MOLI : THE PERSONAIC NATURE OF LANGUAGE

Nirmal Selvamony

For it is language which binds things together (Antoine de St. Exupery)

The Tamil word moli means not only language, but also joint or articulation . This paper proposes to probe into language in the light of the latter meaning.

Language, as joint, connects the speaker or language producer, the world and the sacred. Now, what does ‘connection’ mean here? In order to understand the connection effected by language, we need to explain what kind of connector the joint is. For a joint to be able to connect two parts of the body, two things are essential -- firstly, the parts so connected should have something common to them, and secondly, the joint itself should share the nature of the parts in some way or the other. For example, bone joints. In this respect, a joint is unlike some types of fastener that neither partake of the nature of the parts connected, nor others which connect things totally different in nature.

Language, as moli, like bone joints and plant nodes, connects only similar entities and also partakes of the nature of the entities so connected. In other words, there is something common between language and the similar entities connected by language. Since language proceeds from a person situated in a context, we need to see what elements constitute the context. The two basic elements of the context are the world and the sacred. It should be possible to find language bear the nature of personhood and of the other elements of the context. It is for this reason that language cannot be regarded a symbol system, for, the symbol does not partake of the nature of the person. If the symbol does not adequately represent the nature of language, does the metaphor do so? This is implausible, for, language as a whole is regarded as a symbolic system and not as a metaphorical system. Only parts of language, like the noun (as in ‘he is a lion’), the verb (as in ‘swallow your pride’), the adjective (as in ‘she is quite bright’) and the idiomatic phrases (like ‘letting the cat out of the bag’) are metaphorical. This paper purports to argue that the nature of personhood so shared commonly by all the elements of language is persona. This implies that non-personaic entities also assume the nature of personhood when jointed by language. This nature is both physical and metaphysical. If the acoustic shape of language constitutes its physicality, the meaning this shape communicates is its metaphysical dimension. Meaning is never wholly contained in the verbal body emanating from the speaker. Rather, it is a point where the nature of the person, world, and the sacred converge. For this reason, meaning tends to become connotative and open-ended. For example, the basic Tamil verbal roots i, a, and u assume the nature of the three entities (person, world, and the sacred) onto which they converge. If i, being a front vowel, is produced by the front part of the articulatory apparatus, a, a back vowel, by that part of that apparatus that lies at the back farther from the front part, and u, another back vowel, by the protruding lips which are also frontal articulators. It is not hard to see how front articulation can suggest nearness, back articulation farness, and lip protrusion, the beyond. Obviously the world participates in these roots in terms of one of its basic features, namely, space. The spatial dimensions of nearness, farness, and the beyond/height which inhere in these roots secure these roots with the world. Space and time are metaphysical ontically and in that respect, they constitute two vital dimensions of the sacred . Being space signifiers, these roots also assume the nature of the sacred. Consequently, by virtue of their being the fundamental roots of a language, they invest the entire language system with these traits they have absorbed from the person, the world, and the sacred.

Etymologically, the term mol\i is closer to the Latinic ‘articulation’ than the Gallo-Romanic `language'. The Tamils called language mol\i because they conceived of it as a jointed system much like a skeleton made up of both articulate and inarticulate bones.

As an articulated system, mol\i evinces three significant characteristics: segmentability, activatability and interfacialness. The first two may be dealt with sketchily, but the third deserves in-depth treatment.

Segmentability

Segmentation at the morphological (verbal) level called `agglutination' is a well-known phenomenon and so does not require any discussion here.

Segmentation can be seen at two levels -- internally and externally. If the former refers to identifying distinct parts within, the latter refers to segmentation in the domain outside of the language system as such.

Internal Segmentation

Within the language system, segmentation occurs at four levels -- phonological, verbal, syntactic and semantic.

At the phonological level, a language like Tamil identifies two distinct members known as uyir (vowel) and mey (consonant). Tamil grammar texts like tolkappiyam have studied these phonological/verbal segments in great detail and have also postulated rules for their combination. For example, if combinations such as

V + C [al : a (V) + l (C) ]

C + V [ka: k (C) + a (V)]

C + V + C [kal: k ©(C) +a (V) +l (C)]

are possible, the following, C+C+V [kla: k (C) +l (C) +a (V)] is not, for, in Tamil two initial consonants can never be articulated with a vowel.

Syntactically, one could see either a bipartite segmentation of subject and predicate which is the traditional articulating mode or a tripartite S O V pattern (in Tamil) as against S V O in other languages like English.

The language units of a syntax can be articulated differently for special purposes. For example, in poetry one comes across such types as niralnirai, cunnam, atimari, and molimarru . Later grammars like nannul speak of others such as virputtu, tappicai, alaimaripappu, and kontukuttu .

Again, the line is also segmented without paying attention to meaning. Here is an example:

karunkot terumai kayirupa rintacaiyi

Similar dislocative articulation is found in English prosody too:

His house / is in / the vil / lage though

He will / not see/ me stop / ping here .

