remain two distinct (unglued words) with only the deletion of the case markers while here we have a single CW with agglutination.


Now it can be said that the CWs with sixth case marker (the possessive) show no difference between CW formed with case marker deletions and semantic fusion (poruL tokai). Well the answer is, it is not so, there is a difference. When said “marakkoodu’ (the tree branch) we have a semantic fusion where two nouns are glued together. But when said “ marattin koodu” (the branch of the tree), it becomes a Morph CW (urubu tokai) and stands as two independent words i.e. without agglutination. We have to understand the difference by noting the intention behind their use. Since it is defined that “VeeRRumait tokaiyee veeRRumai viyala” (The case CWs are essentially of the nature of case notions), we have to note that the case markers are recoverable through transformations and that the agglutination proceeds as a way TRANSFORMING (Mozi MaaRRu) a phrase with deletion of case markers but without change of meaning. The NP “malayatidai” (the space between the hills) becomes “idaimalai” and ‘malayatakam’ (the inside of the hill) becomes ‘akamalai’ and so forth.


Notes (Loga)


The Case CWs are semantically and transformationally understood.  We have implicit in such an account a generative transformational grammar where while the Case CWs in the explicit form and transformed form are said to have the SAME meaning, this may not be so with other kinds of phrases, which are only  superficially i.e. in Surface Structure  are so. There may be deletion of case markers but that in itself would NOT mean the phrase generated is a Case CW. For example “marakkoodu” is lexical fusion of ‘maram” (tree) and ‘koodu’ (branch) but would NOT mean a Case CW if it means simply the branch of a tree as opposed to such branches. The Case CWs must have in their Deep Structure the case morphs and which get deleted only in the transformations in the generation of the Case CW through a process of agglutination and where the MEANING remains invariant i.e. the case meanings are available in both forms.



The Eccaviyal of Tolkaappiyam-16


Ǣ  Ũ


Notes( Tev. Cont)


ȡ, “ ĸ â, Ҹ  ¢ , Ţ Ţ Ǣ, ڿ ȡ, ڿ, 쨸, Ţ   šϾ ” ġ ‘’ ɡ ġ, զ¡ ɢ, š â ġ . š ɦɢ, 󦾡 . š ɦɢ, ȡ . Ţ ȡ ¡ ܼ . ȡ ¡, . š ɡ ȡ. š ġ 񦼡ȡ 򾾡. š զ .


Now leaving aside all that, it can be asked in view of the clause “ Ulakam uvappa valan eerbu tiritaru, palar pukaz njaayiRu kadal kaNdaaGku, oovaRa imaikkum ceeN viLaGku avir oLi, uRun_rt taaGkiya matan udai n_oonRaaL, ceRunart teeytta, cel uRal tadakkai, maRuvil kaRpin vaaNutal kaNavan” where we have several attributives terms are enumerated but all end up as the attributes of KaNavan,  claiming that only two words become fused to become a single Glued Word may not be right. 


The reply to this will be that even when it comes as such unless they are nested together as two words fused into one, such clauses will not be generated. The generation of such clauses is similar to the production of a garland of flowers.  First two different flowers are taken and bound together. Then to this bound flowers another third one is added and all three are bound together. Then a fourth flower is added to these and all of them made into a single strand by joining them together and so forth. In the same way such a long and complex Glued Words are generated pair by pair.


Comments (Loga)


While Teyvaccilaiyaar proceeds to give a very detailed and lengthy explanation of the GENERATIVE processes underlying the generation of such complex NP’s as above,  what is clear is that he is thinking in terms of pair-wise nesting processes and which in such agglutinating languages as Tamil can go on quite indefinitely, terminated only on grounds of rules of communicative efficiency.


Symbolically we say that first a pair X1 is generated with gluing together  ‘x’ and ‘y’ as [x-y].  To generate another and a  more complex phrase we have [X1] bracketed and the third word added to generate [X2] as [z- [x-y]] and so forth. Such a process of NP generation where several attributes are added to a single word, here KaNavan (the husband) shows that the agglutinating process is more specifically that of gluing together and not compounding. Such a process is typical of Tamil and Tamil related Dravidian languages and we can see the beginnings of it in SumeroTamil (and perhaps other African, Nubian Egyptian and other ancient languages related to Tamil). For example we have the phrase  “ ki-bala-gul-gul-lu-ja (Sir. 125) where we have first ‘ gul-gul-lu jia’ (You are the one who (ji-a) destroys (gul-gul-lu)) and which is further expanded as [ ki-bala[gul-gul-la-ji-a] ] where ki-bala means the enemies. This process can go on indefinitely though in Sumerian it is limited to just TWO steps such as this.



