In 1786, Sir William Jones announced in Calcutta earth-breaking news in the
linguistic field. He stated that Sanskrit and the European languages
"have sprung from some common source which, perhaps, no longer exists."
These words set a trend in the linguistic study of Sanskrit and the use
of the term "Aryan," to describe not only the Indian languages, but
those of Europe, first came into effect.
However, the factuality of Jones' announcement will be
brought into question in this work. While the Vedic speech, or
Chandas, as Panini calls it, is an inflectional one resembling, in
many ways, the ancient Avestan, the argument here will be that
Classical Sankrit, or Bhasa, is a native Dravidian one with
heavy Austric and Indo-European influence. The same argument
will be made concerning the modern vernaculars. The term,
"Indo-European" in this context will refer to the European,
Iranian, Kurdish, Hittite, Afghan and Chandas, but not to the
other Indic languages generally classified as such.
Chandas, itself, shows heavy Dravidian and Austric influences
already by the time of the Vedas. Bhasa, also, has been very
significantly altered by Chandas, but the two are not dialects,
or branches of the same family. The theme of this work will be
that Bhasa was a highly artificial language designed by the native
speakers as a literary language. Probably it was never a spoken
language before its conception as a literary one. The thesis
here is that indigenous people attempted to create something
close to Chandas, the first scriptural language (excepting
possible Harappan texts), but with a form that more readily
conformed to their own spoken language.
As already stated, there had been already much borrowing between
the Indo-European and the Dravidian, so by the incorporation
of some features of the scriptural language of the Vedic peoples,
the indigenous peoples may have sought to create a new sacred
speech. To a certain extent they failed, as many of the features
they tried to absorb into Bhasa fell into disuse when the Dravidian
speakers were unable to use these effectively or with confidence.
The root languages in this theory would be the ancient Prakrits.
Prakrit itself means "original" or "natural". Even Chandas
possesses "prakritisms" and Panini describes Prakrits as the
forms use in everday speech. Prakrit comes from the same root that
forms "prakriti," the primordial substance from which other things
T. Burrow of Oxford, in his treatise on the Sanskrit language,
asserted that while Dravidian and other languages had influenced
Classical Sanskrit in terms of vocabulary, the structure of
Chandas and Bhasa were essentially the same. However, S. K.
Chatterji was indeed much closer to the truth when he noted
that Bhasa and the post-Vedic Aryan languages seemed altered
due to a non-Aryan people trying to speak the Vedic language.
He noted in his works that in terms of syntax and phonology,
Classical Sanskrit and the other post-Vedic Indian languages
had more in common with Dravidian than with Indo-European.
Perhaps, Dr. Chatterji was too shy to assert, against the bulwark
of Western opinion, that the post-Vedic languages were wrongly
classified. Either way, he certainly made many of notes of the
peculiar and numerous instances when these languages showed
much closer relationship to the Dravidian. Indeed, the ancient
literature itself, seems to classify Chandas and Bhasa as
different languages rather than Bhasa as a development of
Chandas. It may be that the Vedic people knew of an
indigenous Austric-Dravidian people they associated with the
Devas, in the same way they associated other native peoples
further east, with the Asuras. Thus, just as many modern
scholars conjecture a language and culture of the Asuras,
the "Deva Bhasa" may have been the language of the Devas, i.e.
the native allies of the Vedic peoples. The alliance of the
Indo-Iranian Vedics with the native Deva people resulted in a
fusion of language, but with the written scriptures of the Vedics
initially being the accepted ones. Later on, the Dravidians, in
our hypothesis, asserted their own Deva Bhasa, into the sacred
literature attempting to conform it even further to the language
of the Vedas.
We will use an argument here that asserts the importance of
morphology and phonology, and particularly morphology, over
the current emphasis on lexicon used by the genetic, Nostratic
and other schools. Our contention is that morphology and
phonology are much more difficult to borrow than lexicon,
even more so than word lists like the Swadesh list. An
in-depth investigation into this matter is beyond the
scope of this work, but we will make mention of the stability
and durability of both morphology and phonology, and the
comparative ease in borrowing lexicon.
While we will not disregard the evidence of the roots, we believe
that morphology and phonology are even more important in the
proper classification of Classical Sanskrit and the other
post-Vedic Indic languages. Our belief is that the word
"aryan" has been misappropriated by the West, much like
swastika symbol, as its roots are acutally Austric, and
probably Austronesian. To start our discussion, we can offer
first a lexical comparison of the word "bhasa" itself, showing
that rather than being an Indo-European root it actually has
Austric origins: bahasa "language," Indon., Malay.; basa
"to read," Phil., also basahin "to read;" basa "language,"
Kawi; vosa "to speak, say , word, language," Fiji; waha
"mouth, voice," Maori; waha "saying, word, mouth, voice,
language," common Polynesian; vasa "to speak," Sesake,
vasana "speech," visiena "speech," Api; bosa "to speak,"
Florida, Ysabel; bacah "language," Proto-Philippine,
phaasaa "language," Thai; -bisi "to say," Visina, Mapremo,
Nikaura, basa "to speak," Efate.
The idea with this example is to demostrate that a great deal
of the proposed IE lexical correspondences between the Indic
and European languages actually have other explantions rather
than genetic relationship. We will use what the term
"Oceanic line," to describe the limits of Hindu influence
east of India, and this barrier will help us establish that
many roots often considered IE actually come from the Austric
substratum in India.
Bhasa as Agglutinative
While there are certainly elements of inflection in Bhasa, and
some words that have a rather complete inflection borrowed
from Chandas, most words in Bhasa and the vernaculars follow
agglutination as the principle form of conjugation. In such
cases, the root of the verb is not inflected although the
suffix or affix may be inflected as it is in the Dravidian.
