In the foregoing we noticed that certain Vedic syllables had a special association with certain regions and that these were absorbed in the languages spoken there. We also learned from this that the Vedas flourished in all countries. There was never a period in Tamil Nadu, the land we know intimately when Vedic dharma was not practised there.
The name "Tamizh" itself has the "zha" characteristic of the Talavakara Sakha of the Samaveda. Am I right in making such a claim? Or is it all the other way around? Suppose the argument goes like this: it is the "zha" characteristic of Tamil and the "ja" characteristic of Northern tongues that are seen as the distinguishing phonemes in the Vedic texts prevalent in Tamil Nadu and the North respectively. In other words what was already present in the regional languages came to be absorbed in the Vedic sakhas prevalent in the areas concerned. Did I put the whole thing topsy-turvy when I made the statement that the Vedic "zha", "ja" and "ba" became characteristic for the Tamils, Northerners and the Bengalis respectively, that these were reflected in the speech of each of these linguistic groups?
That the rules of the Siksa sastra had their impact on the regional languages is the correct view. The rules of the Pratisakhya do not apply to one area alone but to all those parts where the Vedic recension concerned is followed. If there is a Brahmin chanting the Talavakara in Kamarupa(Assam) or Kasmir, he will use "zha" where others use "da" or "la" in the mantras. A Brahmin who chants hymns from the Krsna-Yajurveda has to use "da" instead of "zha" or "la" whether he belongs to Gujarat or Maharastra or any other place in India. In the same way, it is not only the Kannadiga, any Rgvedin anywhere will use "la" where others use "da" or "zha" in chanting the mantras. The Pratisakhya determines the sound of Vedic mantras not for a particular area alone but for the whole country. In course of time the local language takes on the characteristics of the sakha where it is practised.
The name of the month "Margasirsi" is derived from the fact that generally the full moon falls on the day to which is conjoined the asterism of Mrgasirsa during that month. Margasirsi is Margazhi in Tamil. "Si" changed to "di" and "di": to "zhi". It is according to the genius of that language that "sa" becomes "da". "Purusa" is called "purudan" in Tamil and "Nahusa" is "Nag(h)udan" in Tamil poetry. Kambar calls Vibhisana "Vidanan". But, if Margasirsi changed to "Margasirdi" and then the "sir" in the middle dropped, should not the word have the final form of "Margadi"? How do you explain the presence of the "zha-kara"? In other words, how does the name of the month finally take the name "Margazhi"? The "zha-kara" must be attributed to the Talavakara Sakha that was predominant in Tamil Nadu.
People belonging to this recension use "zha" and Krsna-Yajurvedins use "da", don't they? This habit they still retain unconsciously. The Telugu Vaisnavas sing the Tamil Divyaprabandham during worship in the temples. In Tirupati the Tamil Tiruppavai is sung before the Lord. It starts with the words "Margazhi-t-tingal". "Zhi" is difficult for Telugus to vocalise. How is it that they do not say "Margali" or "Margali" then? They say "Margadi-t-tingal", that is with the "da-kara" instead of the "zha-kara". When they chant hymns from the Samaveda that is prevalent in Tamil Nadu they unconsciously use the " da-kara" for the "zha-kara". "Da is in the blood of the Yajurvedins, so they say "Margadi" instead of "Margazhi".
For a general background, please see here