 The Eccaviyal of Tolkaappiyam-17


Ǣ  Ũ


Notes( Tev. Cont)



 “ ĸ â, Ҹ  ¢ , Ţ Ţ Ǣ, ڿ ȡ, ڿ, 쨸, Ţ   šϾ ” 񦼡ȡ š ɦɢ, ‘ĸ’ ĸǡ á ‘’ ¦ 򧾡 . ‘ɡ’ Ţ . Ĩ . ֦ɢ, š Ţ Ģ Ģ ɡ, ɡ ¦ , ‘’ Ȣ ¦ â ¦򧾡 .


We have the complex clause with pair-wise nesting  “ Ulakam uvappa valan eerbu tiritaru, palar pukaz njaayiRu kadal kaNdaaGku, oovaRa imaikkum ceeN viLaGku avir oLi, uRun_rt taaGkiya matan udai n_oonRaaL, ceRunart teeytta, cel uRal tadakkai, maRuvil kaRpin vaaNutal kaNavan” , and now we shall  explain how there is pair-wise gluing of words in the generation of this clause. The word ‘ulakam’ (the world) stands as a name for the people at large, and finds its completion with verbal participial  of the ‘ceya’ type i.e. uvappa ( be pleased) . Now in the participial  construction ‘valanaaka” we have the deletion of ‘aaka’ because of the requirements of  prosodic coherence. However despite this deletion at the surface level, it continues to be at the deep and original level (muRpadda nilamai) with no change. You can take this to be something like a dog which has lost its tail and ears but still continues to be called a ‘dog’. The verbal participial ‘ uvappa valanaaka’ along with another present tense verbal participial ‘eerbu’ ends up with the nominal participial (peyareccam) ‘tiritaru’ ( that which moves about)


Notes( Loga)


What Teyvaccilaiyaar  is trying to communicate here, can be understood better in terms of Transformational Grammar where we have Surface Structures (SS) that are transformed versions of the original  Deep Structure (DS). The DS is ‘ulakam uvappa valanaaka’ and its SS is “Ulakam uvappa valan” where we have deletion of ‘aaka’ from ‘valanaaka” . But despite this surface level absence, ‘aaka’ continues to be at the DS level as there is NO CHANGE in meaning. He also implies here that the genesis if verbal and nominal participial constructions ( vinaieccam and peyareccam) are related to another level of such transformational processes that keep the original meanings in tact.



‘á’ ȡ ڨ վ 츽, ‘’ .

Ҹ ¡ . ɢ Ţ ɢ . Ҹ ¦, 񧺡 ȡ Ţɦ¡ Ҹ¢ զġ . â ¦ ¢.


The case-suffixed NP ‘palaraal’ that has ‘aal’ , the instrumental case marker. In accordance with IlakkaNam, or grammatical process of coming with the deletion of the case marker, stands here simply as ‘palar’ .  It also stands as the qualifier of ‘pukazapadda’ (that which is praised). The meaning of ‘adai’ the adjectival qualification is the same as viceedaNam(in Sk), the attributes.  The nominal participial ‘pukazappadda’ , like ‘uncooRu” ( food being eaten) indicates concrete objects and becoming a Verbal Compound , generates ‘pukaznjaayiRu” ,  the words glued into one and made to function like a single word.




Teyvaccilaiyar explains here how the passive  sentence “palaraal pukazappadda njaayiRu” ( the sun that is praised by many) gets transformed into the SS ‘palarpukaznjaayiRu’ through a process not of gluing two distinct names but through another species agglutinating, where the word fusion takes place because of deletion of the case marker “aal” and passive marker ‘padda’. Though  the word fusion is made possible by such deletions they are still available in DS as the meaning remains the



The Meaning of IlakkaNam


We should  also note the special significance of the clause ’”urubu toka varutal ennum iLakkaNam”. Here the word ilakkanam corresponds with the English word grammar but however the ‘grammar’ is understtod as the PROCESSES underlying the generation of such phrases and clause. The deletion of case markers in the transformation of a clause into a Glued Word or Compound Word, involve some linguistic PROCESSES implying that the task of a linguists is to UNEARTH such processes already OPERATIVE in the language. This makes it clear that Linguistics is a Hermeneutic Science where we DISCOVER the processes already at work in the generation of grammatically correct phrases and clauses.


