The Diglossic situation is well established in recent years since the publication of Ferguson 's (1959) article on diglossia.  The clevage between the H(igh) and L(ow) varieties of Tamil is found at all linguistic levels starting from phonology to Syntax.  Though the dialossic situation remains relatively stable all through these years, the gap or the distance between the two varieties seems to vary now and then in the history of the language. This alternating convergent and divergent situations existing between the two varieties of Tamil diglossic are due to linguistic as well as socio-political reasons.  Within the linguistic sphere, Tamil script, as it is used for representing both the varieties, plays a small but significant role for the unstable distance the two varieties.  This paper aims at studying the role of writing in the on-going dynamic situation existing between the two varieties of Tamil diglossia.

            The script of a language is devised to represent mainly the sounds of that particular language.  However, this representation is always found to be inadequate and there exists a gap between the script and the reading pronunciation.  There seems to be no faithful rendering of the script in its true sense.  A script of the language is mostly associated with its (classical) literary writings and its long standing tradition bound association with the classical literature of the language, creates an amount of prestige and further, the script becomes a source of historical information, besides being a marker of social and cultural identity of that particular language group. Sometimes, script indirectly helps to stabilize the clevage between the two varieties at linguistic and social levels by the availability and relative permanence of written records and its effect as `a model for those who aim at safeguarding and accomplishing the marks of excellence in usage' (De Silva, 1976:33).  The constant inflow of new words in H-variety to express new ideas and concepts and the standardization process undertaken in the H-variety of the language (Shanmugham, 1975) seems to narrow the gap between the two varieties of the diglossic language.  The amount of loan words being used assimilated or unassimilated the growing trend for a simple and easy writing are some of the indications for close relationship existing between the two varieties of the language.  However, it must be noted here that there is a hard-core H-variety form of Tamil used by pandits for classical writings, where all these modernization processes (such as loan words, borrowings, etc) are kept at low ebb and `pure-Tamil' seems to be the order of the day.  Non-literary writings such as writing novels, short stories, journalistic writings, etc. are the one in which we see a kind of convergent situation in progress.

            It is true that script represents the sounds of that particular language.  But when there is a diglossic situation in the language such as in Tamil 3, the existing script system seems to be adequate to some extent to represent the sounds of the H-variety of the language, but it remains a burden or constraint to those who wish to go closer to speech form, the L-variety of the language.  A script may have been originally a direct reflection of the speech through a process of high level abstraction (Thirumalai, 1978:36), but its long standing association with classical literary writings, the existing diglossic situation and the necessity for using the above prestigious script to represent spoken variety makes one frustrated when he cannot represent as such the nuances of the word he hears.  The constraint imposed by the script forces a writer to create variations and inconsistencies in his writings.  Usage of non-native words to express new concepts and ideas in writing system representing H- or L-variety remains unavoidable in any developing society.

            Most writers use generally a type of a standard spoken forms of Tamil (which belongs to L-variety) which `is not a representative speech of a region, caste or religion (Vasantha Kumari, 1976).  Annamalai (1976) mentions that the standard spoken variety is obtained' by eliminating all the stigmatized, stereotyped or marked features of the home dialects.  Thus, writers use neutral items brings forth the standard spoken Tamil near to written, the H-variety form of Tamil (Gnanasundaram, 1980).  As said earlier, a writer could not go closer to the speech form in his writings due to the inadequacy of the script for a faithful representation of it.  Further, the writer himself does not want to have such a faithful representation because any such representation will hamper one's reading speed.  Thus, technically, a writer needs a spoken effect in his work for which he manipulates the script system in order to get the feeling that he is representing the spoken variety.  Evidently, the above situation makes a writer to represent his L-variety with a sizable mixture and influence of the H-variety of the language.

            Some of the constraints laid down by traditional grammarians for the proper use of the Tamil script, are generally over-looked by many writers in representing H- as well as L-variety of Tamil diglossia.  Initial occurrences of liquid consonants such as r, l, etc. are disallowed in the writing system.  The above liquid consonants however can occur only preceded by the vowel `i' in the initial positions as laTTu written as illTTu `laddu'.  It is not clear how words such as lilli' linkam will be represented.  Further, plosive consonants such as k,t, p, c, etc. do not occur in final positions as pure consonants.  External sandhi rules across word boundaries such as occurrence of gemination of word final plosive consonant when followed by another word initial plosive consonant, are not strictly followed in the writing such as tiTTa kamisan `planning commission' pookku varattu kalakam `Transport Corporation', peeccai keeTTaakal, (They) listened to the speech etc. indicate the absence of gemination by the external sandhi rule which is prescribed for the writing system.  The above three constraints are not found in the L-variety of the language.  Any H-variety writing disallowing these three constraints is obviously close to that of the L-variety.

