[Full title: Ideophones in Tamil: a Historical Perspective on the X-eṉal expressives (Olikkuṟippu Āṟṟuppaṫai)]

[…] taññeṉṉun tī yāṇṭup peṟṟāḷ ivaḷ? (Kuṟaḷ 1104) "A fire which says cold, where did she obtain it? "1

Prelude

Although the present essay is devoted to the presentation of a feature that once appeared to me as specific of Tamil —compared to languages I had previously been acquainted with —one of the quickest way to explain its topic to the reader might be, paradoxically, to start with a quotation from a book describing one feature of Japanese2:

The vocabulary of Japanese includes a large number of words that may be broadly characterized as iconic or mimetic, that is, words whose phonetic form is felt by speakers to be imitative of natural sounds, actions and states. […] it is in some cases difficult or impossible to provide even approximate English equivalents for these terms; moreover, those English expressions which do seem partially equivalent to Japanese iconic forms are often considered by native speakers of English to be childish or informal and of marginal lexical status. […] When we examine the occurrence of iconic words in both spoken and written Japanese, we must conclude that such words enjoy a position in the language that is anything but marginal.3

In order to characterize the domain under consideration here,4 I shall start with a list of examples, drawn from various periods of the history of the Tamil language. All contain expressions which I shall designate as instances of the X-eṉal template, and which could be literally translated as "to say5 X", because they contain the quotative verb eṉal combined with various base elements, here referred to as X. Many of these idiomatic expressions are used to express some "quality", in a broad sense of that term. The examples are enumerated here in reverse chronological order.

1. poṭave paḷapaḷaṅkutē "The saree is glittering" (lit. "the saree SAYs paḷapaḷa") [Folk song]6

2. puṭavai paḷapaḷaveṉkiṟatu [H-Tamil equivalent of the former L-Tamil example]

3. kacakaca-ṉṉu irukkīṅka "You are sticky (with sweat)" (lit. kacakaca say-conv Be-pres.2pl.)7 [Film script]8

4. Āṭai kāyāmal picupic-eṉṟ-irukkiṟatu9 "The garment does not dry and is moist "10

5. avaṉ viral-ellām picupicuṉṉu iruntatu "All his fingers were moist "11

6. nīr taṇṇeṉṟatu "the water was fresh" (lit. water cold-said) (Iḷampūraṇam, TC416 i)

7. vaṟkeṉṟu irutti12 "you are harsh" (Tēvāram, 7-50, 4)

8. puṟkeṉṟa kaṇṇum "mes yeux sont affaiblis" (Kuṟaḷ, 1261)13

9. tuṭkeṉṟaṉṟu eṉ tūu neñcam "My poor heart missed a beat "14 ("SAID tuṭku") (Ku, 157.1-2)

10. kāṉam-um kammeṉṟaṉṟu "the forest is hushed to silence "15 ("SAID ´kam´") (Na, 154.1)

In this list, examples (10) and (9) are from two classical anthologies belonging to the most ancient layer of Tamil literature, whereas (8) belongs to a less ancient work. Item (7) is even later, being from the bhakti period, and (6) is from a medieval commentary. The remaining examples belong to the contemporary period, but some of them are from the colloquial L-Tamil variety whereas others are from the more formal H-Tamil.16 All in all, 8 different X-eṉal expressions are represented here because (4) and (5) on the one hand, and (1) and (2) on the other make use of the same item. When one spans such a long period of time and such a variety of style, there are of course huge differences in morphology and in syntax, as must be evident here, but my focus will not be on that aspect. To mention just one specific point, the dominant behaviour of X-eṉal expressions in the case of contemporary Tamil seems to be their appearance in ad-verbal position and not as main predicate as in the examples given here.17 The eṉal component appears then mostly in the form of the converb (or viṉai eccam): eṉṟu for H-Tamil,18 and -ṇṇu for L-Tamil. These forms are characterized by Asher [1982: 242], in his description of Contemporary Tamil, in the following way (his characterization being accompanied by a list of 29 items):

[…] uninflected onomatopoeic forms […] normally followed by the quotative particle -ṇṇu, the whole expression generally functionning as an adverb. […] Reduplication is common. […] The set is at least to some extent an open one.

Asher seems to be the first to have used the term "ideophone" in connection with Tamil19— this is how he designates these items — and I have followed him, although the class of items which I examine is broader than the one he examines, both from a chronological and from a syntactic point of view.20 Other terms are also met with in the literature, but I feel it is more appropriate to discuss terminological questions after exploring the linguistic phenomenon under study, which has here been defined by a formal criterion—namely being an instance of the X-eṉal template.