The ilakkiyam, the texts contain the IlakkaNam, the processes underlying the generation of grammatical coherent phrases and clauses. The ilakkaNam are processes not invented, posited hypothetically and so forth but rather noted as there already in the linguistic processes underlying the generation of grammatically coherent sentences.


(to continue) 17


 The Eccaviyal of Tolkaappiyam-18


Ǣ  Ũ


Notes( Tev. Cont)


[Recall: “ ĸ â, Ҹ  ¢ , Ţ Ţ Ǣ, ڿ ȡ, ڿ, 쨸, Ţ   šϾ ” 񦼡ȡ š ɦɢ . . .



We have the complex clause with pair-wise nesting  “ Ulakam uvappa valan eerbu tiritaru, palar pukaz njaayiRu kadal kaNdaaGku, oovaRa imaikkum ceeN viLaGku avir oLi, uRun_rt taaGkiya matan udai n_oonRaaL, ceRunart teeytta, cel uRal tadakkai, maRuvil kaRpin vaaNutal kaNavan” , and now we shall  explain how there is pair-wise gluing of words in the generation of this clause. ]


¢ ġ ¢Ȣ ¡ .  ¢ . Ģ . . Ш ¢ڧ .


All the words before the word NjaayiRu(sun) collectively stand as the attributive terms. The word ‘njaayiiRai’ stands with the accusative case marker ‘-ai’  deleted . Similarly the NP kadalinkaN ( at the sea) stands with the locative case marker ‘-kaN’ deleted. The NP ‘kaNdaaGku’ stands as the NP kaNdaaRpoola and implicates an analogy - as if seeing. And with all these what is communicated is the collective meaning of ‘like the rising sun”


‘’ Ƣ , , ¡ ¦Ŧ š. ¡ .  . . Ţ ¡ . Ţ Ţ ̾ Ȣ Ǣ §á Ţ򦾡¡, ŢáǢ ġ¢. Ţ ɡ Ţ򦾡¡, ŢŢáǢ զ . ¦ ¢.


ĸ Ӿġ Ш ¢ Ţ Ǣ Ǣ ھġ, ŢáǢ¡ ġ Ȣ .



The word ‘oovaraRa’ meaning oovu (riding ) and oziya aRa ( to decline, to bring about cessation, absence)  is the verbal participial of the formula ‘ceya’. Here the noun ’oovu’ serves as  an attributive term. The verbal participial ‘aRa’ competes itself with the nominal participial ‘imaikkum” ( will see). The NP ceeNinkaN ( at the distant sky) standa simply as ceeN with the case marker deleted. It stands as the attributive term for the finite verb ‘viLaGkum”. The attributive tern ‘avir’ ( to radiate intensely) implicating an excess of brightness, conjoins with the noun ‘oLi’ ( light) becoming a verbal complex and becoming ‘aviroLi’  functions as if a single word. And furthermore conjoining with ‘viLaGku” and becoming the verbal compound ‘viLaGaviroLi” functions as a single word. It also becomes the complement for the noun participial “imaikkum’


Beginning with word ‘ulakam’ all these descriptions function to communicate that the Light that shines is like that of the sun, all words stands as complement to the compound ‘aviroLi’


( to continue)18



The Eccaviyal of Tolkaappiyam-19


Ǣ  Ũ


Notes( Tev. Cont)


[Recall: “ ĸ â, Ҹ  ¢ , Ţ Ţ Ǣ, ڿ ȡ, ڿ, 쨸, Ţ   šϾ ” 񦼡ȡ š ɦɢ . . .


տ ‘տ’ . ‘ ‘ ¦. տ ¡ . ‘’ Ģ, ‘’ â վâ . ‘’ . ‘’ ‘’ š áĢ , Ŧġê ġ , ‘ȡ’ ڨ ¡ . ‘ۨ ȡ’ ‘տ’ ¦ ¢.


The cased ‘urun_ar-ai’ stands as ‘urun_ar’ with the case suffix deleted. ‘taaGkiya’ is noun participial. For this the word ‘urun_ar’ stands as the qualifier.  The word ‘matan’  means strength and it stands as ‘mata’ , a qualifier term that has become transformed. Actually it should be ‘matan-ai’ with the case suffix ‘ai’ but which remains deleted. Now ‘n_oonmai’ means infinite patience. Since the anmas who attain the Feet are countless in number, the Feet is understood as of enormous strength capable of withstanding all, we have the case CW ‘n_nRaaL’ generated and which functions as if a single word. It then comes qualified as ‘ matanudai n_oonRaaL’ and serves as the completion of the noun participial ‘urun_artaaGkkiya”


‘տ’   ‘տ’’ . ‘’ ¦. ‘տ’ ¡ . ‘’ . ‘’ ‘’ ‘쨸’ ¢ ¡ , ‘ 쨸’  . ‘; ¦ ¢.