Only under the strictest classification of agglutinative
languages, in which no inflection at all occurs regarding the verb,
can Bhasa not be considered agglutinative. Indeed, using such
criteria, most widely-accepted agglutinative languages would not
fit into this category. But recognizing that some borrowing of
inflected roots occurred between Bhasa and Chandas,
there is no other alternative but to classify Bhasa as agglutinative.
In many cases, modern Sanskrit grammars written in the West
confuse the case by not adequately separating the form of Bhasa
from Chandas in their works. It seems clear, though, that Bhasa
was the main subject of Panini's work, and that Chandas was
included possibly to help illustrate the likeness of Bhasa with
the older sacred language. According to our theory, the
likeness was in some cases due to borrowing of traits, while
in others it was due to the artificial implementation by
Panini and others in order to create a native religious language.
Panini made no grand declaration of any such intentions, but this is not
exceptional as we can see the same type of subtle introduction of
indigenous beliefs in the Atharva Veda, Upanishads, Epics, Puranas
and Tantras. In most cases, the agents responsible for such introduction
attempt to give a Vedic, or even divine source for the new systems in order
to give them legitimacy. A great deal of this work is even classified
as sruti by its followers, the Sakta Upanishads, for example.
Just as the Puranic and Tantric deities and rituals eventually
superseded the Vedic Indra and Varuna and the homa offering, the
indigenous Bhasa also supplanted the Chandas of the IE speakers.
Again, due to the confusion wrought over the mechanics of Chandas
and Bhasa, it would seem that Classical Sanskrit bore great
resemblance to other inflectional languages in its morphology.
Actually, this is not true due to the fact that Chandas itself is
already a language deeply altered morphologically from other IE
languages. The effect is due undoubtedly to the influence of
agglutinative languages, and it resembles the same process that
was found in the transformation of Latin to Italian, and ancient
Persian to the modern Farsi language. However, these latter examples
were not nearly as altered in terms of lexicon and, especially,
phonology. Chandas, on the other hand, was very powerfully
influenced in all these, and also in idiom.
However, if Chandas had other examples of IE languages which
followed similar, but less extreme, paths, Bhasa in practical
usage followed a different road altogether. For although Bhasa
could utilize all of the same verb morphologies available to
Chandas, according to Panini�s system, it reality the usage was
closer to that of the Prakrits and the modern vernaculars. Like
the Dravidian languages, the participle was used mostly in
passive and/or past constructions as the primary form. Of
particular interest is the Hindi active participle construction
using the suffix, -ta. The suffix in different inflected forms
is generally agglutinated to the end of the uninflected verb root.
In the IE languages, the verb base, which, unlike the Bhasa-based
languages and Dravidian, mostly uses the infinitive, has sounds
inflected right into the root. Although this is only one of the
many morphological differeces we shall discuss, it should not be
Here are the suffixes for the present tense conjugation of the Hindi verb:
1st -ta hum -ti hum
2nd -ta hai -ti hai
1st -te-haim -ti-haim
2nd -te-ho -ti-o
3rd -te-aim -ti-aim
The Dravidian and Munda languages also appear to have conjugation
suffixes which are similar to the -ta of Hindi. For example, the
present conjugation of the finite verb in Khaia and the present of
Bhasa has in its primary conjugation (Bhu class) a number of
suffixes which resemble the -ta suffix:
The full conjugation looks like this:
Sing. Dual Plural
1st bhava-mi bhava-vas bhava-mas
2nd bhava-si bhava-thas bhava-tha
3rd bhava-ti bhava-tas bhava-(a)nti
Five of the nine suffixes in this most common of verb conjugations
shows relation to the -ta suffix. The general preservation of the
root by means of "gluing"on suffixes is a characteristic that can
only be considered as agglutinative. It has been noted that modern
Iranian and Italian show evidence of morphological change similar
to that found in Chandas and Bhasa. However, scholars like
Chatterji have rightly noted that such changes can be traced
to the influence of agglutinative "Turanian" speakers who were
known to have invaded the concerned linguistic regions for many
centuries. We need not go into the historicity of Ural-Altaic
invaders in either Italy or Iran, or the presence of pre-inflectional
agglutinative languages like Etruscan and Elamite being present in
In practical usage, Bhasa also made very liberal usage of the
-ta suffix in past participle constructions in manner similar to
common grammatical usage in Dravidian In Bhasa, sentences are
often constructed without active verbs; something usually avoided
in IE language. The use of long compound chains and mixed compounds
is also clearly an agglutinative trait. For example, here is a
line from the Mahabharata:
"from the yoga-master, from Krsna speaking directly"
Such long chains used as sentences are practically not found in IE
languages, but are part and parcel of a great many agglutinative ones.
In the formation of chains and compounds many rules found in the
grammar of IE languages are gravely violated.
Bhasa shares with the Dravidian and Munda languages a trait of
inserting additional suffixes in a chain before the normal suffix.
For example, in the Ya- class of verbs, the suffix -ya- is inserted
ahead of the regular suffix, i.e., div-ya-ti.
