The Austric files as a .zip fileThe Austric files as a .txt file
updated: 30 Jul 09
Stephen Oppenheimer in his book "Eden in the East" proposes that
the sudden rise in sea level appr. 6000 BCE flooded Sundaland, the
area which is now the sea around Indonesia, and forced people
living there to leave, in the process travelling to many places,
including the Mediterranean, where they became "civilizing
heroes". It is interesting that the initial spread and dispersal
of AfroAsiatic and of Old European (of which Etruscan is a member?)
is generally set at about the same time. At the same time, Proto-
IndoEuropean reached Central Asia from the Fertile Crescent.
If this is true, did these exiles from Sundaland (presumably
speaking an Austronesian or at least Austric language) leave
linguistic traces in the IndoEuropean and AfroAsatic languages?
It would seem they did.
Paul Manansala has published a list on the internet with comparisons
between Sanskrit (and its descendants) and Austric languages: PMA and
between Sumerian and Austric languages: PMS.
I think (but I am not sure) that Manansala proposes that the
Indic/Austric cognates he has listed belong to a substrate of
Indic (or is Indic?). The problem is that the Indic half of the
cognate pairs have well-established cognates of their own in
the other Indo-European language groups. Therefore the existence
of the cognate pairs that Paul Manansala has found does not disprove
Indo-European influence in India (which he doesn't claim, anyway),
rather it would tend to reinforce an out-of-India hypothesis for
Indo-European. Some of the cognate pairs have cognates themselves
in other language groups (Afro-Asiatic, Kartvelian, Sumerian). Which
means, I think, if the Austronesian cognates are outside the Indic
influence area, that there have been earlier (and extensive! judging
from the number of cognates) contacts between Proto-Austric (or
perhaps Proto-Austronesian with Austro-Asiatic as sister languages)
and the block of languages now being proposed as descendants of
Nostratic. In principle the contacts could of course be both ways,
but a post-diluvian dispersal of a boat people from east to west
would be intriguing. We are left with two options for action
1) Assume that it is all a part of the Proto-World language. Add
Austric to the stock of candidate languages as a descendant
2) Assume a heavy borrowing of cultural items in the areas of
astronomy, boats, kingship, measuring, mythology, basic
geometrical structures couched in zoomorphic terms (this
is supposed to mean that when they talk of snakes and trees,
they also mean lines, waves and circles), psychology, magic,
shamanism, etc from East to West. Go chase Proto-Austronesian
(the Austro-Asiatics are landlubbers) cognates in every
language group that has a coastline (well, the speakers of
which live in an area that has etc, you know what I mean).
I go for 2). If you don't, check for Scandinavian bronze age rock
carvings on the net and be convinced.
This has some consequences for my modus operandi.
For each of Paul Manansala's cognates, I have tried to find
IndoEuropean cognates (from EIEC, Encyclopedia of Indo-European Culture)
Nostratic cognates (from IENH, Indo-European and the Nostratic Hypothesis)
and I have hoped each time (especially with Nostratic) that
the cognates I found would belong to a language spoken in an
area with a coastline (not too many Proto-Altaic roots, thank
you very much). I seem to have been lucky. Most of the roots I
found have belonged to places not incompatible with a theory
of sea-borne transmission.
I have supplemented with discussions of root stems and related
derivatives from the works of Møller and Cuny. Although they
are very aged, I have nowhere found a similarly detailed
exposition of the subject. Present-day authors seem to be
content with working with single words (not fair, but sort of
true). Especially around the complexes *H-n-(g) "bend, snake,
fear, kill, sew, straight line", etc and *H-r-(g) "king, erect,
set straight, extend" etc they have been invaluable.
Dumezil is good for setting things straight, when the sense of
words makes no sense.
And some Etruscan cognates thrown in for good
Charles Fillmore (in CFC) has proposed that in all languages
the various NP's (noun phrases) in a sentence each take on one of
a fixed number of roles relative to the verb. He called those roles
deep cases (as opposed to what he termed surface cases of e.g
Indo-European languages like German, Russian and Latin).
The number of deep cases varied as the theory evolved. Here is a
typical list (I cite from memory, accuracy WRT names is not
These are the "non-spatial" cases. The "spatial" (or "spatio-
temporal") may include:
(place where action occurred)
In standard Proto-IndoEuropean, such as it is reconstructed, a
sentence consists of a verb and one or several noun phrases.