The cased word ‘cerun_ar-ai’ stands simply as ‘cerun_ar’ with the case suffix deleted. ‘ The word ‘teeytta’ is a noun participial. For this the word ‘cerun_ar’ stands as the qualifier. ‘cel’ means heavy rain. ‘uRaztal’ means to be similar. This stands as an analogy for ‘tadakkai’ ( a huge hand) and becomes the CW’ celluraz tadakkai’ functioning like a single word. It becomes then the saturating complement to ‘teyytta’ the noun participial.


(to continue) 19


The Eccaviyal of Tolkaappiyam-20



Ǣ  Ũ


Notes( Tev. Cont)


[Recall: “ ĸ â, Ҹ  ¢ , Ţ Ţ Ǣ, ڿ ȡ, ڿ, 쨸, Ţ   šϾ ” 񦼡ȡ š ɦɢ . . .



‘’ . ‘’ ġ, ‘ ’ ¡¢.  ‘š’ Ǣ, ‘ǢѾ’ ‘¢ ӾȢ Ţ’ ¡ á . ‘Ţ šϾ’ ¡ ɨ¡¢.   ¦á ƨ 𼡸 ȡ ڨ ¡, ‘Ţ šϾ ’ ڨ ¡ á¢.


‘ maru” means some kind of defect. ‘ il’ means the absence here.  Then compound ‘KuRRamil’ becomes the qualifier for KaRpu, chastity as in ‘kuRRamil kaRpu”. ‘vaaL;’ means brightness. It occurs as the adjective for the eyes as ‘oLin_utal’ ( bright eyes) but becoming a noun that indicates the owner, it functions as the name for the woman who has them. Then it becomes the compound of Quality  as ‘maRuvil kaRpin vaaNutal ‘  and functions as if a single word. Then  compounding with the word ‘kaNavan’( husband) it becomes indicative of that which is possessed making the whole a  second case Compound Word as in “ maRuvil kaRpin vaan_tal KaNavan”


‘ 쨸 ’ ڨ ¡ . ‘ۨ ȡ’ ɡ š  ‘ȡ쨸 ’ զ ¡¢. ‘ Ţ ŢáǢ ȡ쨸 ’ ¡¢.


š ¡š . ‘ġ ɨ” ȡɨ¡ â


Then it becomes the  second case compound word   ‘celluRaz tadakkaik KaNavan’ and functions as a single word. Then it glues with ‘matanudai n_oonRaaL’ and becomes ‘ n_nRaaddadakkaik kaNavan’ functioning also as a single word. Then with more embedding it becomes “ceeN viLaGku aviroLi n_oonRaaddadakkaik kaNvan” functioning again as a single word.


Thus in this manner in order to understand the grammatical structure of the various compounds, you have to the  words pair-wise and see how they are glued  together and embedded . This is the meaning of the observation of the Tolkaappiyar that all compounds function as if a single word.


Notes( Loga)


This concludes the very exhaustive exposition of Teyvaccilaiyaar on the generative processes of compound words, which also brings together a variety of agglutinating processes. With the exception of Odduppeyar where nouns are simply glued and which can be more than two, and certainly the echo words where again there can be more than two, in all other cases two words are brought together and made into qualified nouns and so forth and then embedded to generate lengthy compound words where, despite the length it functions as if a single word. This functioning grammatically as if a Single Word is one of the most important criteria for identifying the Compound Words.



(to continue)


The Eccaviyal of Tolkaappiyam-21


The Analogical Compound Words






򦾡 ¡ о ѾĢ


- :  Ҩ ¡ -.


Ũ Ƣġ , ɡ , Ţɡ . ɡ ‘Ч ; ȡ Ũ . Ũ , Ţʨ ġ Ȣ ¡ -


‘Ģ ’ ‘ ’, ‘ ’. ‘ ɢ’ ‘Ģ’ ‘Ƣ ’ ‘¢’ ‘ɢ’ .


Sutra 408


Uvamat tokaiyee uvama iyala


Intention: It is intended to explicate the grammar of Analogical Compound Words


Meaning:  Those phrases that have analogy as the essence are called Uvamat Tokai.