            Initial consonant clusters are not allowed in the writing system without the occurrence of enunciative vowel in between the cluster.  Writing system follows this rule in representing both these two varieties.  Words such as kiraamam' pirapala "popular", piratamar "Prime Minister" are written without initial clusters in both varieties.  Words such as piraa `brassiers' are written sometimes with initial cluster as praa, Nasalization of penultimate vowel followed by a nasal in word is not represented in the writing system.  However, this representation is not possible by the available Tamil script.  So the writing system represents such words as non-nasalized vowel followed by a nasal as `avan' `he' paNam `money', in H- as well as in L-  variety of Tamil⁴.  Initial voiced consonants are represented mostly by their voiceless counterparts except in the case of C where the Sanskrit grapheme `G', is used for voiced ones as in `jatakam' `astrology' vs saatakam `practice'. Some writers use bold letters for voiced consonants and some use colophon in front of the respective grapheme to denote the voicing of the sound⁵ (Thirumalai, 1978:39).

            The inadequacy of representing initial consonant clusters and voicing of some initial consonants are compensated by the reading pronunciation by reading in those palaces as consonant clusters and voiced consonants.  Thus, generally the non-standardized script system tends to move towards L-variety by the reading pronunciation also.  Annamalai (1976, b) has shown that some of the traditional grammatical rules have changed to adjust the usage of modern times and the above changes are found in the direction of the L-variety of the language.

            Writers representing L-variety do not seem to have any fixed norm among themselves.  Due to the non-standardization of the script for representing L-variety forms, each writer differs from others and thus feel their own individual identity.  It is also true that even an individual author is not consistent in is own writings in the same piece representing L-variety.  Consider for example, Jayakanthan's short story `rishimuulam' where alternating forms are used in representing one item as own shown below: `paathaa/paarthaa' (she) saw (it)'; kaTTille' kaTTilile `in the cot', keeTTaalnu' keeTTaalnnu `that (she) asked', keeTTu connaal' keeTTucconnaal `(she) listened (to it) and said'.  We further notice lexical items which are more often found in H-variety, being used in L-variety writings.  Some times H-variety lexical items unaffected by `conversion rules'6 are used in L-variety as: paTTri `about', aval `she' etc. instead of fors such as patti and ava, respectively, which are converted into L-variety ones as shown below: VantukoNtiruntaa instead of vantukiTTiruntaa `she was coming'.  Sometimes, particular type of syntactic construction found more often in H-variety is also found to occur in L- variety writings.  For example, usage of desire verbs such as ari `to know', Virumbu `to like' are found more in H-variety where as they are expressed in L-variety more by verbs like teri and piTi respectively.  However, the usage of verbs of H-variety are found sometimes in L-variety writings.

1. a)  naan avarukiTTeeyiiruntu arinceen

          I from him (I) learnt

          I learnt it from him  

          b)    naan inkeyirukkattaan eppavum virumpreen

          I to be here only always (I) like

          I always like to be here only  

2. a)  naam avarukiTTeeyiruntu terincikiTTeen

           I from him (I) learned

           I learnt it from him

          b)    enakku inkeyirukkattan eppavum piTikkum

          to me to be here only always like

          I always like to be here only

Thus, we see, the non-standardization of the script system used for representing L-variety, the writers' compromise between one's reading speed and the representation of L-variety allow for a mixture of H-variety forms in one's L-variety writings.  Similarly, innovations and changes that are taking place in the H-variety, the motivation for easy understanding of the reader forces the H-variety writings to be closed to the L-variety of the language.  The above situation is bound to lead towards a kind of convergence between the two varieties of Tamil language.

The bi-partie situation of Tamil diglossia is further brought closer more in the Radio medium than in the print medium.  Any Radio literature is intended to the aural medium and needs to be read whether it is in H-variety or in L-variety, for the message to get across to the listener.  The difference in the manner in which one's writing is carried over these two media demands for a different objective and technique in the art of writing.  The above fact that the script is going to be `spoken' influences the author writing in H-variety to make his script suitable for reading; i.e. to write in an easy and simple form of H-variety; simple in the sense the sentences are not too long and complex with embedded ones<sup>7</sup>; easy in the sense, the lexical items are not pedantic and archaic, but are known to wider population and easily understandable.  Usage of loan words are found to be more in Radio literature than in print media.  Radio literature aims at a wider section of people educated as well as uneducated, where as the print media can only be utilized by educated mass.  Because of this significant difference with wide listening, radio literature needs to be more simple and easy for everyone who hears it, to follow it and understand the material.  Thus, the radio literature of H-variety seems to be near to standard spoken variety with regard to simple syntax and neutral lexical items<sup>8</sup>.  The essential factor of reading the radio materials for broadcast adds further to the `nearness' between the two varieties by its reading pronunciation.  The occurrence of liquids, voiced consonants and consonant-clusters in word – initial position, occurrence of plosive consonant in word – final positions and the absence of external sandhi rules which are all found in the L-variety, are also found in the reading pronunciation of the H-variety radio literature.  However, non-nasalization of penultimate vowels following nasals, marks the difference in the reading pronunciation of H-variety from that of L-variety besides individual linguistic items belonging to each variety.