The roles (deep cases) of the noun phrases are indicated by
adding suffixes to the component words (articles, adjective,
nouns) of the noun phrase in question. In Proto-Austronesian
(according to PA) a sentence consists of a verb and just one
noun phrase. The role of that noun phrase is indicated by an
affix (pre-, in-, suf-) not of the noun phrase but of the verb.
Here they are:
Actor mood : *-um- (infix )
Object mood : *-ën (suffix)
Beneficiary-Locative mood : *-an (suffix)
Instrumental mood : *Si- (prefix)
There is also a
Perfective mood : -in- (infix)
to which one might compare:
For actor mood and perfective mood
The Proto-IndoEuropean present tense -n- infix (CAIEH 76).
For object and beneficiary-locative mood
The Germanic and Slavic past participle in -n- (CAIEH 77)
(corresponding to Sanskrit -ána). If the language has no
copula (e.g. Russian), a passive sentence will correspond
to a Proto-Austronesian one. Perhaps the Latin absolute
ablative construction is an archaic sentence type that
belongs here too. Compare also the Etruscan suffix *-na
(shuthi "grave", shuthina "grave goods" written on
Alternative: the Germanic infinitive suffix (-an, later
variously -en, -e, nothing), Sanskrit ´-ana.
For instrumental mood
The AfroAsiatic languages makes causative verbs by mens of an
s-preformative (prefix). Herman Møller has suggested that the
Proto-IndoEuropean "s mobile" is a cognate and therefore
originally also was a causative preformative. It is tempting
to compare it to the Proto-Austronesian instrumental mood
prefix *Si- (cf. English "melt" vs. "smelt" (make melt?), and
"reach" vs. "stretch" (make reach?), cf. German "recken" and
"strecken", Germanic inserts -t- into sr-, as in "stream"
from *(s)rw-m-, “flow”).
Beside the s-preformative for Proto-IndoEuropean and Proto-
AfroAsiatic, Møller also postulates the existence of a
w-preformative of no distinct semantics.
Since I assume an Austronesian origin for the cognate "clusters"
I am free to assume whatever alternations that are common there,
so r/n/l/d will be permissible alternates, and I have permitted
n-infixing into tri-consonantal roots C-C-C < > C-NC-C. This
is partly accepted in Indo-European already as present (ie.
imperfective) stem infixes in verbs. Dyen notes a similar
function in Proto-Austronesian: verb with n-infix, noun without.
Also (NEOVF), in Afro-Asiatic, Cushitic Bedawi has an imperfective
-n- prefix (two-consonant stems) and infix (three-consonant
stems), eg. verb -ktum- “to come”, 1st sing. perfective a-ktim,
Further I will not attempt to keep laryngeals apart. This means
a lot of freedom and a good deal less credibility, but then,
this is only a first approximation. It should be noted that
in Semitic, the choice between r/n/l is considered "a matter
of style" (I read somewhere) and that it is generally assumed
that some substrate languages of the Mediterranean had d/l/n/r
vacillation. Consider also old Indo European problems like
(Hittite) laman vs. (many other Indo-European languages)
n-m-, Latin lingva vs. Germanic etc dengwa, and the archaic
neuter -r/-n and -l/-n paradigms. Thus it is possible to
claim a common origin for the *H-n(-g)- "snake" and *H-r(-g)- "ruler" roots.
In general, I think this d/l/n/r vacillation has been ignored.
Verb-forming nasal infix before second consonant of root
tuboh “body” Melanesian
tumboh “grow” Melanesian
yukta-h “joined” Sanskrit
yuñj-mah “we join” Sanskrit
The n-infix of the present may become the basis of another noun which may
replace the original one.
tumboh “growth” Melanesian
jùngas “yoke” Lithuanian
or an n-infix-less verb may be formed from the original noun.
tuboq “grow” Tagalog
tubu “grow” Toba
tuwoh “grow” Javanese
and if both happens, you may have a doublet root, with
or without semantic distinction.
Suffix (verb)-en meaning “something which gets (verb)ed”
-ka:qin “eat” Tagalog
ka:n-in “boiled rice” Tagalog
makan “eat” Mal.
makan-an “food” Mal.
ka-i “eat” Futuna
kan-o “meat” Futuna
bit- “bite” Gothic
bit-an-s “bitten” Gothic
nes- “carry” Old Church Slavonic
nes-en-û “carried” Old Church Slavonic
Austronesian durative reduplication.
nagpu:tol “did cut” Tagalog
nagpu:pu:tol “is, was cutting” Tagalog
mööt “sit” Trukese
mömmööt “be sitting” Trukese
títhe:mi “I put” Greek
quotes E. Benveniste's "Origins of the Formulation of
Nouns in IndoEuropean" (1935) as the origin of the ideas
that lead IENH to these observations on the root structure
1 There were no initial vowels in the earliest form of
Pre-IndoEuropean. Therefore, every root began with a
2 Originally, there were no initial consonant clusters
either. Consequently, every root began with one and
only one consonant.
