It is universally recognised, by Tolkien himself to begin with (through
Lowdham's pen), that the triconsonantal structure of Ad�naic bases was
inspired by Semitic languages. The vowel system and its use in
morphology is however said to be different: "The vocalic arrangements
within the base, however, do not much resemble Semitic; neither does
Adunaic show anything strictly comparable to the 'gradation' of languages
familiar to us, such as the _e/o_ variation in the Indo-European group.
In an Adunaic Base there is a Characteristic Vowel (CV) which shares with
the consonants in characterizing or identifying the Base. (...) The CV may,
however, be modified in certain recognized ways (described below under
the Vowels) which can produce effects not unlike those of gradation"
Browsing through a book about ancient languages, I recently came
across a brief sketch of Sanskrit structure, and was struck by
similarities with Ad�naic in the matter of vowel gradation (== apophony,
ablaut). I do not remember that Sanskrit had been mentioned yet as
a possible inspiration for Ad�naic in a developed way.
The usual Sanskrit transcription makes use of diacritics not supported
by ASCII, which may be problematic in emails, so I will use instead the
Harvard-Kyoto Convention; a description is available on Wikipedia
>. I am not a Sanskritist
at all; if there are errors, I will be happy to be corrected.
Ad�naic as described in Lowdham's report makes a threefold vowel
length distinction: short, long and overlong. In the source, macrons
are used for long vowels and circumflexes for overlong ones, while
short vowels are unmarked. Since macrons too may cause problems
in emails, I will simply double the vowel when long and reserve the
circumflex for overlong vowels.
1) VOWEL GRADATION
The Ad�naic vowel gradations are of two kinds: the CV can be
lengthened, or modified by a-infixion or "fortification". (Nasal
infixion is also possible). The apophonic series are presented so
Basic A: a - aa - �
Basic I: i - ii - �; ee - �i
Basic U: u - uu - �; oo - �u
Sanskrit morphology makes much use of vowel gradation, which
it inherited and modified from Proto-Indo-European (PIE). Three
grades are recognised; Western terminology calls them reduced,
full and long. Traditional Sanskrit grammar sees the reduced grade
as basic and describes the other grades as successive addition of
_a_: a first addition gives the guNa grade, a second addition gives
the vRddhi grade. This gives the following alternations
(reduced - full - long):
Basic nil: nil - a - A
Basic i: i - e - ai (before a consonant); y - ay - Ay (before a vowel)
Basic u: u - o - au (before a consonant); v - av - Av (before a vowel)
Basic R: R - ar - Ar (before a consonant); r - ar - Ar (before a vowel)
Here follow a few examples (the root with its apophonic vowel is
surrounded by asterisks):
pa*pt*uH "they flew"
*vid*maH "we know"
*rud*ita- "wept" (past participle)
va*vRt*uH "they turned"
*pat*ati "he, she, it flies"
*ved*mi "I know"
*rod*itum "to weep"
*vart*ate "he, she, it turns"
a*raud*iSIt "he, she, it wept"
The successive additions of _a_ described by traditional Sanskrit
grammar are very like the fortification of Ad�naic. In both languages,
_e_ and _o_, which exist in long varieties only, pattern as the
diphthongs they earlier were -- this is very apparent in the contrast
between preconsonantal and prevocalic treatment in Sanskrit
(Sanskrit _e, o, ai, au_ < Indo-Iranian _ai, au, �i, �u_).
It can be noted that if the vowel gradations of Ad�naic resemble
Sanskrit (or better, Indo-Iranian), they are quite different to the PIE
patterns, explaining Lowdham's words that "neither does Adunaic
show anything strictly comparable to the 'gradation' of languages
familiar to us, such as the _e/o_ variation in the Indo-European group".