Indeed, in both Dravidian and Munda, the elements of the suffix
chain are nearly always placed before the regular suffix just as
in the Classical Sanskrit. Such suffixes in Bhasa include
-aya-, -sya-, -ant-, -tavya- ,-aniya- and -ya-. Kharia offers
a good example of how these suffixes are placed in front of the
regular conjugation suffix:
Kharia Finite Verb
Future Past Present
-ing -o-ing -t-ing
-m -o-m -t-am
-e -o -ta
-jar -o-jar -ta-jar
-nang -o-nang -ta-nang
-bar -o-bar -ta-bar
-kiar -o-kiar -ta-kiar
-le -o-le -ta-le
-ning -o-ning -ta-ning
-pe -o-pe -ta-pe
-ki -o-ki -ta-ki
Suffix or prefix chains used in verb conjugation in IE are rare,
if found at all. The author is not aware of any examples other
than Chandas, where it is likely due to indigenous Indic influence.
Of course, it is in the root-aorist that we have the true inflectional
category of verbs, and this class is practically extinct in Bhasa.
Note here the conjugation of the root-aorist and compare the
inflection of the root with the agglutination of the Bhu class above;
Single Dual Plural
1. Asravam, agam, ahema akaram akarma, adama
2. Agas, asres, akar agatam, kartam akarta, agata, ahetana
3. Asrot, asthat, akar akartam, adhatam, akran, aksan
The fact that this obviously heavily-inflected category was nearly
absent from Bhasa, and that the agglutinative conjugation displayed
previously was the predominant form goes far in giving a true picture
of the morphological character of Classical Sanskrit.
In reference to the passive verb, we may return to the -ta suffix
addressed previously. The use of the passive participle -ta and
the future passive participle -tavya eventually came to dominate
Sanskrit literature. The suffix -tavya in particular occurs first
only in the Atharva Veda and is mainly found in the classical languages.
Some have commented that the change in use to the pasive constructions
from the multiple verbal forms was brought about by evolutionary
processes, but it is much more likely that this change testifie to the
true morphology of the Classical Sanskrit and the other Bhasa or Prakrit
derived languages. This can be understood if we consider that the
syncretism so common in Indian religion and culture also extended
to language. Instead of placing a great barrier between the
inflectional Chandas and the agglutinative Bhasa, Panini may
have tried to explain them all in the same terms. He knew nothing
of modern division of languages. He simply tried to bring under
the scope of one work the two primary langauges in Indian literature
at the time. Eventually, as Chandas became mostly a liturgical
language, Sanskrit grammarians completely ignored it or included it
only in appendices. The work of Panini sufficed. However, as Bhasa
continued to be used in the literature, and other forms continued in
spoken form, they became the focus of all post-Paninian works.
Indeed, even Panini himself was mainly interested in fixing rules
for the use of Bhasa.
As in religion, Panini did not attempt to draw clear lines
linguistically based on ethnic or racial considerations. He
introduced the indigious, agglutinative elements with all the
subtlety that Krisna introduced the puja offering and the bhakti
mode of worship. When Sanskrit grammarians refer to desi words,
they are not speaking about native indigenous words, but of those
which did not belong to either of the two original religious languages,
one of the inflectional IE, the other agglutinative Dravidian; both
influenced by each other and also by the Austric.
The other source of confusion comes from the Chandas language already
heavily influenced by Dravidian and Austric by the time it arrived
in India. This is made clear by the presence of cerebrals, a
highly-agglutinative morphology and a large Austro-Dravidian
In addition to accepting foreign morphological traits from Dravidian,
it appears that in the area of verbal reduplication, Chandas
displayed Austric influence. The fact that reduplication exists
to some small extent in Greek and a very few other IE languages need
not lead us to believe that it was a native feature. Indeed, there
is every reason to believe that just as agglutination of suffixes
in Iranian and Italian was due to foreign influence, so the minimal
reduplicaton found in European tongues may also be a result of
borrowing. We might add that in the East, the Thai and Japanese
languages also show strong influences from foreign morphologies.
Reduplicaton, by its rare occurrence in the IE languages, and by
its secondary position when it does occur, is highly indicative of
borrowing. The use of partial reduplication of the first syllable
and of, in most cases, apophany clearly relates to the practice
found in Sumerian and Austric verbs. The lack of apophany in
Greek and Hittite seems to be a corruption of the original use
which is better preserved in Chandas. We might note that this
trait is not common in Bhasa, the Prakrits or modern vernaculars,
due no doubt to its absence in the Dravidian forms of speech.
The use of a single suffix -tum, to mark the infinitive places
Bhasa much more close to the Dravidian and Munda than to the
Chandas or other IE languages. The rich use of infinitives is
very widely-attested to in the Indo-European. One of the main
characteristics of the IE languages is there emphasis on
specificity particularly with the verb. Numerous inflections exist;
in some languages there are inflections for practically every possible
facet. The other quality is the general emphasis on separating
the different elements of speech. In agglutinative languages,
simplicity takes priority over specificity. And there is a tendency
to unite the different elements of speech.
To foster the requirement for simplicity, the agglutinating elements
tend to be similar or derived from similar roots. We see this in
the conjugations given above for Hindi, Kannada and Kharia. Both
vertically and horizontally, when looking at tables of conjugations
in agglutinative languages, one tends to find correspondences in
the agglutinated elements. We have already mentioned the importance
of -ta as a root for certain suffixes, especially in the present
active construction. After Panini's likely attempt to create a
Vedicized indigenous sacred language from one of the native tongues,
Bhasa began to shift from the artificial conjugations back to the
use of the -ta suffix by using the passive participle constructions.
Many scholars have recognized this as a way for the native writers to
compose literature in a manner which resembled their actual daily speech.
Thus, in opposition to the normal forms in Chandas, Bhasa expressed
phrases like "he wrote the book," as "the book was written by him."
It may be no coincidence that the most common suffix for the absolutive
in Dravidian is -tu compared with the Sanskrit past participle active
-ta; and that the common suffix for the relative participle is -a,
which deals mostly with the past, and corresponds with the Hindi
past participle passive -a. The latter is also used in some gerunds
and in the accusative of the verbal action nouns in Classical Sanskrit.