Notes( Tey)


The analogy comes in terms fo ” This is like that” and where the analogy may pertain to activities, effects, shape and attributes( or color). The words that name an object that is compared and the analogy itself are uttered without any separation between them and hence as if a single word.


The phrase ‘puliyanna paayttuL’ ( the leap like that of the tiger) stands as the Compound Word “pulipaayttuL”.  Similarly  for ‘mazaiyanna vaNkai” ( the great hand that showers like rain) becomes ‘mazai vaNkai”; ‘tudiyanna idai” ( the waist like the hand-drum) becomes ‘tudiyidai’, ‘ponnanna meeni” ( the body like gold) ‘ponmeeni’ and so forth


Ţâ, ‘Ģ ¡ ’, ‘ƨ ¡ ’ Ţâġ ɡ ž ġ Ũ Ȣ, ڨ ¡ ɢ, . ‘ ’ Ƣ, Ţâ¡ ɢ ը ξĢ, š 츢 ȡ. ɡɧ, Ҩ šá¢.


Now when these  words are expanded they can be as ‘puliyai okkum pyaattuL’ ( the leap that resembles that of the tiger), “mazaiyai okkum vaNkai’  (the generous hand that resembles the heavy fall) and therefore it can be said that there are no analogical CWs as such as there are comparisons that involve the second case marker ‘ai” and hence they can be understood as a species of Case CWs. There is some truth in it.  However it can be expanded into a form without the second case morph ‘mazaiyanna vaNkai”  where the named object is directly compared with another on the basis of some similitude. It is for the purpose of maintaining this distinction that Tol. emphasizes that such phrases are Uvamat Tokai only when taken as analogical in essence.



Ţ Ⱦɡ Ũ  Ȣ . ‘Ӹ’ ‘â” . Ÿ ȡ ɢ, Ÿ ¡, 򦾡 ¡ Ũ . 򦾡 ڨ ¡ žɡ, ɡ ڨ .


Now as Tol. talks of the ‘science of analogy’ as well as  Analogical CWs, we can also include phrases where the compared and the comparison are reversed as in “mukat taamarai’ ( Lotus-face) ‘kaimmaari” ( rain-hand). Now can they be Figurative Images ? Since Tol does not mention ‘Uruvakat Tokai’ and furthermore they do not also belong PaNput Tokai ( Attributive CWs), we can take it that they belong specifically to analogies. And since these kinds of phrases are studied close to Case CWs, there must be some similarities between them


Notes( Loga)


Analogical thinking is very primitive thinking and one can say Indian Logic begins first as studies of analogical reasoning that Tol devotes a chapter to it. It is also the beginning Indian poetics. The analogical reasoning is taken as valid form of reasoning in Nyaya Logic as well as in many other schools of Indian Logic. Now ‘uruvakam’ belongs to Ani (beauty, ornamentation), one of the beauties of poetic language and which is very extensively studied in many books like TaNdi AlaGkaaram and so forth. Uruvakam literally means IMAGING, the casting an object into an aesthetically appealing object like the face of young lady onto aLlotus, the Moon and so forth. Thee is NO comparison but only a recasting an object into an appealing image that expresses the FEELINGS of the observer more than analogical reasoning that borders the logical.


(to continue) 21



The Eccaviyal of Tolkaappiyam-22


Ţ򦾡 :  The Verbal Compound Words


 : 409


Ţ¢ ̾


- Ţ¢ ̾ ġ -


Sutra 409


Vinaiyin tokuti kaalattiyalum


Meaning: The verbal CWs function as one but with a verb that carries tense notions


Notes (Tey)


Ţ ġ Ƣ Ģ, ġ ¡¡, 𼡾 Ţɦ, ¦ , Ţɦ Ţġ ¡¡, Ţâ ¦ . ը Ţâ . Ƣ, Ţʨ ġ ڨ𼾨 Ţ򦾡 ȡ.


- ¡, ¡; , ; â š, â š; , Ţ; , ; , ¡, ɣ, â, Ģ, . .


The VP indicate the action, time and gender differences. Among these the gender indicating  VP  are finite verbs and hence stand unable to become the CWs. Now among the non-gender indicating participial forms of both kinds, the verbal and nominal, as the verbal participial do not complete themselves other than with a finite verb, it is the nominal participial that enters into compounding with deletion( and expansion) of some infixes.  Since the action and  object  named are uttered without pause between them and hence as if a single word Tol calls them Vinait Tokai.