 

To remain in the preliminary remarks, another question that might arise is how pertinent it is to insist on giving "literal" translations alongside the global translations. An answer valid for all periods is impossible, but I shall simply remark that it makes sense explicitely for at least one of the examples mentioned, because (6) "nīr taṇṇeṉṟatu" has been adduced by the commentator Iḷampūraṇar as an illustration for a sūtra asserting that there are occasions when one mentions as saying (something) those things that are not reputed to speak (eṉṉā marapiṉa v-eṉak kūṟutal): this sūtra is found in the Book on Words, the second book of Tolkāppiyam,21 the most ancient Tamil grammar, and the commentators differ as to what is a fitting example. A more recent one, Cēṉāvaraiyar, reproaches Iḷampūraṇar, his predecessor, for his choice of (6), arguing that the eṉṟatu which we have here does not mean "it said" (as it would if it had its primary value). A temporal distance of one or two centuries, of course, prevents Iḷampūraṇar from answering, but at least we learn a few things in this one-way dialogue which took place between scholars from the XI^th and XIII^th centuries commenting on a text generally considered to be, at least, not later than the V^th century:

• there are phrases in Tamil with a word which seems to mean "to say".
• there are indirect ways of describing the properties of things, and these ways are sometimes more eloquent (or expressive)

Having concluded this prelude with the conviction that there is some degree of naturalness in examining the elements falling under the X-eṉal template, my goal, in the following sections, will be to examine the various X items with which Tamil, in the course of its history, has been constantly enriching its vocabulary, in a never-ending quest for expressivity.

X-eṉal expressions inside the Tēvāram

While studying a linguistic phenomenon from a general point of view, a possible convenient observatory is a corpus of texts that belongs to the middle period of the recorded history of the language in question. It allows one to see a state of affairs in a temporal dynamic perspective because it is thus possible to make comparisons both with the earlier periods and with the more recent ones.22 For that reason, the present study, after the prelude part, will have as its real starting point a corpus of Later Classical Tamil texts nowadays called Tēvāram (TEV).23 Its vocabulary contains a number of those items that have been presented in the prelude part, example (7) being one of them, and if we try to enumerate them exhaustively, we come up with the following chart (chart 1):

 

X-eṉal	Tēvāram example (and ref.)	Other ref.
S124: kalleṉal	kalleṉa uḻitarum "who are wandering making a big noise" [2.91.10] (VMS)
S3: taṇṇeṉal	kāra mutu koṉṟai kaṭināṟu tañ eṉṉa "the fragrance of the old koṉṟai flowers that blossom in winter to be cool "25 [4.19.9]
S20: tuṇṇeṉal	tuññeṉṟu eḻuntiruntēṉ [6.13.6] "I woke up startled" (VMS)	2.3.6, 2.4.5, 3.81.7, 5.19.4, 5.77.5, 7.76.2
S34: cikkeṉal	teruṇṭa vāy iṭai nūl koṇṭu cilanti // cittirap pantar cikkeṉa iyaṟṟa [7.66.2] "The skillful spider wove a wondrous dense canopy with threads from its mouth [to shelter you]" (Shulman [1990: 424])	4.59.1, 4.59.2, 4.59.7, 5.31.8, 5.39.2, 5.71.9
S36: kiṇiṉeṉal	maṇi kiñiṉeṉa varu kuraikaḻal cilampu ārkka [2.104.1] "the tiny bell in it to make a sound resembling ´kiṇiṉ´ and the anklet to produce a sound" (VMS)
S37: kiṇṇeṉal	kiññeṉṟu icai muralum tiruk kētāram [7.78.7] "T. where the sweet sound of ´kiṇ´ is produced" (VMS)
S38: vaṟkeṉal	vaṟkeṉṟu irutti [7.50.4] "you are a little bit hard" (VMS)
S39: tiṇṇeṉal	tiññeṉ viṉaikaḷait tīrkkum pirāṉ [4.90.6] "the master will remove the very strong and irresistible karmam-s (of devotees)" (VMS)	1.78.1, 2.4.5, 4.2.1, 7.45.2, 7.46.9
S40: veṭṭeṉal	veṫṫeṉa pēcaṉmiṉ [7.44.3] "don´t talk disparagingly" (VMS)
S4126: miṉṉeṉal	tikaḻtaru mārpiṉil miṉṉeṉa miḷirvatu ōr araviṉar [3.85.6] "has a cobra which is gleaming like lightning on the shining chest" (VMS)
S42: mūceṉal	mūceṉum [7.36.2]" [the cobra] makes a hissing sound" (VMS)
S43: nerukkeṉal	nerukkeṉa nirutta viralāl [3.68.8] " (who crushed) by the toe which practises dance, (so as) to produce the sound ´nerukku´" (VMS)
S44: mummeṉal	mummeṉṟu icai mural vaṇṭukaḷ [1.11.3] "the bees which hum like the sound ´mum´" (VMS)
S45: ummeṉal	ummeṉṟu eḻum aruvittiraḷ [1.13.3] "the collection of streams which produces a sound ´hum´" (VMS)
S46: mommeṉal (Var. ommeṉal)	tāmommeṉap paṟai [3.102.8] "the drum which makes a sound like ´mom´" (VMS)
R1: kaṇakaṇa-veṉal	katam miku karu uruvoṭu ukir iṭai vaṭavarai kaṇakaṇaveṉa [1.21.7] "the mountain Meru, which is in the North, to make a sound resembling ´gaṇagaṇa´27in its nail assuming a great form with excessive anger" (VMS)
X-eṉal	Tēvāram example (and ref.)	Other ref.
R2: kalakaleṉal	ayal nilavu mutu vēy kalakaleṉa oḷi koḷ katir muttam [3.69.5] " [Kāḷatti´s hill, rich in…] bamboos that shed sparkling pearls" (I. Peterson [1989: 174])28
R3: neṟuneṟeṉal	arakkar kōṉai neṟuneṟeṉa aṭarttiṭṭa nilaiyum [6.18.11] "the state of pressing down the king of arakkar… to produce the sound ´neṟuneṟú" (VMS)
R4: kūkūveṉal	kūkūveṉa aḻaikkum [7.50.9] "where [the black cock…] sends its call" (Shulman)
R5: āvāveṉal	"āvā!" eṉa arakkaṉ alaṟa aṭarttiṭṭu [1.89.7] "having pressed down the arakkaṉ to roar saying ´alas´" (VMS)
E1: tiṭukumoṭṭeṉal	tiṫukumoṫṫeṉak kutti, kūṟai koṇṭu [7.49.1] " [Vaṭukar tribesmen…] who jump on travellers with fierce yells, stab them, and rob them of their wealth." (I. Peterson [1989: 179])