Another feature that displays the emphasis on simplicity in agglutinative
languages is the common non-distinction between noun and verb.. This is
not universal, but it follows the same principle as the non-inflected root.
That is, to render the fors easily recognizable for memorization. This
feature is also found in IE, for example, the word, "sailor," comes from
the verb, "to sail," However, there is a great prevalence in IE, more
than any other language family, to form completely distinct verbs and nouns.
Often this is accomplished by using nouns derived from foreign or obsolete
words as in "food," from the ancient IE word "to eat," rather than
In agglutinative languages as we see resistance to such tendencies, in
favor of maintaining some recognition between nouns and corresponding
verbs. In Bhasa, we have a great many words that follow this trend
such as budh "to fathom, penetrate, understand," and buddha "wisdom,
knowledge." A statistical analysis of the percentage of verbal nounds,
nouns of agency, verbal adjectives, etc., in Bhasa, in comparison to
statistics for inflectional and agglutinative languages in general,
would be very helpful in understanding the relationship of Bhasa.
In the formation of compound nouns, Bhasa again shows its
agglutinative nature. While certainly the formation of compounds
can be found in the inflectional languages, the process there
is of an entirely different magnitude than what usually occurs in
the agglutinative languages. This same divergence in magnitude and
scope is mirrored in the difference in nomimal composition in
Chandas and Bhasa. For, if in the verbal and other forms, the
inflectional languages, due to their specific nature, may be much
more varied, in the case of forming compounds, the agglutinative
languages have found this an excellent tool in maintaining
In this sense, we can see the great similarity between Bhasa
and the Dravidian languages. There is no similarlity between
the principal types of nominal composition in Bhasa and
Dravidian languages, found in the IE languages, with the minimal
exception of Chandas. In the IE, compounds are established forms
which are brought into speech over time. In Bhasa,
compounds may be constructed as one speaks. In the IE, compounds
are established forms which are brought into speech over time. In
Bhasa, compounds may be constructed as one speaks. For example,
one does not say "cornerstandingman," instead of "a man who stands
at corners." In Bhasa, there are classes of compounds that can be
constructed at will. Also there is nothing like the multiple
compounds can be constructed in Bhasa in the IE languages. It
may be true that certain IE languages construct long compounds
as place names and as novelties, but not in regular literature or
speech. It may be true that certain IE languages construct long
compounds as place names and as novelties but not in regular
literature or speech.
Some Western Sanskrit scholars have erroneously tried to give an
internal origin for the Bhasa nominal composition, but better
scholars like Chatterji have at least admitted these have come
from an indigenous influence. This thesis here, of course, is
that Chatterji should have gone one step further and simply labeled
the practice as a Dravidian one.
Let us take, for exaple, the Dvandva compounds from Burrows and
compare them to similar compounds recorded by Bloch. Some examples
pitraputrau "father and son," Sanskrit
ingyoembar "father and mother," Kurukh
Other examples of compounds are:
madhuhastya "having honey-like hands," Sanskrit
mudimelan "who has the crown," Kannada
punyalaksmika "having auspicous marks," Sanskrit
nirvarkannem "having eyes streaming with water," Tamil
rastradipsu "injuring the kingdom," Sanskrit
arraludeiyor "possessing force," Tamil
sahasaivashyahanyanta "suddenly they sounded simultaneously," Sanskrit
perumpunenukku "to me who has a great ornament," Kui
In terms of declension and other elements of the noun, there
is the same tendency toward attaching them as prefixes, suffixes,
etc., as found in other agglutinative languages. The interesting
occurrence found in both Dravidian and the Indic languages is the
correspondence of morphemes used in one type of morphology, the
declension of the noun, for instance, with another system like
the case markings of the personal pronouns:
Noun delension Personal Pronoun (1st)
Nom. pit-ay am-i
Gen. pit-ar am-ar
Loc. pit-ay am-ay
Nom manid-an n-an
Acc. manid-anei en-nei
Dat manid-anukku en-akku
Gen. manid-anin en
The vast use of compounds of nouns with adjectives, verbal
adjectives, etc., displays the common desire in all agglutinative
languages to unite certain ideas, or to give certain emphasis or
power to certain words, by uniting them into a single unit.
Thus, in religious or other literature, when a point need
stressing, or great emphasis is to be added to a verse, the
composer will often resort to long, compressed compounds.
This is found repeatedly in agglutinative languages around
the world. Compressing many ideas into a terse compound which
can be interpreted in more than one way is a part of the
ethnological make-up of the agglutinative languages.
In judging the origin of many Bhasa and Chandas words, we must
bring into question the words that contain cerebrals. It is
fair to say that even the earliest Vedic literature was as
different from the main body of IE as any other IE language
in history. This is especially true in the area of phonetics
were the cerebral, unknown in IE is already strongly attested
to in the Vedas. The question is whether the words containing
cerebrals are actually non-IE in origin. Most Western
scholars explain the use of cerebrals in words athat are
claimed as IE, is due to the change of pronunciation by the
IE speakers themselves. Chatterji and other have made a
more reasonable hypothesis of the IE speakers imposing their
language on a predominantly Dravidian and Austric populace
who could not correctly pronouce the IE. This makes sense
as even the modern Indian Gazeteer, published by the Indian
government, classifies the Dravidans and Austrics as the
"bedrock" and the "basis" of the Indian population. However,
the one problem with this scenario is that while there certainly
can be no doubt that the indigenous folk may have massacred a
foreign tongue, so to speak, there is no reason that the IE
speakers themselves should have done the same.