3 Two basic syllable types existed (A) *CV and (B)
*CVC where C = any non-syllabic and V = any vowel.
Permissible root forms coincided exactly with these
two syllable types.
4 A verbal stem could either be identical with a root
or it could consist of a root plus a single derivational
suffix added as a suffix to the root: *CVC-VC-. Any
consonant could serve as a suffix.
5 Nominal stems, on the other hand, could be extended
by additional suffixes.
Similarly for Afro-Asiatic
1 There were no initial vowels in the earliest form of
Proto-AfroAsiatic. Therefore every root began with a
2 Originally, there were no initial consonant clusters,
either. Consequently, every root began with one and
only one consonant.
3 Two basic syllable types existed (A) *CV and (B) *CVC
where C = any non-syllable and V = any vowel. Permissible
root forms coincided with these two syllable types.
4 A verb stem could either be identical with a root or it
could consist of a root plus a single derivational
morpheme added as a suffix to the root: *CVC-VC-. Any
consonant could serve as a suffix.
5 Primary (that is, non-derivational) noun stems displayed
similar patterning, though, unlike verbs stems, they
were originally characterized by stable vocalism.
PA quotes Dempwolff for the folowing limitations on the
2 CVCVC, of which some are
CVNPVC, where N is a nasal and P is a stop, homorganic
with each other ie -mp-, -mb-, -nt-, -nd-, -n(g)g-,
-n(g)k-. CVNPVC and CVPVC may be or not be semantically
equivalent, in which first case they are written
3 CVCCVC, where CVC = CVC,
ie two identical syllables. abbr. CVC2
4 CVCVCVC, trisyllabic roots, of which some are
CVNPVCVC, CVCVNPVC, CV-CVC2
5 CVCVCVCVC, very few
Many have the form cv-, therefore the result becomes
one of the above, eg. cv- + CVC > cvCVC, cv- +
CVCVC > cvCVCVC
All infixes have the form -vc-, and are inserted
after the first C. eg. -vc- + CVC > CvcVC, -vc- +
CVCVC > CvcVCVC
Many are in the form -vc, eg. -vc + CVC > CVCvc,
-vc + CVCVC > CVCVCvc
This looks very similar to the structure of Proto-
IndoEuropean + nasal infix as explained above by Dyen.
VISW claims the following of the forms of
1 Pronominal single consonant stems C-
2 Two-consonant stems C-C-
3 Reduplication of two-consonant stems C-CC-C ((C-C)2 ?),
eg. gw-l “roll, turn” >
gw-lgw-l, Hebrew gilga:l “wheel, circle”
and as simplification "simple reduplication",
a consonant is dropped.
IndoEuropean drops the second consonant
kw-kw-l- , kweklo-s “wheel”
Afro-Asiatic drops the third consonant
gw-l-l “roll”, Amharic gw-l-l
4 Addition as suffix of the consonant in anlaut
eg. gw-r “swallow” > gw-r-gw, Old High German querca
“throat”. Called "half reduplication", elsewhere "broken
5 Two-consonant roots might be extended by a one-consonant
suffix, a "determinative". In some cases the meaning of
the determintive can be seen from its later history in
Indo-European, eg. -d and -n from -t and -n in Indo-European,
where the form participles. Since most of the determinatives
are no longer productive in Indo-European, they must be pre-
Indo-European. In Afro-Asiatic the determinative is tightly
bound to the root, since native Sprachgefühl demands three-
6 Two-consonant roots may become three-consonant by a one-
consonant infix. A nasal infix, which is rare in
AfroAsiatic eg. Ethiopian kanfar “lip” from *k-p-, has
become very much used in Indo-European. Further, in
Afro-Asiatic two-consonant roots may also be extended
by a one-consonant prefix. These prefixes may also be
found as infixes with no or small semantic change.
plus 8 more paragraphs.