Beside variations in vowel quantity (quantitative ablaut), PIE extensively
used variation between the qualities _e_ and _o_ (qualitative ablaut)
-- the typical alternations were between e - o - nil (_e_ and _o_ being
short). Long grades in _�_ and _�_ existed but their use was quite
restricted. In the Indo-Iranian branch branch of Indo-European, this
system was profoundly altered because _e_ and _o_ (short or long)
merged in _a_ (short or long); qualitative ablaut became prominent
and was much expanded (notably by a change of _o_ > _�_ in certain
It is also interesting that nasal infixion is reconstructed as a
morphological feature of PIE; there is a class of verbs with nasal
infixes / suffixes in Sanskrit.
2) ELLIPTICAL DUAL
Another similarity between Ad�naic and Sanskrit is the use of the
elliptical dual: in a conventionally paired opposition of terms, one
in the dual can represent both elements. �douard Kloczko already
pointed out this similarity (pg. 104 of his _Dictionnaire des langues
des hobbits, des nains, des orques_. Argenteuil [France]: Arda, 2002.
179 p. ISBN 2-911979-04-4). Lowdham provides us with the Ad�naic
example _uuriyat_ "the two suns" == Sun and Moon (IX:428). Sanskrit
dyAvA "the two heavens" == heaven and earth
mAtarA "the two mothers" == mother and father
varuNA "the two VaruNa's" == Mitra and VaruNa
For such pairs, Vedic Sanskrit also used copulative compounds with
both members in the dual and accented (a kind of compound known
as dual dvandva); beside _varuNA_ it had _MitrA-varuNA_ for "Mitra
and VaruNa". We know that Ad�naic for oppositive pairs used
compounds too as well as elliptical dual (alone or completed with
the second term in the singular): for "sun and moon" it had beside
_uuriyat_ the forms _uuriniil(uw)at_ and _uuriyat niiloo_(IX:428).
However, this is not fully like a dual dvandva, for we would then
expect _*uuriyat-niil(uw)at_ .
3) TENSES IN THE VERBAL SYSTEM
We do not know much about Ad�naic verbs, but know that it had
four tenses; aorist, continuative present, continuative past, past.
Sanskrit had a very complicated verbal system, with the following
present, imperfect (built on a present stem),
aorist (built on an aorist stem; it was past in meaning),
perfect, and in Vedic a residual pluperfect (built on a perfect stem).
The semantic distinctions between imperfect, aorist and perfect
are said to have blurred with time, but this is a tense system is
quite reminiscent of Ancient Greek, and in Neogrammarian times
it was thought to reflect the PIE verbal system (as far as I know this
idea is met with much reservation today).
There is some parallel with Ad�naic:
-- The two Ad�naic continuative tenses would correspond to the
present and imperfect built on a present stem. In Ancient Greek the
present stem had an imperfective meaning (for Sanskrit it is not clear
from my limited reading -- hints?)
-- The Ad�naic past is also used in association with other tenses as
a "relative" tense shifting even further in past than the associated
tense. Could it be considered as a kind of perfect?
-- Finally, the existence of an aorist. An interesting terminology,
even if the values are not quite the same, since the Ad�naic aorist is
said to be present as well as past -- it is somewhat reminiscent of the
rather present-oriented (or timeless) Quenya aorist.
I do not know if Tolkien really studied Sanskrit at length, but he
cannot have ignored it because of its great importance in the
comparative grammar of Indo-European languages, even more so
in the time of the Neogrammarians than now. Evidently it left some
traces in Ad�naic. For nasal infixion and the tense system I think
it is possible that Tolkien, more generally, was influenced by the
reconstruction of PIE current at his time; but the vowel gradation
patterns and the elliptic dual are likely to be more specifically
Haudry, Jean. _L'indo-europ�en_. 3rd revised ed. Paris: Presses
Universitaires de France, 1994. 127 p. (Que sais-je ?; 1798). ISBN
Varenne, Jean. _Grammaire du sanskrit_. 2nd revised ed. Paris: Presses
Universitaires de France, 1979. 127 p. (Que sais-je ?; 1416). ISBN
Woodard, Roger D. (ed). _The Cambridge Encyclopedia of the World's
Ancient Languages_. Cambridge: Cambridge University Press, 2004.
1162 p. ISBN 0-521-56256-2.