Unless, we accept the suppostiion that the literary and ruling
class in Vedic India was primarily indigenous, we cannot see the
IE speakers suddenly mispronouncing their own words! It may be
true that if an Englishman were raised only speaking Hindu, and
then were introduced to English as a second language he might
make some mispronunciations of a certain order. But, the IE
community in India cetainly seemed large and well-established
enough to properly preserve their language. Even modern Indian
do not use cerebrals when speaking English as a second language.
The more plausible explanation is that the IE speakers, who had
already drastically altered their language, accepted a large
corpus of Austric-Dravidian words with cerebrals. That such
words, or related words, am appear in other IE languages does
not discount this theory. They may have been borrowed and
pronounced the only way the non-Indian IE speakers knew how to
pronounce them. But in India, the situation was such that the
proper pronunciation would be preserved through close, continous
contact with the original speakers.
Certainly, there is also the possiblity that some true IE words
were mistakenly categorized with cerebrals by the scribes. But
any wholesale conversion of IE words by mispronunciation seems
untenable. In general, the rules governing the cerebrals in
India are the same found in the Dravidian. The main exceptions
to this are the relatively few words with cerebrals as the first
letter. This formation is not found in the Dravidian. The
Austric languages may provide the answer, and this will be
discussed more in the section on phonology.
There are some interesting parallels in form between the pronouns
in Bhasa and those in Dravidian and Austric. These similar forms
do not always correspond to usage, however. But this same
characteristic is found in the Dravidian and Austric languages
themselves. For example, avam is the singular masculine nominative
of the demonstrative or third person pronoun in Kannada; however in
the inferior group, avam signifies the plural accusative, in Tamil
avan stands for the singular masculine-feminine nominative. In Bhasa,
avam is found as the dual construction of the 1st person dual pronoun.
In Old Kannada, the 1st person dual pronoun am, bears
similiarity to the same pronoun in Bhasa, and theuse of the
suffix -am in the declesion of pronouns in both Bhasa and
Dravidian is to be noted. Also, in relation to the ma- based
first person pronouns in Sanskrit, the common na-based first person
singular pronouns in Dravidian would corresond by the
interchangeability of |n| and |m|. This interchangeability can
be clearly demonstrated in these 1st person plural inclusive
pronouns in the Dravidian:
Old Tamil nam
Old Kannada nam-
Likewise in the second person oblique plural:
If we exchange the "n" with "m" we can imagine how these first
singular pronouns in Dravidian and the modern vernaculars are
ma - Danni na - Korava
mai - Wl Hindi nan - Tamil
ma - Tali na - Kannarese
mu- Pahari nanu - Badagi
mama- Singhalese nanu - Irola
The trait which Classical Sanskrit shares with Dravidian and Altaic
is its unusal declension of the pronouns. Kannda, for example, has
inflection for the nominative, accusative, dative and genitive of
all the pronouns, even the interrogatives have genitive, accusative and
dative inflections. In this snese, the Dravidian, and also the Munda,
depart from the usual simplicity of agglutinative languages. However,
most inflectional languages do not use case inflection with pronouns.
It might be noted that in agglutinative languages, the pronoun is
generally the most specific element in speech. In Austronesian, it
is the pronoun which usually is the only ellement to show person and
number. Although it does not show gender or case, it does have
inflection for the inclusive and exclusive similar to many of the
Munda languages. Also, it can be very specific in terms of number.
For example, the Melanesian has four numbers, the single, dual, trial and
plural. Thus, the complex case sytem of the pronouns found so extensively
among both the Dravidian and Munda languages may be a local
extension of this specificity found in agglutinative languages,
but here including case also.
Another area in which Bhasa and Dravidian show common
origin is in the similarity between the pronouns and verb
terminations. For example, here are the pronouns and the
verb terminations in Telugu as given by Bloch:
1 enu -n, -nu
2 ivu, nivu -vu, -vi
3 vadu, adi -du, -di
1 emu -mu
2 iru -ru
3 varu, avi -ru
This same feature can be seen in the Munda languages, such as Kharia:
1 ing -ing
2 am -em
3 adi -e
1 injar -jar
2 am(b)ar -bar
3 ar-kiyar -kiar
1 ele -le
2 ampe -e
3 arki -ki
In Classical Sanskrit the relation looks like this:
1 aham (ma) -mi
2 tvam -si
3 sah (ta) -ti
1 avam -vas
2 yuvam (tvam) -thas
3 tau -tas
1 vayam (ma, sg.) -mas
2 yuyam (tvam, sg.) -tha
3 te -an-ti
While there is a definite similarity between the verb terminations
and the corresponding pronouns, Classical Sanskrit does not show
this feature in as pure a state as most Dravidian languages, or
even as much as Kharia. Among the latter Indic languages a
simplified pronoun system developed without the full inflection
of case, person, etc., found in Bhasa and Chandas.
Also, most lacked any relationship between the pronouns and the
verb terminations. Needless to say these terminations are very
different in nature than those in Bhasa, and this may be due
to evolutionary changes, other foreign influences or simple variation.
The use, in Chandas, of prepositions as verbal prefixes in prefix
chains is most probably a borrowing from Austric/Sumerian influences.
The fact that prefix chains may occur rarely in other IE languages like
Greek does not refute this, since it was probably borrowed there as
well. The Greek, of course, is not pure language, and has already
been demonstrated to have an enormous substratum influence.
