Letters and languages

Kakatiya inscriptions use three languages: Sanskrit, Telugu and Kannada. Many records mix them. J. F. Fleet found that the Thousand Pillar inscription opens with eight lines of old Telugu and continues in Sanskrit verse.

The early records of the family often use Kannada. The Guduru pillar of 1124 carries Kannada text on three sides and Telugu verses on the fourth side.

A. C. Burnell, writing in 1878, traced the Telugu and Kannada alphabets back to the southern form of the Asoka script. His table of derivation places that script at 250 BCE. G. Bühler called the script of the Asoka edicts the older Maurya alphabet. He thought that it was probably in use at least in the second half of the fourth century BCE and in the third century BCE.

Ojha's Plate 4 gives the letters of ten inscriptions from the Bhattiprolu stupa in the Krishna districts, and dates them to the second century BCE. Bühler named their script the Dravidi of Bhattiprolu and dated it to about 200 BCE. He found it connected with the southern variety of the Maurya script, though it includes many very archaic signs.

Bühler wrote that the script of the Kannada and Telugu districts appears first in the Kadamba inscriptions of the fifth and sixth centuries. In the east he placed its archaic stage in the Salankayana plates and in the plates of the first two Chalukya kings of Vengi. Burnell reproduced a grant of the Salankayana king Vijayanandivarman and called its letters the Vengi alphabet. Burnell referred these records to the fourth and fifth centuries, while Bühler called their date uncertain. Both scholars found these letters still close to the older cave characters.

Bühler dated the archaic Chalukya inscriptions between 578 and about 660. In this period, he wrote, the letters of the western and the eastern records do not differ much. D. C. Sircar found that the Amudalapadu plates of 660 use the old Telugu-Kannada alphabet. In his view their letters closely resemble those of the published records of the Chalukya king Vikramaditya I, who issued the grant.

Bühler placed a middle stage of the script between about 650 and about 950, in the west and on the copper plates of the Chalukyas of Vengi. His last stage is the alphabet that Burnell called transitional and Fleet called Old-Kanarese. Bühler wrote that it differs little from the modern Kannada and Telugu scripts. In the east it first appears in the Vengi inscriptions of the eleventh century.

Telugu and Kannada once shared one script, which scholars call the Telugu-Kannada script. Epigraphia Telanganica names a record of 660 CE at Amudalapadu as the earliest Telugu inscription in Telangana. That record uses old Telugu-Kannada letters.

Parabrahma Sastry writes that the common Telugu-Kannada script had divided into two separate scripts by about 1150. He adds that Kannada inscriptions are rare in the Telugu country after that time.

Burnell, writing in 1878, placed the split later. He could find no distinction between the Kannada hand and the Old Telugu hand at the end of the thirteenth century. In his view the two alphabets drew apart only after 1300.

H. Krishna Sastri described the Padmakshi pillar of 1117 as written in bold and clear Kanarese characters of the twelfth century. Kanarese is an older English name for Kannada.

Editors describe the letters of each record. Fleet called the letters of the Thousand Pillar inscription of 1163 old-Canarese, meaning old Kannada. L. D. Barnett called the letters of the Ramappa pillar of 1213 fairly good Telugu characters of the period.

E. Hultzsch called the alphabet of the Motupalli pillar of 1244-45 Telugu. He noted that some letters, such as the dentals, the linguals, r and v, are hard to tell apart without the context.

Parabrahma Sastry states that students of epigraphy can easily tell thirteenth-century Telugu letters from the Kannada letters of the same period. He reads the letters on Kakatiya gold coins as Telugu, not Kannada.

Some copper plates use a third script, Nagari. The 1925 editors of the Kotagiri plates of 1273 wrote that the letters resemble modern Devanagari in large measure. The Karimnagar plates of 1246 write even their Telugu prose in Nagari letters.

The old spelling differs from modern Telugu spelling in several ways. Engravers usually doubled a consonant that follows the letter r. Barnett notes this habit on the Ramappa pillar, and Parabrahma Sastry notes it in Telugu records.

Parabrahma Sastry lists Telugu spellings such as udiyince for udayince and varushamu for varshamu. Barnett found hardly any difference between the long and the short vowel i on the Ramappa pillar. The same engraver sometimes wrote b in place of v.

The Kotagiri engraver wrote the vowel ri as ru almost throughout, according to the 1925 editors. Readers of the plates therefore meet spellings such as pruthvi for prithvi.

The small angled mark on top of many Telugu and Kannada consonants has a long history. Burnell traced it to a short cross-stroke that early engravers cut at the top of a downward line. The stroke marked the end of the line clearly. Telugu calls this mark talakattu.

Burnell rejected the view of the Telugu grammars that the mark is a form of the vowel a. Two other letter shapes help to date a record. The modern u sign, a hook to the right of the consonant, appears in the eleventh century and prevails in the twelfth. Older records write u below the consonant.

An r before another consonant was written above that consonant until about 1300, according to Burnell. After 1350 the r is always written to the right of the consonant.

Inscriptions also use signs for numbers and units. Parabrahma Sastry shows that a vertical stroke can stand for a gadyana, a gold coin. Barnett explains a sign for the land measure martu in the Uparpalli inscription.

Numerals with place value came into common use about 1300, according to Burnell. In South Indian writing the zero is always a small circle. Burnell's earliest examples of these numerals were on the rims of the plates of a grant of 1087.

The Thousand Pillar inscription of 1163 already writes its Saka year in numerals. Line 6 reads Saka-varshamulu 1084 in J. F. Fleet's edition, and Fleet's plate shows the four numerals. Burnell noticed that in Telugu inscriptions the numeral 5 sometimes lacks its middle stroke, and the numeral 3 is generally upright.

Some poets wrote numbers as words. The poet of the Ramappa pillar gave the Saka year as earth, moon, worlds and arrows. Barnett read these four words as the year 1135.

Burnell found that from 600 to 1300, nine out of ten dated inscriptions give the date in such words. In this system the units come first, then the tens, the hundreds and the thousands. Burnell's example rishi-naga-kha-indu means 7, 8, 0 and 1, which is the year 1087.

The Ramappa text itself reads sara-loka-indu-bhu, which means arrows, worlds, moon and earth. These words stand for 5, 3, 1 and 1. Read from the last word back, they give 1135. Barnett's English gives the four words in the opposite order.

The pictures on this page show these differences. Compare the rounded Telugu letters of the Ramappa pillar, the Kannada letters of the Padmakshi pillar and the Nagari letters of the Kotagiri plates.