Martínez de Vaan (2014) Introduction To Aves Tan

Introduction to Avestan
Brill Introductions to
Indo-European Languages
Series Editors
Michiel de Vaan
Alexander Lubotsky
volume 1
The titles published in this series are listed at brill.com/iiel
Manuscript J2, 478v, Yasna 43.6 (end) Yasna 43.8 (beginning)
Introduction to Avestan
By
Javier Martnez
Michiel de Vaan
Translated by
Ryan Sandell
LEIDEN | BOSTON
Library of Congress Cataloging-in-Publication Data
Vaan, Michiel Arnoud Cor de, 1973-
[Introduccin al avstico. English]
Introduction to Avestan / By Michiel de Vaan, Javier Martinez ; Translated by Ryan Sandell.
pages cm. (Brill introductions to Indo-European languages ; Volume 1)
Published in Spanish by Madrid : Clsicas, 2001 as " Introduccin al avstico".
Includes bibliographical references and index.
ISBN 978-90-04-25809-9 (pbk. : alk. paper) ISBN 978-90-04-25777-1 (e-book : alk. paper) 1. Avestan
languageGrammar. I. Martnez Garca, Francisco Javier, 1965- II. Sandell, Ryan, (Translator) III. Title.
PK6103.V3613 2014
491'.525dc23
2013040365
This publication has been typeset in the multilingual Brill typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more
information, please see www.brill.com/brill-typeface.
issn 2214-5605
isbn 978 90 04 25809 9 (paperback)
isbn 978 90 04 25777 1 (e-book)
Copyright 2014 by Koninklijke Brill nv, Leiden, The Netherlands.
Koninklijke Brill nv incorporates the imprints Brill, Global Oriental, Hotei Publishing, idc Publishers and
Martinus Nijhoff Publishers.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910,
Danvers, ma 01923, usa. Fees are subject to change.
This book is printed on acid-free paper.
Contents
Translators Note ix
Preface xi
Symbols and Abbreviations xiii
1 Introduction 1
1 Avestan; the Iranian Languages 1
2 The Avestan Texts 2
3 Writing System: The Avestan Alphabet 4
4 The Transmission 5
2 Phonology 7
5 Alphabet and Phoneme Inventory 7
6 Historical Phonology 9
7 Vowels 10
8 Epenthetic Vowels 17
9 Anaptyctic Vowels 18
10 Diphthongs 19
11 Consonants 21
3 Morphology 39
12 Introduction 39
13 Nominal Inflection 39
14 Case Endings of the Singular 43
15 Case Endings of the Dual 44
16 Case Endings of the Plural 44
17 Inflectional Classes 46
18 Consonant Stems 46
19 Vowel Stems 53
20 The Adjective 64
21 Numerals 66
22 Pronouns 69
23 Prepositions and Preverbs 77
24 The Verb 78
25 Component Elements 78
26 Present Stems 80
viii contents
27 Aorist Stems 82
28 Perfect Stem 83
29 Moods 83
30 Personal Endings 83
31 The Augment 87
32 Paradigms 88
33 Non-Finite and Nominal Forms 96
4 Syntax 99
34 Syntax 99
35 Number 99
36 Case Syntax 100
37 The Use of Verbal Moods 101
38 Clausal Syntax 103
39 Negation 104
40 The System Changes 105
5 Texts 107
41 Introduction 107
Bibliography 119
Glossary 123
Word Index 131
Topical Index 159
Translators Note
Like any dutiful translator, I have at all times striven to obtain an accurate and
clear rendering of the original Spanish text into English, while still maintain-
ing as muchof the original phrasing andstyle as possible. Giventhe objective of
this book, tointroduce the facts concerning the historical andsynchronic gram-
mar of Avestan, I have prized clarity above all, and therefore have occasionally
added additional explanatory clauses where I felt that a merely accurate trans-
lation did not convey the intended point. Similarly, the availability of a more
extensive number and gender agreement systemin Spanish often necessitated
that I restore full nominal referents in many places, again for the sake of clarity.
For technical reasons, the files in which the original Spanish manuscript was
composed were not usable. To produce the translation, I therefore retyped the
entire book froma printed copy of the original 2001 Ediciones Clsicas edition,
and typeset it using , in order to take advantage of direct Unicode input.
First thanks go to the authors, Javier Martnez and Michiel de Vaan (who,
incidentally, taught the first course of Avestan in which I sat, now almost
five years ago), for having entrusted me with this project. Chiara Bozzone
kindly read various portions of my text, suggested better translations, and
offered moral support whenever necessary. I owe a considerable debt to Jesse
Lundquist, who read a draft of the manuscript through the portion on nomi-
nal morphology, and thereby saved me frominnumerable typographical errors.