The 21 expressions which are listed in this chart have, according to my initial assumption, a common morphology, but the French (or English) reader who considers them one by one will be tempted to say that some of them (like S36, S42, R4, etc.) are onomatopoeic and that R5 should be called an interjection. However, categorizing other items might appear to him more problematic. I shall postpone the discussion on whether it is acceptable to group all these items together until a later stage; some arguments for an answer will be given when we discuss the formulation found in Tivākaram.

 

The initial letter (S, R or E) of the labels in the first column indicates subcategories inside the set of all X-eṉal expressions. Those labels that start with "R" indicate "R-items", having internal reduplication, as what we see in kaṇakaṇa. Similarly, we talk about "S-items" for those that are simple. Lastly, we have to make a provision for "E-items", a category which has become much more visible in later Tamil, as we shall see, although it is almost not represented here.29

The Tēvāram data in a dynamic perspective

To put things into perspective, the data which have been given for the Tēvāram in chart 1 have to be compared with the data from other texts. As of today, the most comprehensive source of information about Tamil vocabulary (without distinction of period) is contained in the more than 4400 pages of the Tamil Lexicon (7 vol.). In a recent study, I have made a preliminary examination of that data.30 The distribution of the items is summarized by the Chart 2 a (see below) where 613 X-eṉal expressions listed by this dictionary have been taken into consideration. The contents of S and R columns have already been illustrated by items from Chart 1; the E and IP columns will be explained shortly. As for the lines of the chart, they reflect the distinction between items "without suffix" (like for instance S1, S3, S20, etc.) and items "with suffix", like for instance S38 (vaṟkeṉal) and S43 (nerukkeṉal), which both contain the "-ku" suffix.31

Zoom in

Download the image (.jpg, 17.3 KB)

 

Chart 2 a: Distribution of X-eṉal expressions in the Tamil lexicon32-33-34

The main problem with the data found in the Tamil Lexicon is of course that it does not give us a precise picture for any given period, because its scope is global. It thus needs to be completed by explorations limited to specific samples, such as our Tēvāram. On the model of Chart 2 a, we can draw a Chart 2 b that will give us a profile of the Tēvāram corpus:

 

Number of items	S (simple)	R (redup.)	E (echo)	Total
without suffix	12	5	1?	18
with suffix	3			3
Total	15	5	1	21

This profile can then be compared with profiles built for other homogeneous corpora of texts. To give just a hint of what the result of a comparison with contemporary Tamil might be, I have made a casual search through the scripts of some recent films: Alaikaḷ ōyvatillai (AO, 1981) and Muntāṉai muṭiccu (MM, 1983). The items found are (in the Tamil alphabetical order):

kacakacaṉṉu (MM, p. 16),
korkorṉuṭṭu (MM, p. 49),
koḻakoḻaṉu (MM, p. 19),
caṭṭupuṭṭuṉṉu (MM, p. 33),
carruṉṉu (AO, p. 4)
tirutiruṉṉu (MM, p. 15),
toṇatoṇaṉṉu (MM, p. 15),
nēkkuṉṉu (AO, p. 5).
nainaiṉṉu (MM, p. 20),
nainainainaiṉṉuṭṭu (MM, p. 35),
naiyinaiyiṉṉuṭṭu (MM, p. 49),
vatavataṉṉu (MM, p. 10),
veḷḷaiveḷērṉu (MM, p. 19),

 

The profile of the distribution is as follows:

Zoom in

Download the image (.jpg, 10.08 KB)

 

Chart 2 c: Distribution of X-eṉal expressions in 1980 s film scripts

In these examples, the eṉal component appears under various forms: -ṉṉu, -ṉu, - ṉṉuṭṭu, -ṉuṭṭu and the X component belongs to one of 4 types: S, R, E or IP. Since the S-type and the R-type35 have already been explained, it will be sufficient to point out the two remaining types: the E-type (with echo reduplication), which is illustrated by caṭṭupuṭṭu-ṉṉu, and the IP-type, which is illustrated by veḷḷaiveḷēr-ṉu. It is a feature of contemporary spoken Tamil, that R-type expressions are more frequent than S-type and E-type. As far as IP-type expressions are concerned, they seem to occur only for the small set of colour terms. A detailed study of the R-type36 has been made in Malten [1989], but some of the items which we have just found in these contemporary films dialogues are still missing from it.37 No similar detailed study seems to be available for the two other types. Tamil is not as well studied as Japanese in this respect.