At the semantic level also segmentation is noticeable. Semioticians, after the manner of Saussure, look upon the word as a sign made up of signifier (sound image) and the signified (concept) . If the connection between these two segments was arbitrary to Saussure, it was necessary to Benveniste . Now, even if the latter were true, the language user cannot be discouraged from dislocating a signifier from a denotative signified and articulating it with another signified, which may be termed connotative. For example, the signifier of ‘tree’ could be dislodged from its denotative articulation, and tied up with the connotative signified, namely, the cross or a tree diagram. Since the connotative possibilities are indeterminate, semantic articulation, namely, linking up a signifier with a signified, is also indeterminate.

Significantly, the word porul itself connotes ‘articulation’: oru to fit, to match; p+oru -- poru to fit + l – porul that which is fitted, articulated meaning which results from the sound shape (col, signifier) fitting with an idea (porul, signified).

External Segmentation

Language is possible only when such entities as persons (who stand in relation to each other), and the non-human world (which consists of animate nature, inanimate nature, and the sacred) engage a paticipatory relationship. This relationship, like the one between the units within language, is also segmented by the personaizing power of language. Persons are able to relate to the non-human world by virtue of their ability to assume the proper persona. The non-human world neither remains a set of objects nor signs nor symbols nor metaphors, but participates in the personaic community of the language field in the capacity of grammatical person (tan\mai or mun\n\ilai or patarkkai). The entities that stand in such relationship may be spoken of as segments in the sense that these entities are both discrete and connected to each other and also participate in the nature of the whole that is so formed.

Segmentability is inevitable in the case of language since no person can represent himself/herself in terms of oneself. Whatever one ‘says’ about oneself is always in terms of another, be it another person or non-human entity. As stated earlier, the first person is first person only in reference to a second and third, and these in a total language field. In other words, language necessitates what we may call ‘otherization’. This otherization could be understood as a kind of masking or impersonating. The mask of x cannot be x itself; it can only be that of non-x. Language, as mask, both otherizes a person and makes the person participate in the nature of the other.

Activatability

Closely related to the segmented nature of language is its activatability. Even as the jointed segments of a body could be activated in certain ways, the joints in a language system allow the segments to be activated in certain ways. Activation here means `use' which could be either emphasising or deemphasising of one or more segments in an utterance.

The rules regarding segment use are elaborately dealt with by grammarians. In such uses as derivation, conjugation, declension and other such, not all the segments of a particular segment set are used, and the ones used could be emphasising or deemphasising the respective segments. For example, the last vowels in kenmiya and nakiyatu are shortened with the result that they help under-emphasize the linking vowel. This places the stress on the stem vowel and the terminal vowel. Similarly, the final `u' in words such as `pa$kku', `na$kku' and so on is underemphasised in order to activate the initial vowel of the base.

Again, as a rule, one could say that the consonant is not activated fully in words and sentences. But for poetic or musical purposes, it could be as in the following example:

kan; naen manam mika /

alaipayu te ; ; ; //

Here the consonant n is lengthened for filling the time measure.

In a sentence, a part of it can be inactivated by dropping it as in

(ni) va.

To activate a segment different devices have been adopted of which some common ones are – stress, pitch variation, positioning strategically, lengthening the vowel besides others.

External activatability

Though the external segments of language stand in an a priori relationship, at any given moment of language use these segments are engaged only in a certain unique manner and this engagement may be characterized as activation. For example, if the non-human world appears as the third person in a given context, it could as the first in another. Again, even in the former, no two appearances are alike.

Interfacialness

The next feature of the articulated system of language is interfacialness. Though one could speak of the interfacial features in the internal organisation of language itself, like the infix, utampatu mey (uru + akum – uruvakum), the euphonic particle (puli+am\ + pal\@am – puliyampalam) and other such, this paper will examine the phenomenon of interface in any language situation.

In order to speak of the interface one should first of all identify the entities that create the interface. It is these entities that are required to create mol\i, and the sum total of all these which constitute the language field.

Language Field

Basically, there are only two members in any language field, the speaker and the listener. Let us consider these in some detail using the traditional Tamil terms for these, namely, tanmai (first person), munnilai (second person), patarkkai (third person).

The equivalent for the speaker is tanmai, for, all utterances are ultimately attributable to it. Regardless of the ‘person’ indicated by the utterance, tan\mai is the source.

varukiren (come I)

-- nan varukiren\ (I come)

va (come) [imperative]

-- ni va (you come)

vantan (came he)

-- avan vantan (he came)

vantal (came she)

-- aval vantal (she came)

vantatu (came it)

-- atu vantatu (it came)

If tan\mai is speaker, mun\n\ilai is not the only listener, there is patarkkai too. If the former is proximate listener, the latter is remote listener.