Los Angeles, June 2013
RS
Preface
Around the second millennium bce, at the same time that Judaism was tak-
ing shape further to the west, another monotheistic religion, impelled by the
prophet Zarathustra, arose among the Iranian tribes that inhabited the area of
present-day Eastern Iran and Western Afghanistan. Zarathustra incorporated
the old Iranian deities into this newreligion, and he reorganized themwithin a
dualistic system characterized by the battle between Good and Evil. The Maz-
dayasnian creed attained dominance in the Achaemenid (559336bce) and
Sasanian (ca. 224651 ce) Persian Empire. Later, following the Muslim inva-
sion, the Mazdayasnian religion was nearly annihilated; nevertheless, in spite
of everything, Mazdayasnianism survived until the present day in a couple of
Iranian cities, in the west of India (Bombay [Mumbai], Gujarat), and through-
out the diaspora (United States, England).
Of the totality of texts belonging to the canon, which was formed in the first
half of the first milleniumbce and was successively transmitted by priests, only
a tiny fraction has survived, transmitted in manuscripts since the Middle Ages.
Withthe acquisitionof a goodportionof those manuscripts inthe 18thand19th
centuries, the academic study of the Avestan language and the Mazdayasnian
religion began in Europe.
Presently, grammars and monographs on Avestan exist in various languages,
but anupdated modernintroduction, mainly intendedfor students of Compar-
ative and Indo-European Linguistics was still lacking. In view of this gap, we
first published in 2001 an Introduccin al Avstico, which was quite successful,
and rapidly sold out. Limitations owing to the original language and the small
size of the first printing made the book a rarum.
The start of the new series Brill Introductions to Indo-European Languages
allows us to seize the opportunity to produce an English translation based on
the Introduccin, as many students of Avestan and Comparative Linguistics
requested since the Spanish book appeared.
Considering the growing number of scholars interested in the study of
Ancient Languages and Cultures, the present grammar has a dual objective.
In the first place, it aspires to be a clear and concise manual of Avestan for
those who wish to study the texts from a historical and cultural perspective,
but it also intends to bring out the history of one of the oldest and most archaic
Indo-European languages.
On the whole, the aim of this new edition was to change the text as lit-
tle as possible, because of both the positive feedback received after the first
edition and in order to maintain the concision and handiness of the original.
xii preface
Nevertheless, we have taken advantage of this occasion to correct typos and
other minor errors borne inthe first edition, andwe have also addedsome titles
to the bibliography.
Academic works are always subject to revision and the present book con-
stitutes no exception. After a period of further study, the historical grammar
of the languages indeed requires new explanations as well. Real progress has
been made in Avestan philology over the last decade (in particular, the results
of studies made by Jean Kellens and by Alberto Cantera and his collab-
orators), and the current text has consequently been updated. In matters of
historical phonology, M. de Vaan has, in a number of instances, modified our
previous views in agreement with the findings of his 2003 study on the Avestan
vowels and other subsequent papers.
We would, at this point, againlike to remember the teachers who introduced
the study of Avestan to us: Helmut Fischer and Ralf-Peter Ritter on the one
hand, and Robert Beekes, Alexander Lubotsky, and Jochem Schindler on
the other.
Other friends and scholars also deserve our gratitude for their advice con-
cerning this book. Carlos Jordn Clera (Zaragoza) revised and gave many
comments on the original Spanish version. Alberto Cantera (Salamanca) has
given graciously of his time and has likewise offered many helpful comments,
and Douglas Fear (Heidelberg) was able to untighten his schedule and read
the final manuscript. Lastly, we are deeply grateful to Ryan Sandell for will-
ingly entrusting himself to the translation of the Spanish text, and for being a
patient and solicitous assistant at every stage of this work.
A considerable number of the textual materials used for the revision and
updating of the present book, such as the frontispiece image of the Ms. J2, have
been extracted from the corpora belonging to the TITUS Project.
Oviedo / Leiden, June 2013
JM & MdV
Symbols and Abbreviations
hypothetically expected form

+
improved reading (of
Geldner)
conjecture not present in

mss.