The Tēvāram data compared with its past

Charts 2 a and 2 c have given us clues concerning the future developments of the Tēvāram data. We shall now examine the evidence which can be gathered concerning the past, where the Tēvāram has its roots, and which is presented in chart 3 immediately below. This is a list of all the X-eṉal expressions which I have been able to find inside four Early Classical Tamil anthologies: Akanāṉūṟu (Ak), Puṟanāṉūṟu (Pu), Naṟṟiṇai (Na) and Kuṟuntokai (Ku). They have been ordered by decreasing frequency, starting with the most frequent of them all, kalleṉal, which is met with 45 times, and the columns on the right hand-side give the number of occurrences in each of the 4 anthologies.

Zoom in

Download the image (.jpg, 77.17 KB)

 

Zoom in

Download the image (.jpg, 114.19 KB)

 

Zoom in

Download the image (.jpg, 125.49 KB)

 

Chart 3 a: X-eṉal expressions found in Early Classical Tamil (ECT) 38-39-40-41-42-43-44-45-46-47-48-49-50-51-52-53-54-55

Zoom in

Download the image (.jpg, 16.02 KB)

 

Chart 3 b: Density of X-eṉal expressions in ECT56

As will be clear from the examples and from the notes that accompany them, many of these items would deserve a full-fledged study, to try to get as close as possible to their original intended meaning, and many of the available translations are probably not precise enough. However, for the time being, I shall only remark that in ECT, X-eṉal expressions were often inflected,57 and could be found in ad-nominal position (verbal root or relative participle forms), in ad-verbal position (converb forms) or in predicate position (finite verb form). This inflectional freedom was reduced in the course of the history of Tamil, and this is why the modern heirs of these items could be characterized as "uninflected" by Asher [1982: 242].58

Establishing the X-eṉal spectrum of specific texts

The descriptive (DLH) task which was started in the preceding sections could, in a slightly simplified way, be explicitely summarized in the following way:

Provided that they fulfill a morphological condition, which has been stated as "being an instance of the X-eṉal template", and provided that they are recognized as "expressions" by" (extended) native speakers of classical Tamil",59 some strings belong to the "spectrum of X-eṉal expressions" for a given text.

Using a functional notation, we could now re-state the first results of our investigations:

(11) X-eṉal-Spectrum [Ak+Pu+Na+Ku] = {kalleṉal (S1), pulleṉal (S2), …, kiṭiṉeṉal (S22)} (12) X-eṉal-Spectrum [Tēvāram] = { kalleṉal (S1), taṇṇeṉal (S3), S20, S32, S44, …, E1}

Such a task can be accomplished on other texts, and I briefly give here now some of the results obtained by me, singling out the items not met with so far:

 

Work	X-eṉal-Spectrum (conservative part)	X-eṉal-Spectrum (innovative part)	Density60
Patiṟṟuppattu (PATI)	kalleṉal (S1), pulleṉal (S2), taṇṇeṉal (S3), iḻumeṉal (S8), aiyeṉal (S10)		0.47 %
Aiṅkuṟunūṟu (AI)	kalleṉal (S1), pulleṉal (S2), taṇṇeṉal (S3), naḷḷeṉal (S5), olleṉal (S9), tuṭumeṉal (S17)	nalleṉal (S24)61	0.6 %
Pattuppāṭṭu (PP)	kalleṉal (S1), pulleṉal (S2), taṇṇeṉal (S3), naḷḷeṉal (S5), oyyeṉa (S6), immeṉal (S7), iḻumeṉal (S8), olleṉal (S9), katumeṉal (S11), melleṉal (S13), ñerēreṉal (S16), tavveṉal (S18)	metteṉal (S25)	1.18 %
Paripāṭal62 (PAR)		kommeṉal (S26)	0.05 %
Kalittokai (KAL)	kalleṉal (S1), pulleṉal (S2), taṇṇeṉal (S3), paiyeṉal (S4), oyyeṉa (S6), immeṉal (S7), olleṉal (S9), katumeṉal (S11), teṟṟeṉavu (S14), valleṉal (S15)	ammeṉal (S27), taiyeṉal (S28)	1.02 %

 