As can be seen, the limitations of VISW correspond pretty
well with those of IENH and PA. In general, I think it sad
to see how Møller in generally is written off as a
"philologist" or "semiticist" who undertook an "abortive
attempt" to find a common origin for Indo-European and
Afro-Asiatic, by people who obviously haven't read his
As for the e - o ablaut gradation IVSQA maintains
that this should reinterpreted as a ë - a gradation.
So did RVCFRN.
(Conclusion by the author, I. Dyen)
I must confess that I am impressed with the extent to which I
have been successful in gathering matchings between the
reconstructions of the two families. Granting that thus far I
have not wavered in my stedfast belief in the likelihood that an
Austronesian-IndoEuropean relationship can not be demonstrated,
should I waver now?
I do. It should be noted that the article was written not to prove such
a relationship, but, rather, given the impossibility thereof since no
connections between the corresponding cultures before appr. AD 1000 is
(was) known, to test the validity of the comparative method (by providing
a counter-example, in Popper's sense). Dyen (personal communication), however,
doesn't see it this way.
The consonant cluster “ng” has been noted, by Chaterjee
and others, in word like "Ganga", which is related to words
meaning simply “river”, in many Austro-Asiatic languages.
Other words like sanga “having limbs”, Sanskrit seem to indicate
that intervocalic "ng" may be connected with words associated
with appendages, limbs, extensions, offshoots and the like. The
Austronesian words darnga, tarnga, kalinga, etc. all meaning “ear”,
certainly seem to be related to Sanskrit karna “ear”, the latter
having lost the “ng” consonant cluster. This has also occurred
in Austronesian languages were we find kalna, kalina, talina, etc.
Other possible examples of this trait are:
This *-ng- “appendage” infix wouldn't happen to that horrible serpent
*-n-g- again? Or am I getting paranoid? At least I would claim "anguli"
“finger” and "anga" “limb” for it.
And how about Indo-European *dnghuH2- “tongue”?
ang-, esp. to designate joints of limbs
lithus “joint” Gothic
*lei- “bend” Gothic
ángam “joint” Sanskrit
angúri-h “finger, toe”, Sanskrit (from which
anguli:yam “finger ring” Sanskrit)
anguisthi-h “big toe, thumb” =
angushta “toe” Avestan
anginn “corner” Armenian
añjali-h “both hands rounded
held together” Armenian
punga “heap, mound”
sunga “sheath of a bud”
matanga “elephant (from trunk)”
vangsa- “bamboo and other cane”
Initial retroflex consonants
There can be little doubt that the common occurrence of retroflex
consonants in the various languages of India are due to common
influence upon each other. Many of the retroflex consonants in Indic
and the Austric languages of India are almost certainly of Dravidian
origin as they are abundantly attested to in those languages but not
found outside India. On the other hand, the initial retroflex
consonants of Indic, cannot be found, or at least are very rare in
the Dravidian languages. Initial retroflex
consonants are quite common in Austric languages, both Austro-Asiatic
and Austronesian. Initial retroflex consonants do not occur in
IE languages in Iran or Afghanistan or further to the north and
west. Retroflex consonants themselves can be found as far east
as Formosa, Papua, Micronesia and the Santa Cruz Islands.
Re: the retroflex consonants in Swedish and Norwegian which are
sometimes cited as an example of "spontaneous" "retroflexisation"
(without super- or substrate influence); the "thick" or cacuminal
l occurs in the dialects of those areas where bronze age rock carvings
are found with motifs that might be interpreted as connected with
South East Asia (tree of life, birds, ceremonial axes, boats).
The retroflex r (which causes the immediately following consonants
(n, d, l, s) to become retroflex too) is limited to Sweden and Norway,
in Sweden historically to north of Scania into which area it is now
spreading, according to (indignant or jubilant) letters to the editor
in the local papers.
There exist an interesting similarity between the Vedic concept
of the sun as the “eye of Varuna”, and the Hindu one of the orb
as Lokachakshu, or “eye of the world”, and also as one of the
eyes of Siva; and the very common Austronesian term “eye of
the day”. In Thailand, we see the phrase as 'sa ven' “eye of
the day”, Li, 'da-ben' “sun”, Jer., 'tau wan', Dioi, Sek.
In Indonesia, and Malaysia there is 'mata wari' and
'mata hari' respectively for “sun = eye of the day”. In Malagasy
there is maso andro “sun = eye of the day”. In Amboyna and
Ceram, we have a slight variation with riamata “shining eye = sun”.