# beginning or end of a word
before or after a shortened
word
corresponds to
. syllable boundary
< develops regularly from
<< develops indirectly from
> develops regularly to
>> develops indirectly to
replaced by
abl. ablative
acc. accusative
act. active
aor. aorist
Av. Avestan
ca. circa
ch. chapter
cs. centuries
caus. causative
dat. dative
esp. especially
f(em). feminine
FG full grade
fn. footnote
fut. future
gen. genitive
Goth. Gothic
Gr. Greek
HD hysterodynamic
Hitt. Hittite
i.e. id est
id. idem
IE Indo-European
IIr. Indo-Iranian
impv. imperative
impf. imperfect
ind. indicative
inf. infinitive
inj. injunctive
inst. instrumental
Ir. Iranian
Lat. Latin
LG lengthened grade
Lith. Lithuanian
loc. locative
m(asc). masculine
mid. middle
mss. manuscripts
N.B. nota bene
ne(ut). neuter
nom. nominative
OAv. Old Avestan
OCS Old Church Slavic
OHG Old High German
OIr. Old Irish
OP Old Persian
opt. optative
part. participle
pass. passive
PD proterodynamic
PE primary ending(s)
PIE Proto-Indo-European
perf. perfect
pl. plural
plupf. pluperfect
PN proper noun
posp. postposition
ppp. past passive participle
pres. present
xiv symbols and abbreviations
prev. preverb
rel. relative
resp. respectively
SE secondary ending(s)
sg. singular
Skt. Sanskrit
subj. subjunctive
superl. superlative
Umbr. Umbrian
v.l. varia lectio
voc. vocative
w.a. without attestation
YAv. Young Avestan
ZG zero grade
koninklijke brill nv, leiden, 2014 | doi: 10.1163/9789004258099_002
chapter 1
Introduction
1 Avestan; the Iranian Languages
Avestan is the language preserved in the sacred books of the Parsis,1 the ensem-
ble of which is called the Avesta. Avestan is an Indo-European language and
belongs tothe Indo-Iranianbranchof the family. Inturn, Avestan(Av.), together
with Old Persian (OP), is the oldest transmitted Iranian language.
Precision is necessary with respect to the term Iranian because of possible
confusion beween the linguistic and the geographical use of the term. The Ira-
nian languages are not geographically restricted to the borders of present-day
Iran, but are also found scatteredthroughout the whole area of the Middle East:
Turkey (Kurdish and Zaza [Zz]), Georgia and Russia (Ossetic), Azerbaijan
(Tt), Iraq (Kurdish), Iran (Persian, Kurdish, Balochi [Balo c ], Pashto [Pas to ]),
Afghanistan (Pashto, Ormur, Parachi [Parc ], Tajik [Tak], etc.), etc.
As has already been noted, the oldest attested languages of the Iranian
group are Old Persian and Avestan, of which two varieties are known: Old
Avestan (OAv.), also called Gathic Avestan or Avestan of the Gathas [Gs],
and Young Avestan (YAv.). The differences between both varieties are as much
chronological (diachronic) as dialectal (geographic).
There are other old Iranian languages of which we are aware, but unfor-
tunately, nothing more than scarce and badly preserved remnants have been
passed down. Such is the case for Scythian, of which we have information
through Greek writers; for Median, of which direct testimony does not exist;
etc.
In Figure 1, one can get a general view of the family tree of the Iranian lan-
guages. Note that the diagramdoes not faithfully reflect the historical reality of
the languages, and that there are still obscure points regarding the synchronic
position of some languages.
Neither where nor when Avestan was spoken is known with certainty,
though it is possible to surmise that its area of origin was Eastern Iran, while
it is thought that Zarathustra must have lived before the 10th c. bce.
1 Followers of the Mazdayasnian religion (Zoroastrianism).
2 chapter 1 introduction
figure 1 Iranian family tree (non-exhaustive listing of modern languages)
2 The Avestan Texts
The Avestan language has been transmitted to the present day by means of
manuscripts, the oldest of which dates from the 13th or 14th cs. ce (!).2 The
monumental edition of Geldner is preceded by a series of Prolegomena,
in which all of the manuscripts utilized for the edition are classified in a
detailedfashion, according tothe traditional methodof Classical Philology. The
Prolegomena furnish some valuable information, particularly due to the fact
that some of the manuscripts seen by Geldner have since been irremediably
lost, and the whereabouts of many others are unknown.3
The difficult work of reconstituting the Avestan texts rigorously combines
philology and linguistics, as the results that follow from the exegesis of the
2 This is K7a.b., which Barr dates to the year 1288 or 1268; cf. Geldner Prol. VIIa.; Barr 1944:
XIII f.; andHoffmann/Narten16. This manuscript is accompaniedby a Pahlavi translation,
i.e., inMiddle Persian. The manuscripts that containonly the text inAvestanare termedpure
(sa de), most of which are generally much younger than those called impure.
3 The Ms. Mf4 is an exceptional case, as it was not collated by Geldner. This important
manuscript has been edited in 1976 by JamaspAsa. As it happens, some of the mss. have
been rediscovered: cf., for example, F1.