Certainly many of the prepositions used in these prefix chains were
of IE origin, but the constructions are foreign. When the Bhasa
languages of India took over, the prefixes were permanently attached
to the verbs as the Dravidian makes no use of them except in words
also borrowed from Chandas. Thus, we find the indigenous
inflences of Indian displacing earlier influences of probable
Austric provenance that occured very early on among the Vedic peoples.
Few areas support the contention of the Dravidian provenance of
Classical Sanskrit and other similarly derived languages as that
of phonology. Only the morphology and grammatical structure of
the languages is more convincing.
The theory that the cerebrals in Bhasa are due to natural evolution,
and the use of the Scandinavian examples as proof, can safely be
rejected. First of all, it can hardly be coincidence that the
Indic possess cerebrals in abundance like their Dravidian and Austric
neighbors while the Iranians and Afghans do not. Secondly, there
is no real proof that the cerebralization of dentals in the
Scandinavian languages is not due to borrowing from some language
that has not survived. We know that there are many non-IE languages
in this region and that they have been there since for quite a
long time. Even today, there are non-IE speaking Lapps right
in the heart of Scandinavia.
Many so-called IE words in Sanskrit have only connection with the
Avestan, and this evidence tends to show that this was due to
late Indian influence in Iran, rather than vice a versa. While it
is beyond the scope of this work, the author will only mention his
belief that Avestan, like the Gypsy dialects, was an Indic language
that migrated West out of India at an early date.
By its use of cerebrals even Chandas differs more in its
phonetics than anyother IE language does from the common stock.
Classical Sanskrit makes slightly more liberal use of the cerebrals
than does Chandas, but differs mainly in the use of the "l," for "r,"
the softening of hard consonants, the dropping of the final consonant,
the use of long vowels, the use of apophony (wrongly assigined to
ancient IE) and in the use of the svarita accent.
The existence of cerebrals at the beginning of a word is a feature
that cannot be explained by the Dravidian.. The cerebrals are, of
course, formed by raising the tip of the tongue and drawing back to
the top of the mouth. However, there is little similarity between
the sound of the cerebral in Scandinavian with that of the modern
Indian vernaculars or Bhasa, which on the contrary are very
close to the sounds produced by the Dravidian and Austric
Following is a list compiled by Chatterji (On the Development of
Middle Indo-Aryan, pg. 61) that he classifies as shifts from OIA
to MIA, but which the present author believes mostly describes a
failure of artificial alterations to Dravidian languages. In
some cases also, natural foreign influences from IE may have
given way to Dravidian resurgence, particularly due to pressure
from the South and East. In some cases, the phonological changes
probably are not due to Dravidian at all, as in the case of
initial retroflex consonants:
(1) Vowel quantity subject to speech rhythm and to quantity of entire syllable.
(2) Pitch accent gives was to fixed stress accent, also affecting vowel quality.
(3) Assimilation of consonants in certain groups and clusters.
(4) Cerebralisation of dental stops and aspirates through contigous r-sound.
(5) Voicing of intervocal consonants.
(6) Loss of intervocal stops and aspirates.
(7) Flapped pronunciation of intervocal -d(h)-.
(8) Spontaneous nasalisation.
(9) Weakening of final vowels.
(10) Change of intervocal sibilants to -h-.
(11) Change of simple -m- to mere nasalisation.
The Bhasa and Dravidian syntax have a great many things in common.
Chatterji nots that the Bhasa languages are much closer to Dravidan
in this regard than to any IE languages. Even Jules Bloch admitted
that the Sanskrit sentence with its "successive inclusions and with
unique agreement of the Dravidian," likely was based at least on
what he calls a "psychological model" provided by the Dravidian.
Nouns in both the modern Bhasa languages and in Dravidian have
special functions which are dependent on oblique suffixes and the
postposition of additional words. We have already discussed the
copious use of long compounds, a trait which the Bhasa languages
share with Dravidian and other agglutinative languages.
When both languages studied by speakers of inflectional languages
there will be equal difficulty in identifying the regular forms
which are common in the inflectional. The action words will often
be missing; the great differentiation of function found in the
Chandas and IE languages will be mostly absent, the sentence will
be simple and terse and with alternate meanings; and long
compound phrases completely unlike anything in the inflectional
tongues will appear repeatedly.
Although the subject can be placed before the verb, it is common
for it to come afterward. For example:
sidanti mama gatrani "are quivering my limbs"
instead of: "my limbs are quivering," or:
etan nirikse ham "all these may view I"
instead of "I may view all of these"
Austric langauges often have this feature. For example, in the
Philippines we have:
bumasa ka lahat "read you all these" Tagalog
mamangan siya "is eating he/she" Kapampangan
In this regard, the Dravidian differs as is almost always placed at the
end of the snetence after the both the subject and the object. The
Sanskrit feature of placing he subject directly behind the verb may
be due to Austric influence, possibly even Austornesian, which regularly
uses this order. Eventually, though the written Bhasa-derived
languages reverted to the spoken form in which the verb came at the
end of the sentence, as it does in Dravidian and in the Munda
languages. Hindi, for example, places the verb after the subject.
Even in early Classical Sanskrit this feature was already found:
pralinah tamasi mudha-yonisu jayate "being awash in ignorance,
among animals is born."
Sa gunan samatityaitan brahma-bhuyaya kalpate "he the
modes of materialsim transcending all these to the status
of Brahman becomes."