It is assumed that most of the L-variety forms are derived from H-variety by a set of conversion rules (Schiffman, 1978, Arokianathan (ms.), Deivasundaram, 1978).  However, in some cases we have independent forms belonging to each variety which are not derivable.  Some L-variety forms such as ava `she', uuttu `to pour' are derived from the corresponding H-variety ones aval, uurru respectively, by a set of phonological rules.  We have morphemic sets such as koNTu/kiTTu `marker for continuous aspect' where the former is used mostly for H-variety and the latter for L-variety.  Sometimes, conversion rules may apply to its maximum or may stop at some arbitrary intermediate stages as needed, as shown below: H-variety form piTittaan   piTittaa

PiTiccaa           piTiccaa            puTiccaa

`He caught (it) !' The more the number of rules apply, the more colloquial or sub-standard the form appears.  Since, the word needs to have a spoken effect to be understood by all and read over the prestigious radio medium, radio literature of L-variety often aims to remain with less application of the conversion rules, thus, being closer to H-variety.  Further the direction of conversion in a word is found always from right to left i.e., a H-variety form gets converted to L-variety starting from the last syllable of word.  A conversion rule, in other words, cannot abruptly affect in the middle of a word. The non-occurrence of the forms such as *puTittaan, *piTicaan (H-variety form piTittaan `he caught (it)') illustrate the direction of change as right to left in a word converting from H-variety to L-variety.  Thanks to the necessity of reading involved in the radio medium a writer of the Radio literature of L-variety can freely use H-variety forms (which are convertible to L-variety) and shrug the responsibility of converting to L-variety to the one who puts them across the medium for the listener.  This makes the radio reader (announcer, talker or actor) to do the conversion at the time of presenting the material over the medium and the quantum of conversion depends on the radio reader's ability

and experience in the art of conversion.  This situation can well be understood by comparing the L-variety material meant for broadcast and recording it.  Generally, sentences in L-variety radio literature have verbal endings in L-variety, Words of scientific and technical terms are in H-variety in L-variety radio literature e.g. visai telippaan `power sprayer', porukku vitai `improved seeds'.  Loan words are used in L-variety as well as in H-variety with no hesitation in the radio medium.  Passive construction which is found only in H-variety of Tamil is also found in L-variety radio literature.  Thus, in L-variety radio literature, grammatical as well as form words, phrases and even whole sentence appear in H-variety.  On contrary, L-variety forms are used in H-variety sometimes `assimilating' them phonologically to suit the H-variety as esaku pisaku a L-variety form being used in H-variety as isaku pisaku `petty mistakes' where usage of the above item is non-extent in H-variety.  Forms such assnaalu `four' whose corresponding H-variety form is aanku `four' are more often found to occur in H-variety  radio literature.  Whenever there is a confrontation for selection of a non-stigmatized word, radio literature uses invariably either H-variety forms or loan words which even is easy to understand.

            Some of the sentences given below are taken from a L-variety play broadcast in an agricultural programme from All India Radio, Trichy on 25^th March 1976 .  The sentences given below will indicate the amount of H-variety forms (which are underlined here) being used in a L-variety radio programme.

3.  Inraikku nam munne vantirukkira valakku

    today us before have come case

    patti cila vaartiakal

    about some words

Here are some words about the case which has come to us today.

4.  porukku vitaikala malivaana vilaikku koTukkutu

    selected seeds cheap price it gives

    It gives improved variety of seeds for a cheap price.

5.  tinam pala ariya vivasaaya visayankalaiyum

    daily many important agricultural news also

    oliparappi varraanka  

    broadcast (they) are coming               

    Daily, they keep broadcasting also many important agricultural news.

6.  aanaa, atu illaammee nam coompeerittanamum

    but that without our laziness

    muyarci inmaiyaalum tean pattaakuraikku

    non-incentive also for in-sufficiency

    kaarnamaayirukku

    it is (the) reason.

But, besides this, our laziness and non-incentive attitude become the reasons for our insufficiency (in food products).