Work	X-eṉal-Sample (conservative part)	X-eṉal-Sample (innovative part)	Density
Kuṟaḷ (TK)	pulleṉal (S2), taṇṇeṉal (S3), katumeṉal (S11), tavveṉal (S18)	puṟkeṉal (S29), poḷḷeṉal (S30)	0.23 %
Nālaṭi63 (NĀL)	kalleṉal (S1), paiyeṉal (S4), immeṉal (S7), olleṉal (S9), teṟṟeṉavu (S14), kāreṉal (S22)	kaṫukkeṉal (S31)	0.44 %
Cilappatikāram (CIL)	kalleṉal (S1), pulleṉal (S2), oyyeṉa (S6), iḻumeṉal (S8), olleṉal (S9), tuṇṇeṉal (S20)	cevveṉal (S32)	NC
Maṇimēkalai (MAN)	kalleṉal (S1), pulleṉal (S2), taṇṇeṉal (S3), iḻumeṉal (S8), aiyeṉal (S10), teṟṟeṉavu (S14), veḷḷeṉal (S19)64	iḻiṉeṉal (S33), cikkeṉal (S34)	NC
Cīvakacintāmaṇi65 (CIV)	kalleṉal (S1), pulleṉal (S2), taṇṇeṉal (S3), paiyeṉal (S4), oyyeṉa (S6), immeṉal (S7), olleṉal (S9), melleṉal (S13), tuṭkeṉal (S21)	kommeṉal (S26), ammeṉal (S27), puṟkeṉal (S29), poḷḷeṉal (S30), iḻiṉeṉal (S33), pommeṉal (S35), kiñiṉeṉal (S36)	NC
Work	X-eṉal-Sample (conservative part)	X-eṉal-Sample (innovative part)	Density
Tēvāram (TEV)	kalleṉal, taṇṇeṉal, tuṇṇeṉal,	cikkeṉal (S34), kiṇiṉeṉal (S36), kiṇṇeṉal (S37), vaṟkeṉal (S38), tiṇṇeṉal (S39), veṭṭeṉal (S40), miṉṉeṉal (S41), mūceṉal (S42), nerukkeṉal (S43), mummeṉal (S44), ummeṉal (S45), mommeṉal (S46)	NC
Tivviya Pirapantam and commentaries (TIV)66		metteṉal (S25), puṟkeṉal (S29), tiṇṇeṉavu (S39), veṭṭeṉal (S40), cilleṉal (S), cummeṉal (S), koḷḷeṉal (S), civiṭkeṉal (S), kaṇakaṇeṉal (R), neṟuneṟeṉal (R3), moṭumoṭeṉal (R), mocumoceṉal (R), coṭṭuccoṭṭeṉal (R), kāṟukāṟeṉal (R), kīcukīceṉal (R), parakuparakeṉal (R), mocukumocukeṉal (R), kaṇārkaṇāreṉal (R), cīṟupāṟeṉal (E), calārpilāreṉal (E)
Tiruvācakam (VĀC)	tuṇṇeṉal	mukēreṉal (*T12=S), [naṟumuṟuttal]67 (*E)
Īṭu (ĪṬU)		metteṉal (S25), civīleṉal (S), cilukucilukeṉal (R), pacukupacukeṉal (R), paṭaṉpaṭaṉeṉal (R), naṟukumuṟukeṉal (E)

It is of course a complex task to analyse this data. One possible rough distinction between these LCT texts could be tentatively established between those that are conservative in their vocabulary, not containing many new items, and those that are innovative, probably reflecting more closely the spoken language of their period. For instance Cilappatikāram could be said to be conservative, whereas Tēvāram could be said to be innovative, because although it contains a sizeable number of X-eṉal expressions, a majority of them are "new" if we compare them with the usage of ECT. A special case would be Cīvakacintāmaṇi, which can be said to be both conservative and innovative, on the basis of this criterion. One should also remark that it is of course not enough to simply give the list of X-eṉal expressions. The number of occurrences should also be given, in order to see which items were growing, so to speak, out of (linguistic) fashion.68 Another parameter is of course the inflectional freedom which was already alluded to at the end of the last section. This is only a preliminary survey. This being done, a new list of items can be compiled, extending our original list of 22 items which was based only on Ak, Pu, Na and Ku.

 

X-eṉal	Works
S24: nalleṉal	AI (doubtful)
S25: metteṉal	PP
S26: kommeṉal	PAR, CIV
S27: ammeṉal	KAL, CIV
S28: taiyeṉal	KAL
S29: puṟkeṉal	TK, CIV
S30: poḷḷeṉal	TK, CIV
S31: kaṭukkeṉal	NAL
S32: cevveṉal	CIL
S33: iḻiṉeṉal	MAN, CIV
S34: cikkeṉal	MAN, TEV
S35: pommeṉal	CIV
S36: kiṇiṉeṉal	CIV, TEV
S37: kiṇṇeṉal	TEV
S38: vaṟkeṉal	TEV
S39: tiṇṇeṉal	TEV
S40: veṭṭeṉal	TEV
S41: miṉṉeṉal	TEV
S42: mūceṉal	TEV
S43: nerukkeṉal	TEV
S44: mummeṉal	TEV
S45: ummeṉal	TEV
S46: mommeṉal	TEV

HLD: The Tivākaram and its aṉukaraña-v-ōcai-s

In the preceding sections, I have compiled several lists of X-eṉal expressions, and tried to set up criteria for them. I am of course not the first to do so. The first attempt in these matters for Tamil seems to have been that of the traditional Lexicographers who composed non-alphabetical lexicons, of the kośa-type, as a result of the influence of Sanskrit.69 The most ancient one to be preserved is the Tivākaram (ca. 7 th-8 th century) and one of its verses enumerates 8 items, namely

immeṉal (T1 = S7), kalleṉal (T2 = S1), iḻumeṉal (T3 = S8), valleṉal (T4 = S15),
pommeṉal (T5 = S35), olleṉal (T6 = S9), poḷḷeṉal (T7 = S30) & ñeḷḷeṉal (T8)

and characterises them as being aṉukaraṇa-v-ōcai "imitative sounds".