It is, of course, rare to find the SOV word order of these verses
in any IE language. We have already discussed the particular types
of compounds based on apposed terms. Some more examples from the
mai-mansal "woman and man"
bai-mui "mouth and nose and face"
Both Dravidan and the modern vernaculars use a great deal of
idiomatic expressions which are very similar in meaning and
form. They are quite unlike anything outside of India and
thus they probably originated among the indigenous speakers. Here
are some examples in Gond as given by Bloch:
rohci simt "having sent, sent"
si simt "giving give"
arsi hattul "falling he was, he falls,"
hanji mandakat "having gone, we shall stay"
Bloch also gives the following description of the Dravidian proposition:
"The sentence is variable in dimension and form. It can consist
of a single word, which is not necessarily a verb; the verb 'to
be' in particular can be missing."
Such a description would also fit Bhasa when one considers the compound as
a "single word." The only major difference is in the word order of Bhasa,
which was unlike Dravidian or IE, but resembled certain Austric languages
like the Austronesian. Eventually though, the Dravidian word order came
to the fore in the later Bhasa languages. The Dravidian proposition
contains fragments which are formed into long compound, something which
mirrors the lengthy compounds of Sanskrit literature. Eventually, in the
modern languages, all elements in the sentence: proposition and nouns and
verbs themselves, came to have their morphological determinants placed at
the end of the order just as in Dravidian, and to a lesser extent, the
Munda languages. The primary exceptions were noun forms with prefixed
prepositions that came in from Chandas and Sanskrit.
Indeclinables of Place
Indeclinables denoting place and time are formed using the ending
-tra in Sankrit:
here - atra
there - tatra
where - kutra
everywhere - sarvatra
in various places - bahutra
at one place- ekatra
wherever - yatra
in another place - anyatra
in heaven - paratra
In Austric languages, here and there are often denoted with elements
from a root like ta/te/ti/to/tu or ka/ke/ki/ko/ku and the like:
Sud-Est e|ke| e|ko|
Budibud i|to|n |to|none
Kukuya taina tanoi
Tawala geka noka
Garuwahi wedahosi nodahosi
Sinaki nekai wakai
Duau beka yoka
Kurada tenina tenem
Dobu gete gote
Mwatebu iga nage
Galeya kamele kano
The roots in Bhasa, having a great many roles to play allowed words to
take on a substantial number of meanings, a process which had already
begun in Chandas. This meanings, a process which had already begun in
Chandas. This, however, is not a typology of IE were the tendency is
toward specification. It is not unusual in Bhasa to find words, with
over 20 different meanings.
The compounding of words in Bhasa also showed the highly-agglutinative
nature of the language, a process which seemed only to be borrowed in
Chandas. The use of the absolutive or past particple in the proposition
becomes firmly established in the Bhasa and related langauges, Indeed.
in the use of the past tense, and the absolutive we see a tremendous
correlaton between the Dravidian and Sanskrit of a nature comparable
to the general non-inflection of Bhasa verb roots. Also, the
prevalence of SOV word order and the frequent occurrence of the
subject occurring after the verb, as in the Austronesian languages,
also mark tendencies which can safely be classified as non-IE.
The nature of agglutinative languages is that they provide simplicity
and unity in spech at the sacrifice of specificity and independence of
forms. Often the tendency toward secrecy in these languages, stemming
from the general associated culture, is accomplished through multiple
meanings, word formulae, etc. All these features are found in Bhasa,
but are not typical of IE languages.
The morphological similiarities between Bhasa and the derived
languages, with the Dravidian cannot be easily put aside. The
evidence shows that morphology is not as easily
borrowed as modern philogists assert. Indeed many thousands of
tribal languages have retained their morphological and grammatical
structure despite centuries of highly-intrusive exposure to Western
culture and language. Prof. Chatterji (Indo-Aryan and Hindi) states
concerning the difference of the verb in MIA as compared with OIA,
or really, mostly Chandas rather than Bhasa:
"...the past tense of the transitive verb in this from was
really in the passive voice - in the formation of the past,
therefore, the verb became in its nature an adjective. In this
matter, Aryan altered itself in the direction of the Dravidian
habit which saw in the verb an adjective."
Our question is whether this was the true nature of Bhasa and the
Prakrits, and whether the inflected forms added to Bhasa were not simply
done so to make it look more sacred like the liturgical Chandas.
Speaking of the use of verbal and nominal post-positions, Chatterji
"This post-positional habit, if it may be so called, brought the
Indo-Aryan speech nearer to Dravidian and Austric (Kol); and
in later MIA. their number was on the increase, so much so that
a good number of these, mostly nouns and a few verb forms, were
in use widely over the Aryan language area. In the NIA. stage
there were more addtions of verbal post-positions (of the type
of Gujarati thi and thaki), and this was a still greater
approximation to Dravidian."
Again, we have to wonder whether this feature, despite its apparent
chronological development, was really a native feature of the
Bhasa-derived languages rather than a borrowed one. In Sanskrit,
and possibly also some of the Prakrits, there may have been a
conscious effort to use the inflected forms that came through the
influence of Chandas. This may have been much
stronger in the written language rather than the spoken. With time,
the emphasis on imitating these forms may have faded, and the written
language and the spoken came to accord more and more with the NIA
languages. The difference in morphology of the verb between Chandas
and Bhasa was most striking. The subjunctive, inflected past forms,
aorist and imperfect were not found in Classical Sanskrit. The
optative and perfect were only meekly represented. The middle voice
did not exist and the passive was found only with the indicative
present, although in NIA the perishrastic passive takes the place
of the inflected passive. The dual number, which seems to have
Austric connections, was dropped from the verb and also from the
noun and pronoun. The passive participle took the place of the
inflected past in Bhasa, and although the inflected future was
used, it gradually gave way to the future passive particple and
post-positional forms in most NIA languages. Prof. Chatterji
points out concering these apparent changes in Development of
Middle Indo-Aryan (pg.92):
"Functional simplification was brought about in MIA in the
above way, and herein unquestionably a good deal of non-Aryan
-- Dravidian, Austro-Asiatic and Sino-Tibetan -- influence
can be legitimately surmised. The preponderant use of the
particple made the spirit of the IA verbal construction change
from the purely verbal to the adjectival, which is charateristic
of Dravidian. The wider use of conjunctives and other gerundial
forms and of verbal nouns which gradually became quite the
characteristic of OIA as it is of NIA also points to the
In the declension of the noun, a similar process occurred with
agglutination taking place of inflection. On this Chatterji states:
"This agglutinating habit came into prominence from Transitional MIA, when
nouns, adverbial words, and verbal formations (participles and
other forms) came to be added to the noun to indicate case."