Examples : The phrase ‘kollum yaanai’ ( the elephant that kills) and ‘konRa yaanai’ ( the elephant that killed) become ‘kolyaana’i. The  phrases ‘uNNum niir’ ( drinking water) ‘uNda niir’ ( the water that was drunk) become ‘uNNiir’. The phrase ‘ari vaaL’ ( the sword that cuts) and ‘arin_ta vaaL’( the sword that cut) becomes ‘arivaaL’ .  The phrase ‘ cellum idam’ ( the palce to go0 and ‘cenRa idam” ( the place on went) become “celidam”. The phrases  ‘uNNum pozutu” ( the time for eating) and ‘uNda pozutu’ (the time food was taken) become “uNpozutu’. Similarly  the phrases “ kollu kolai” ( the killing that is murderous) and ‘konRa kolai” ( the murder that killed) become ‘kol kolai’. The same kind of analysis goes also for many other such compounds.


(to continue)22



The Eccaviyal of Tolkaappiyam-23


 The Attributive Compound Words




Ţ Ţ ͨ¢

̽ ѾĢ

¢ Ц

Ţ ȡ




ͨ Ȣ, , Ǣ ̽ վ Ȧ¡ Ţ Ԩ ġ ¡ -


Cuuttiram : 410


VaNNattin vadivin aLaviR cuvaiyin

Anna piRavum atankuNam n_utali

Yinna tituvena varUu miyaRkai

Yenna kiLaviyum paNbin Rokaiyee




All  compound words where in the nature of describing the various attributes like color shape measure taste and so forth pertaining to an object are called Attributive Compounds.


Notes( Tey)


â ̾, Ĩ( Ĩ), , Ţ ̾, Ĩ, , .


Ⱦɡ   Ӿġ¢ .


Notes( Tey)


The statements ‘ The horse is black” ‘The plate is round” “ the staff is long” “ the sugarcane is sweet” and so forth become the Attributive Compounds “ the black horse” “ the round plate’ “the long staff’ “the sweetish sugarcane ‘ and so forth


Now as it is said that similar terms can also be included, we can include such terms as TaNNiir ( cold water) Venniir ( warm water) and so forth


Notes (Tey)


, â ̾ , ¦ š , Ȣ ŢȢ ʾġ Ţ ɢ, š Ţâ, 򾾢 ھĢ ¦ š ŢȢ ġ Ţ ¡ šá¢ .


Notes( Tey)


Now it may be contended that such compounds can be included within the category of  Verbal Compounds instead of a different category by virtue of the fact that such compounds can also be expressed as Kariya Kutirai, Nediya Kool ( where the attributive term takes the adjectival marker ‘-a” and a nominal participle forms) , it should be noted, as mentioned in  Ezuttatikaaram, that the Attributive Compound words do not have a single formula.  Even such Nominal Participle forms are implied Verbal Compounds and Tol enumerates them along with Verbal Compounds.


Notes( Tey)


ըԨ ̾æ ڨ ¡ ɢ, Ц Ȧ¡ Ţ վ . š ڨ ¡. ڨ , , Ţ , Ҩ ҨԨšĢ, ɢ¡ ռ Ȣ.


Notes( Tey)


Now it may contended that such compounds can be understood as Horse with Blackness and so forth  and hence as Case Compounds, it should e noted that there is no more the sameness in meaning - when it is said “this is that” there is a qualification of a named object with a attribute and hence not a case compound at all. The Case Compounds the Analogical Compounds the Verbal Compounds and the Attributive Compounds all have a partial similarity and because of which PaaNini called them TaRpuruda Samaasam.



Notes (Tey)


ɢ ‘ Ц Ţ” Ⱦɡ, Ȣ վȢ, âɡ Ȧ¡ Ţ , ¡ . š , . , â š, ɢ վĢȢ, Ĩ¡¢, Ӿ Ţ â, ¡ .


Notes( Tey)


Now in view of the general grammar of phrases of the sort “this is of this sort” we can also include within this category not only those with ascribing attributes but also those which have a qualifying  kind of term within the complex and hence such compounds as :”vaaNikac Caattan: ( Cattan the trader) and Caaraippaambu ( the snake that crawls) that are agglutinative. Even two nouns where we have a subject-attributive constructions but not agglutination such as Ayan Caattan ( the Cattan the shepherd) , Aciriyan Nallantuvan ( the Teacher Nallantuvan) and so forth where the subject terms are further qualified can also be taken as a species of such Attributive Compounds.


(to continue)