There is also another significant difference between munnilai and patarkkai. If the former consists only of human listeners, the latter includes both animate and inanimate listeners. The neuter singular aspect of patarkkai includes not only inanimate and animate non-human beings, but also the sacred.

Though tanmai is the ultimate source of all utterances, it is not to be identified with an autonomous speaker-person, but with a language persona that speaker-person occupies. The ontological territory of tanmai is penetrated by its complement, namely, munnilai with the result that language itself is redeemed from the verge of solipsism. In all language events including the first person utterances like nencotu kilattal or soliloquies or monologues, the listener’s presence and influence cannot be ruled out. In fact, tan\mai speaks another’s word, not its own.

munnilai

(mun, before + nilai standing, posture; munnilai, that which stands before)

The term munnilai implies that there is always something before it, even as the word murram (front yard) implies a house before which the yard lies. The meaning of the term itself shows how tanmai and munnilai are inexorably related to each other.

Even as tanmai is the speaker in all language events regardless of the grammatical person ascribable to them, munnilai is the axiomatic listener in all language events. Sometimes the listener is real; at other times, imagined. Not all language fields have real listeners. For example, the one in a newspaper report is a socially constructed hypothetical listener who may vary from newspaper to newspaper. If one targets a discriminating one, another does not mind a vulgar glutton craving sensational diet. However, unless the speaker formulates the utterances in relation to a definite listener, they are not likely to be effective.

A listener either real or imagined, may either fully or partially participate in the language event or withdraw from it. If the participation is complete in such events as dialogue, and letter (which expects a reply), it is partial in others like the conversation in which the listener merely nods or punctuates with an interjection or exclamation, and in monologues where (s)he withdraws and remains behind the scenes.

At times, the listener is interiorized as in monologues like nencotu ucatal . Nevertheless, the interiorized kind derives from the actual exterior listener who is typical and primordial. This is so because language is a communitarian praxis learnt in contact with real, other people.

patarkkai

(v. patar, n. to spread; that which spreads; thought/disease/suffering; the servants or attendants or messengers who maintain a distant relationship).

From its etymological meaning we can infer that patarkkai refers to that member in a language field, like the attendants and servants, who are not intimately but remotely related to the speaker. However, since patar implies a source from which it spreads out, it is also a relational persona like tanmai and munnilai.

As for tanmai, it can never be listener, even as munnilai and patarkkai can never be speakers.

Unlike munnilai, patarkkai includes not only human listeners, but also non-human animate and inanimate ones because of its neuter reference. Among these one could find the sacred also, for, in Tamil tradition the sacred is considered, in many occasions, a neuter (a-gendered) phenomenon1.

Modern Western linguistics keeps the sacred out of its domain; in the language field imagined by the Tamils, the sacred has played a significant part. This does not mean that tolkappiyam has grammatical rules about the play of the sacred. But one cannot rule out the participation of the sacred in the sphere of language. Take for instance, the notions of uyir and mey. Grammarians may like to see uyir as only vowel, but how could they wish away its reference to the soul? How could they assert that eluttatikaram is only about vowels and consonants, and that it is devoid of philosophical insights into body-soul relationship? Similarly, how could `porul ’ be emptied of its reference to reality, bliss or even, God? Other linguistic phenomena like mantiram, ar\am, col are not without a touch of the sacred. These and other such may be seen as the nodal points of the sacred which inform every member of the language field.

Yet another significance of patarkkai is that it is the domain of valakku and marapu. Though patarkkai includes all sorts of people (noble and ignoble), valakku, which pertains only to the noble kind, also derives from this language sphere.

According to tol., the sacred can charge moli with power (arral, collarral) when the latter is spoken by those (niraimoli mantar, literally, persons of complete language) who see language in its relation to all members of the language field. The other two conditions for sacralising language are speaking under oath (anai) and speaking words that have hidden meaning.

The relationship between speaker and listener is complementary and constant. No language event is possible outside of this relationship. However, the personaic nature of the speaker – whether a mother or daughter; teacher or student etc. – is variable, and open-ended. It is not a priori like the former speaker-listener complementarity. Like the former, the latter, namely, the speaking persona and listening persona also constitute a dyadic pair. At the third remove is the person or persons behind these personae. Though it is possible to have one and the same person become the two dyadic pairs mentioned above, usually a minimum of two separate persons are involved.

To recapitulate, one or more persons assume specific social personae (which are usually social roles like motherhood or guruhood etc.) and the linguistic personae known as speaker (tanmai) and listener (munnilai or patarkkai).

So far we have tried to understand how the members, namely, the speaker and the various types of ‘listeners’ including the world and the sacred constitute an interactive and interrelational language field. The most consequential implication of this interrelationship is that none of these members can be isolated from this primordial context and privileged for constructing theories of language. This means that language cannot be explained in terms of any one of these members, like the speaker or the world or even the sacred. This has to be so, for, language does not reside in any one of these members but among them in the form of a jointing/articulating persona that segments, activates and interfaces in several complex ways.