 

A first remark is that five of these items (T1, T2, T3, T4 & T6) are to be seen in Chart 3. Of the remaining three, two more (T5 & T7) are found in Chart 5,70 but for the last one (T8) no attestation could be found by me in any text.

Sanskrit roots of Tivākaram terminology.

Regarding the technical term aṉukaraṇa-v-ōcai which is found in the Tivākaram, it appears that its first part was borrowed from Sanskrit and apparently goes back to the pāṇinian grammatical tradition: Pāṇini himself, the ancient Sanskrit grammarian, when referring to some items in the Sanskrit language, uses once the term anukaraṇa "imitative" and twice the expression avyaktānukaraṇa "imitation of an inarticulate sound".71 According to some commentators and to some of the modern explanatory translators of the Aṣṭādhyāyī, the lexical items which are referred to by the following expressions, could possibly be those listed in Chart 2, below, some of them being based on the particle iti. It is to be noted, in that respect, that the peculiarities of the use of the quotative verb (or of the quotative particles) in the Dravidian languages have been, in the XXth century, compared by Emeneau with those of the particle iti in Sanskrit. If the items gathered in Chart 2 are indeed what Pāṇini had in mind when he used anukaraṇa and avyaktānukaraṇa, Emeneau, in his observations on the parallelism between Indo-Aryan and Dravidian, can thus be seen as walking in the footsteps of the ancient Tamil lexicographer who composed Tivākaram.72

 

Items with marker iti	Example quoted (or explanation given)	Pāṇini sūtra
khaṭad iti	khaṭad iti kṛtvā vs. khaṭatkṛtya	I.4.62 (Renou [1966])
ghaṭad iti		VI.1.98 (ibid.)
paṭ iti	paṭ iti karoti	V.4.57 (ibid.)
paṭad iti	(becomes paṭ iti)	VI.1.98 (ibid.)
paṭatpaṭad iti		VI.1.99 (ibid.)
paṭatpaṭeti	(from. paṭatpaṭat + iti)	VI.1.99 (ibid.)
śrad iti		VI.1.98 (ibid.)

 

Items with marker ḌāC	Example quoted (or explanation given)	Pāṇini sūtra
paṭapaṭā	paṭapaṭā karoti	V.4.57 (ibid.)
paṭapaṭā		VI.1.100 (with ref. to VIII.1.12, vt.)
kharaṭakharaṭā	kharaṭakharaṭā karoti	V.4.57 (ibid.)

R-Type (and suffixed S-Type) expressions in traditional lexicons

S. Vaiyapuri Pillai, the editor of TL, remarks in his introduction (TL, Vol. 1, p. xxvi-xxvii), commenting upon the Tivākaram, that:

The first edition of this work (1835 A.D.) by Tāṇḍavarāya Mudaliar covers the first ten sections only and contains many sutras, admittedly composed afresh and added by him. Excluding these additions, about 9500 words are dealt with. The later editions fail to distinguish between the original sutras and these additions. There are indications showing that a like process of addition had gone on even prior to the first edition. The original could have been only a work of moderate size, though sufficiently comprehensive for the time.

As a matter of fact, it so happens that the verse (or sūtra) containing the 8 items (T1 to T8) which are characterized as aṉukaraṇa-v-ōcai, seems to belong to the first stratum of the Tivākaram, but that it is followed in the 1835 edition by 2 extraneous sūtras (which are explicitely discarded in the 1990-1993 critical edition). The first of these sūtras contains 6 items, also called aṉukaraṇa-v-ōcai, 5 of which are remarkable for the suffix (-ēl or -ēr) which they contain:

kommeṉal (T9=S26), viṭēleṉal (T10), terēleṉal (T11), mukēreṉal (T12),74 ammeṉal (T13), carēleṉal (T14).

The second extraneous sūtra contains 10 items, characterized as being upayavōcaiyiṉ peyar:

paṭapaṭeṉal (T15), kaḷakaḷeṉal (T16), patapateṉal (T17), koḷakoḷeṉal (T18),
moṭumoṭeṉal (T19), tiṭutiṭeṉal (T20), mokamokeṉal (T21), kalakaleṉal (T22),
neṭaneṭeṉal (T23), kaṭakaṭeṉal (T24)

We meet here, for the second time, with Sanskrit terminology, because upaya is the tamilized form of ubhaya-. The upayavōcai are also called iraṭṭaik kiḷavi in the Tolkāppiyam —they are those items which we assigned to the R-Type.75

 