Later, these forms decayed into inflexions, but new forms were brought
into play as Chatterji surmises: "...these became the genuined and
distinctive case-indicating post-positions, in NIA, comparable e.g.
to the post-positions in Dravidian. "
Despite the availibility of numerous inflections in Bhasa, they
were often hardly used at all, especially in the later period. This
may have been due to the difficulty of writers being unable to
incorporate forms they rarely, or never used in their regular daily
speech. That conscious efforts were made to "Chandasize" speech
elements occurred can be seen in the modification of Prakrit words
that were introduced into Bhasa. Chatterji states on this matter:
"...a whole host of Prakrit roots and verbal bases both of Aryan
and non-Aryan or uncertain origin were slightly altered to
look like Sanskrit and bodily adoped."
Some examples of these words are vata > vrta and lanchana > laksana. In
the hybrid Sanskrit that developed in latter Buddhism, we see this same
process take place in its entirety. This is not to say that all the
phonological and morphological IE elements in MIA were artificially
introduced. Certainly some were the result of genuine natural borrowing,
but there was also an artificial attempt that further complicated the
situation. Eventually, the indigenous nature of the languages
reasserted itself, not as a result of IE languages adopting new
native elements, but of elements introduced local languages
gradually fading away.
We also see many addtional grammatical elements from Dravidian that were
used in Bhasa such as the agglutination of different types of words in
writing and speech:
yathakasa-shtito "just as situated in the sky,"
In IE languages, compound come as pre-bounded forms of speech, although
Chandas had already shown some influence in terms of compounding in
the Vedas. We can give one last example to illustrate the close
correlation in morphology and sntax as represented by the use of
the past participle together with a related verb in both Bhasa and
drstvedam manusam rupam tava sauyam janardana idanim asmi samvrttah
"seeing this human form of yours, very beautiful, destroyer of enemies, now I
phalam tyaktva minsinah janmabandhavinirmuktah
"results giving up, great persons from birth and death are freed."
pavu kacci aransanu sattamu "the serpent having bitten him, the king died."
kanda sukham aduda "having seen the dear one, the joy was produced."
In demonstrating the correspondences between the Indo-Aryan and the
Austric languages, we attempted to show that many of the lexical
links between IA and IE were, to say the least, suspect. Not that
every one of our examples was necessarily considered IE by all
sources, e.g. bala, is listed as of Dravidian origin by some
specialists. However, over all many words seen as IE cognates may
not be so. On the other hand, the number of Dravidian words in IA,
which has been aptly demonstrated in other works, seems much
stronger although a close examination of this would have been
to lengthy for this work. As many have already postulated that
Dravidian, Austric and other words already made up a large portion,
and among many experts such as Kittel, Kuiper, etc., even the
majority of the IA lexicon, the study we have presented further
erodes the basis of classifying IA with the IE family. Indeed,
this basis has from the start been based mainly on lexicon, and m
any scholars, including Chatterji, have noted that in terms of
phonetics, morphology, syntax, idiom, etc., the IA languages,
excluding Chandas, have more in common with Dravidian than with
IE. Of course, most have assigned this similarity to external
influence, while this work suggests that the similarities in the
MIA to Chandas were due to the borrowing of both language forms
from each other. That is, the Prakrits, as Dravidian languages,
had borrowed from Chandas and other related forms, and these
latter languages had also borrowed from the Prakrits, in some
early epoch. Bhasa, or Classical Sanskrit, which looks like
sort of an intermediary stage between the two, actually was,
in our estimation, and attempt to make a local Dravidian language,
or Prakrit, look like the ancient liturgical Chandas. We have
already given examples of MIA words that were absorbed into
Sanskrit and altered somewhat to give them an appearance of
Chandas by altering the doubled consonant. It may be that
the entire corpus of Sanskrit words, which are believed to
have undergone a reverse process, such as ratra > ratta need
to be re-examined.
On the other hand, dharma > dhamma seems like a legimate example
of a Prakrit assimilation of the second consonant. The author�s
own investigations into Austric languages reveal that the latter
form is the more natural and "primitive" one. Indeed, in
the case of the Polynesian and Melanesian languages in
comparison to the ones of the Malay Archipelago, we seen
many examples of single or double consonants from the former
groups, probably representing the earlier and purer form of
Austronesian, breaking up into consonant clusters with varying
Without the lexical basis, the IE classification of Bhasa and
related Indic languages fails miserably. This becomes more
apparant when we examine the morphological and phonological
correspondences between MIA and Bhasa, or at least the late
developed Bhasa, with the Dravidian. Again, we see the earlier
Bhasa of Panini as more of an attempt to create a local sacred
language by adopting certain phonological and morphological
traits of the older religious languages, which did not exist
in the local speech. Eventually this mostly failed in
morphological and grammatical terms, as writers began to write
Bhasa in a manner similar to the way in which the Prakrits were