Thus, we see a great amount of H-variety forms in L-variety radio literature and vice versa.  Thus, the influence or the overlapping situation between the two varieties is found to occur regularly and systematically, thus, leading to a possible convergence between the two varieties of Tamil diglossia.  The reason for the above situation partly lies within the Tamil script system which is non-standardized and inadequate for writing.  The whole problem further culminates in the radio medium by the reading pronunciation when one's writing needs to be transmitted only by reading through the aural medium for its prospective listeners.

NOTES :

1. Shanmugham (1960) refers to the expansion of domains of L-variety and predicts a possible merger of the two varieties in Tamil diglossia.  Zvelebil (1964) also refers to a possible merger of the two varieties in Tamil.  Annamali (forthcoming) observes that changes in H-variety are in the director of L-variety and thus predicts a possible merger.  Though everyone predicts a merger, yet they have difference of opinion regarding the direction in which one variety moves towards the other.  Schiffman (1978) believes that there exists a built-in mechanism in Tamil which averts any possible merger.
2. Scholars like Manicka Naikar (1971) have studied the mysticism that surrounds around the origin and the shape of Tamils script.  There are scholars who claim that the shape of the Tamil script represents evolution of human life, nature and Hindu philosophy.
3. Tamil script system is alphabetic in nature and represents phonemically the sounds of Tamil language (Thirumalai, 1978: 41).
4. Some writers omit the nasal phonemes in words like naan which are written in spoken variety as naa:he'.  If generalized, this will lead to further confusion in the case of avan `he' when written as ava, which is homphonous to the spoken variety of aval `she'.
5. Sometimes, graphemes English language are used in Tamil representing spoken variety.  Tuglak magazine uses `f' for bilabilal fricative sound as in `form', `file'.
6. For conversion rules see Schiffman (1978), Arokianathan (ms.).
7. c.f. Thirumalai (1978).
8. Neutral items are those which occur frequently in both varieties (c.f. Gnanasundaram, 1980).
9. c.f. Schiffman and Arokianathan (1980).
10. Gnanasundaram (1980) refers to such a situation in conversational environoment where stigmatized spoken word is substituted by a standard spoken form modeled on H-variety of Tamil.  

REFERENCES :

Annamalai, E. 1976

`Standard Spoken Tamil': some observations' in R.V.K. Thampuran, (eds), Studies in Dialectology vol: 1:2. Trivandrum : Linguistic Department mimeo.

Annamalai, E. (1976 b)

"Grammar and Error" Tamizhaayvu. 4.

Madras : University of Madras .

(forthcoming) "Convergence of Literary and Colloquial Tamil" Pakha

Samajam, University of Punjab . 

Arokianathan. S. "Vowel shift in Diglossic Tamil". (ms)

Devisundaram, N. 1978 "Can we measure . . . ? "Pulami" 4:2. pp. 173-182.

DeSilva, M.W.S. 1976. Diglossia & Literacy, Mysore : Central Institute of

Indian Languages.

Ferguson , C. 1959.

"Diglossia" word 15:2 pp: 325-40.

Gnanasundaram, V. 1980. "The role of Diglossia in the Emergence of

Standard Spoken Tamil" paper presented in 2^nd International Conference in South Asian Languages & Linguistics, Hyderabad .

Manikka Naickar, P.V. 1971. The Tamil Alphabet: Its Mystic Aspect,

Madras : Siva Siddanatha publication.

Schiffman, H. 1978. A grammar of spoken Tamil Madras: Christian

Literature Society.

Schiffman, H. 1978.

 "Diglossia and Purity/Pollution", in Maloney (eds.) Contributions to Asian     Studies, pp. 98-110.

Schiffman, H. 1978 & Arokianathan, S. 1980.

"Diglossic variation in Tamil film and fiction" paper presented in 10^th International Congress of Anthropological and Ethnological Sciences, Mysore : C.I.I.L.

Shanmugham, S.V. 1975. "Modernization in Tamil".  Anthropological Linguistics, 17:2, pp. 53-67.

Shanmugham pillai, M. 1980.

"Tamil Literary and Colloquial" in Charles A. Ferguson (eds). Linguistic Diversity in South Asia , IJAL:7.

Tirumalai, M.S. 1978.

Stylistics: A Linguistic Approach. Madras : University of Madras .

Vasantha Kumari, T. 1976 "The standard Spoken Tamil" in Gopinathan

 Nair (eds.) Studies in Dalectology vol.1:1, Trivandrum . University of   

 Kerala.

Zvelebil, K. 1964.

"Spoken Language of Tamil Nadu"

Archive Orientalini, 32:2, pp. 237.64.