Whatever be the source from which items T9 to T24 have been taken from,76 we can observe that several of these items are already familiar to us, because we have (see chart 5):

(13) kommeṉal (T9) = S26 (see chart 5),
(14) ammeṉal (T13) = S27 (id.),
(15) kalakaleṉal (T22) = R2 (see chart 1)

Moreover, even though an item like T23 (neṭaneṭeṉal) is not found in the Tēvāram,77 we find there some close equivalent (if we are to believe TL when he gives two items as semantically equivalent), though in a very peculiar condition because it occurs without its eṉal component, in the following phrase:

(16) neṭuneṭu iṟṟu viḻā" [My mind was meditating on the feet whose toe was fixed so that the heads and shoulders] fell down crackling and splitting with a sound ´neṭuneṭú "78

Emancipation of X-eṉal expressions from the eṉal component

The phenomenon just mentionned is not an isolated occurrence. It happens several times with R-type expressions. Other examples are for instance:

(17) neṟuneṟu (Tēvāram 6.18.11),79
(18) calacala (Tēvāram 1.22.1, 1.22.3, 2.91.3)
(19) kaṟakaṟa (Tēvāram 7.54.5)

It has been noted that in modern Tamil, the emancipation of the R-type from the presence of an eṉal component is frequent.80 It is however not compulsory.

 

The emancipation from the eṉal component in Later Classical Tamil is not reserved to the R-Type. We can mention at least one case where it occurs with a S-Type element, characterized by a suffix. The item T14 (carēleṉal) which is listed in an extraneous sūtra of Tivākaram can be seen, without the eṉal component, here:

(20) manti pāyac carēlac corintum murintu ukka pū "flowers which shed bending, and pouring, when the female monkeys suddenly leap" (Tēvāram 2.114.3, translation VMS)

We should finally add that in the case of S-type elements, the disappearance of the eṉal component transforms them in some sense into simple adjectives.81 For instance, after the classical period, the item naḷḷeṉal (=S5) seems to disappear: what we find is a simplified form, in formulas like naḷḷiruḷ (14 occ. in Tēv.), instead of the former formulas: naḷḷeṉ kaṅkul (7 occ. in Ku. + Na. + Ak.), naḷḷeṉ yāmattu (10 occ. in Ku. + Na. + Ak.).

The rise of suffixes

Apart from the increasing frequency of R-Type X-eṉal expressions, and the concomitant emancipation from the eṉal component, another important feature in the history of X-eṉal expressions is the rising importance of suffixes. We have already given in Chart 2 a brief outline. We are now in a position, drawing from the texts explored so far, to mention a few suffixes. The first one to be mentioned could be the one just seen in T14 (carēleṉal), which possesses two variants (ēl and ēr) in LCT and which is found in 4 forms in Modern Tamil (-ēr/- ēl/-ār/-īr). It is to be noted that the suffix was already present in ECT, in an item like S16 (ñerēreṉal). The mention of 4 items belonging to this group in the extraneous Tivākaram sūtra is well in accordance with the rising visibility of this suffix. Other suffixes with rising importance are -iṉ, -um, -ku, as one can see when comparing charts 1, 2 b, 3 a, 4 a & 4 b with the following chart 782

 

	S (simple)	R (redup.)	E (echo)	Total
no suffix	120 items	231 items	41 items	392
- um suffix	9	2	1	12
-ēr/-ēl/-ār/-īr	44	6	1	61
-iṉ	5	4	1	10
- kku/-ku	45	39	10	94
-ā	3	6	12	21
Other suffixes	9	13	11	33
Total	235	301	77	613

The rise of the E-type

So far, the only E-type X-eṉal expression which I have pointed out is one found in Tēvāram (E1= tiṭukumoṭṭeṉal), and it is a doubtful one because it does not exactly fit into the pattern of those E-type expressions which I mentioned for contemporary Tamil. However, if we examine other works, we are able to find the following items, already used in ancient texts:83

(21) calārpilāreṉal (Tiv. periyāḻ. 1.7.1)
(22) cīṟupāṟeṉal (Tiv. tiruppā.)
(23) naṟukumuṟukeṉal (īṭu, 6, 9, 5)
(24) calukkumolukkeṉal (kaliṅ.)
(25) naṟumuṟuttal (Tiruvācakam, )84

These examples are of course better than the one found in the Tēvāram (E1 in Chart 1). They testify that this phonologically peculiar class of words is ancient.85

HGD: The grammarianś analysis: a semantic triad

Compared with the lexicographers´ task, which may simply have been a patient and cumulative labour once the category (or the label) had been given its place under the sun, the grammarians´ task was a more arduous one. Their interaction with the X-eṉal expressions seems in some sense to have been indirect, because we find in the T several statements concerning either the X or the eṉal component but the appearance of X-eṉal expressions as wholes is found only in examples given by the commentators, as Iḷampūraṇar and his successors. Therefore, we can never be sure that those items are really what the author of T had in mind. However, we are not in a position to simply dismiss T´s commentaries as irrelevant, and I shall proceed to list the X-eṉal expressions that are adduced by them, before examining the corresponding T sūtras.