The words 'mata-alo' of Celebes and 'mata-alon' of Baju probably
have the same meaning. In Makatea, we have 'mata-ra', “eye of
the day = sun”, from a word for “day”, that is the same as
the word for sun in other New Hebrides languages.
Moving into Oceania, there is 'mata ni siga' of Fiji
and San Cristoval ahve the same meaning of “shining eye”,
for the sun. In Espirito Santo and Duke of York we have
simply 'maso' “eye = sun”, and 'make' “eye = sun”, respectively.
In Api, 'mat ni ele' “sun”, is quite the same as the 'mata-alo'
and 'mata-alon' mentioned above. At Leper's Island, 'matan aho',
related the newly-risen sun to the eye. In the New Hebrides,
the Western Eromangan has 'nipmi-nen' “eye of the day”.
Probably connected to this are the Sanskrit 'dina' “day”,
and 'dina-kara' “sun”. In Malay, there is the phrase 'dina hari'
“day break”, while in Motu, Mekeo, Kuni and Doura we have 'dina'
“day”. In Arosi, Sau, Saa, Kwaio and Ulawa there is 'dani',
'dangi', 'danigi', 'dinga', etc., all meaning “day”. In Motu and
Proto-Central Papuan 'dina' means “sun”, while in Indonesia
'tinag' means “torch”.
We can find many other examples of common terms and
phrases. For exmaple, in India it is common to denote
beautiful eyes by the term 'mina-kshi' “fish-eye”. The
eyes of the fish are also considered beautiful among
Austronesians, and thus we have examples like 'kole
maka onaona' “sweet-eyed kole (a fish) = beautiful
person”, due to the fact that the eye of the kole
was considered beautiful.
It is obvious (to me, at least) from Stephen Oppenheimer's
book "Eden in the East", that there was some kind of contact
from Sundaland (the land around the present Indonesia that
disappeared following the rise in water level at the end of
the last ice age) to India, Mesopotamia and Europe
(specifically Scandinavia). According to a recent (May, 2000)
article in Genetics some pig races in Europe (i.e. Yorkshire
and Swedish Countryside) contains Asiatic gene material.
Note the appearance of "hog" in the above list of
You may have noticed some enormous clusters in the cognate
*A-d-m “master, lord” | “build, something built” |
“domesticate” (driving out chaos)
*H-r-(g-) “king” | “reach, stretch” | “build with wood,
pile up, mount onto”
*H-n-(g-) “snake” > “constriction, suffocation, murder”,
“fear” | “snake” > “bend, wave”, “sew”
*m-n- “(noble) man” | “sexual desire, magical power” |
“spiritual body” |
“something protruding, stick, churning”
*t-n- “physical extension, physical body
(as opposed to spiritual, magical body)”
Words for swimming, travelling, boats.
The opposites heavy/light, appearing as if originating as
predicates of cargo on some floating transportation.
As for the four first clusters, here is what I think "they"
thought about it.
They knew about the Kundalini serpent, deep in the mind.
They saw the terrible conflagration in the depths of the oceans
at the Flood (whichever) (volcanic activity?, meteoric impact?)
as that same serpent rising and causing havoc in the physical world.
The *H-r-g- is not a surveyor, as Dumezil would have it. He is the
captain of a ship, also responsible for its construction (in wood),
and knowing about geometry because he sets the course by the stars.
With respect to the clusters stemming from *H-r-g- and *H-n-g-, they
may express the two opposite concepts of the straight vs. the crooked.
The chaotic vs. the orderly.
Or there may be a three-way opposition *A-d-m-, *H-r-g-, *H-n-g-;
creation, orderder, destruction.
The following is taken from memory, and I might be wrong:
One of the pioneers in the mathematical field of topology was the
German mathematician and linguist Grassmann. I believe that what
basis vectors do to a topological space in German is "aus-dehnen",
a cognate of *t-n- "extend, stretch". Coincidence (or bad memory
on my part)? In comparative linguistics, he is known for
"Grassmann's law" in Sanskrit, so he must have had done some
reading in that language.
Torsten Pedersen's BibliographyPaul Manansala's BibliographyPMA
dektu-worgo- “net-makers” Mycenian Greek
díktuon “hunting/fishing net” Greek (-i- from dikein “throw”)
e:kt- “net” Hittite
aggati- “catch-net” Luvian
áksu- “net” Sanskrit
The Greek forms represent neuter nouns with
(prefixed?) *d- as in the word for “tear”