Zoom in

Download the image (.jpg, 79.93 KB)

 

Chart 8 a: X-eṉal expressions inside T́s commentaries (Classical items)86

Zoom in

Download the image (.jpg)

 

Chart 8 b: X-eṉal expressions inside T́s commentaries (New items) 87-88

The triad and the rise of "quality"

There are of course many comments to be made on these data and on the way they are related to the T sūtras for which they are given as illustrations. However, before making any observation, it seems necessary first to explain the way these expressions are analysed by grammarians. The main fact seems to be that the most "tangible" (or ascertainable) element inside the X-eṉal expressions seemed to be for them the eṉal component, most often quoted in one of two inflected forms, eṉa and eṉṟu, considered however as "particles" (iṭaic col). As far as the X component is concerned, it is sometimes taken as a topic for discussion but as being part of an unanalysable whole.89 Thus, paradoxically, classifications of the different semantic value-types of the X component can be found in fact, however with a few exceptions, inside the list of possible values of the particles eṉa and eṉṟu. And it is inside the commentaries to sūtras dealing with eṉa and eṉṟu that we find many of the examples that are mentioned in chart 8. These sūtras enumerate six different values for these particles, and among these six, three are illustrated by examples containing X-eṉal expressions, these three being characterized by the words icai "sound", kuṟippu "notion, idea, intent" and paṇpu "quality".90 It has also to be noted that these three terms, which I shall designate globally as the semantic "Triad", are also found together, without additional terms in other contexts where they seem to be possible characterizations of the X component.

Conclusion

I had announced I would postpone the terminological discussion which touches on the essence of X-eṉal expressions, i.e. what makes them to be what they are, until after completing (at least partially) the census that has taken the better part of this essay. Any such discussion has to be based on facts. I have tried, until now, to remain as close as possible to the morphological ground, but one may try to find out whether these items have also something in common semantically, beyond their singularities. Before the advent of "ideophone",91 many other terms have been used although the different authors do not always agree on what they mean by them.92 In the western grammatical tradition, the technical term onomatopoeia has for a long time been in use to refer to the process of imitative word formation, although this was not its original meaning in Greek.93 Other vocabulary elements are "onomatopoeic words", "expressives", "iconicity", etc., to which we should add anukaraṇa ("imitative") which we found in the Sanskrit tradition and the triad (icai, kuṟippu, paṇpu) which our Tamil grammarians have used. The perspective under which some of these discussions have taken place is, for instance, summarized by M.L. Apte [1968: 5], who invokes E. Sapir [1929] but does not discuss Pāṇini,94 before embarking himself on an exploration of Marathi:

Although Sapir has no general discussion on the usefulness of the terms ´echo words´ or ´onomatopoeic words´ he has described in detail in his article ´A study in phonetic symbolism´ [95] the experiment he carried out to find how far the ´expressive [´] symbolism existed in any language apart from the ´referential´ symbolism which is the very essence of linguistic form. The results of his experiments according to him go far to support the theory that such an ´expressive´ symbolism does exist (Apte [1968: 5])

Sapir himself had said:

The symbolism of language is, or may be, twofold. By far the greater portion of its recognized content and structure is symbolic in a purely referential sense; in other words, the meaningful combinations of vowels and consonants derive their functional significance from the arbitrary associations between them and their meanings established by various societies in the course of an uncontrollably long period of historical development. […] This completely dissociated type of symbolism is of course familiar, it is the very essence of linguistic form. But there are other types of linguistic expression that suggest a more fundamental, a psychologically primary, sort of symbolism. […] We may call this type of symbolism ´expressive´ as contrasted with the merely ´referential´ symbolism which was first spoken of. (E. Sapir [1929]96

However, if we try to make a working definition starting from the intuition which is expressed here and try to apply it to the description of a specific language, we may run into several difficulties.

• If the language is a living language, the differing perceptions of native speakers and of non-native speakers and the echo that will arise among linguists97 because of these differing perceptions can be very puzzling

• If the language is a dead language, there will be puzzling traces of differing perceptions.98

This article will not be able to provide an answer to a general question which would be: "What is an ideophone?" I could only try to answer the more restricted one: "What has been in the course of time, the behaviour of what I perceive as a morphologically coherent group of items, with this additional éclairage that they seem to behave in ways which resemble those of ´similar´ groups of items in other languages?" It is quite possible that further studies will bring me closer to less ineffable explanations on what "expressive symbolism" is. For the time being, I have to go on reading texts and trying to recreate artificially (mine is ceyaṟkai) for non-native recipients what spontaneous perception I suppose exists for native recipients (theirs is iyaṟkai). Among the items which I regret not to have had the time to present in this essay are all the X-eṉal musical elements,99 which are found frequently, especially in the Tēvāram,100 and which might have an even more efficient claim for real iconicity. Therefore, as a compensation for their absence, I shall give the last word to one of them:

tantattintattaṭameṉṟa aruvit tiraḷ pāyntu pōy…

"the collection of streams flowing with the sound tantattintattaṭam… "101

Apprendix: alphabetical list of the X-eṉal expressions discussed102

Zoom in

Download the image (.jpg, 98.63 KB)