2 the avestan texts 3
figure 2 The Iranian lands in antiquity
texts, together with textual criticism, must be compared with the results from
linguistics (historical and comparative). In virtue of this method, the Erlangen
circle, under the guidance of K. Hoffmann, has made considerable progress in
the study of Avestan philology during recent decades. One of the achievements
of Hoffmann consists precisely in positing an archetype from which all the
preserved Avestan texts ultimately derive. In effect, despite the many variants
that one may observe both in the texts and in their parts, all the manuscripts
retain a great uniformity, and must derive from an archetype put together
in early Islamic times (see 3). The restitution of the text that underlay the
archetype (cf. 4.2.5) is the principal task that Avestan philology undertakes.
It seems that selectedAvestantexts were organizedinto a canonina manner
similar to other liturgical texts such as the Bible or the Veda. Nevertheless,
even after the creation and diffusion of the Avestan script, as well as the
compilation of the Avestan canon, the transmission of the Avesta continued
to be primarily oral. Today, only a meager portion of the ancient canon is
preserved. Inthe Pahlavi (Middle Persian) literature, many texts are mentioned
that, unfortunately, have not been preserved. The archetype reconstructible
for the totality of the extant Avestan manuscripts probably reflects a tradition
made during the Sasanian period (ca. 224651 ce). This collection consisted
of descriptions of Zoroastrian rituals. They give the Avestan texts which are
to be pronounced during the ritual and some short stage instructions to the
officiating priests (given in various contemporary languages). The collection of
texts that was made for the archetype is found to different degrees in various
manuscripts. Its original use was for instructing future priests in the religious
schools, rather than actual use during religious ceremonies.
Within the Avestan corpus, the texts belonging to Old Av. are: (a) the 17
songs conceivably composed by Zarathustra himself (ordered into five gathas:
Y 2834, 4346, 4750, 51, 53), (b) the Yasna haptahiti (Y 35.241.6), which
could also be attributed to Zarathustra, and (c) several fragments dispersed
throughout the Yasna. The remaining texts of the corpus are catagorized as
Young Av.: for example, the Yasna (Y), Yat (Yt), the Vsperad (Vr), the Nyyis n
(Ny), the Gh (G), the Sro za (S), the A fri nga n (A), the Vi devda d (Vd), etc. They
are composed in prose with some small remnants of poetry. On the texts, cf.
41 ff.
3 Writing System: The Avestan Alphabet
Avestan is written with an alphabet created expressly for the purpose of com-
mitting the corpus to writing; the creation of this alphabet is probably to be
situated between the middle of the 7th c. and the middle of the 9th c; cf. below.
The Avestan alphabet is very complete and contains a detailed inventory of
graphemes (it may, perhaps, be the first phonetic alphabet), whose goal was
to represent with precision, in writing, a (liturgical) recited text, which would
have had variant forms depending upon the speed of recitation, etc. This fact
indicates that the Avestan alphabet is a deliberate creation, and does not result
from a lengthy process of adoption. Indeed, the alphabet must have had a spe-
cific creator, who could equally have been a lone individual or have come from
a school of recitation.
The immediate model for the Avestan alphabet is a variety of the Pahlavi
script (which itself ultimately derives from the Aramaic script): Book Pahlavi,
used by theologians of the Zoroastrian church for their writings. Its influence
is clearly seen in the letters a i k xv t p b n m r s z and s . Some
characters, however, have been taken over from another, more archaic variety,
used for a translation of the psalter, and known because they have been found
in a manuscript from Turfan (7/13th c. ce): cf. j and d . The rest of the
letters result either from the use of diacritics (for example, the small line was
abstracted from Pahl. L // > o , as a sign of length, and added to make the
corresponding longs to i and u, i.e. i and u , while being subtracted fromPahl.
L /o / to write o ) or from pure invention (a
as a ligature of + [ + ] or
in its two variants and ).
The establishment of a relative chronology for the script is a complicated
issue, and has been a recent topic of discussion. Until now, an inscription in
4 the transmission 5
Book Pahlavi script on a sarcophagus discovered in Istanbul, whose archeo-
logical dating demands a date no later than 430ce, was usually adduced as
important evidence. Some scholars even allowed the possibility that the cre-
ation of the script had taken place during the reign of S a buhr II (310379ce).
At present, the dating of the sarcophagus has been revised, and a much later
date is admitted, probably the 9th or 10th c. Recently, a new approach to the
evidence provided by the Pahlavi books and other sources, like coins, points
toward the existence of a Sasanian Avesta. The invention of the Avestan alpha-
bet could accordingly be dated to around 500ad. It is quite probable that the
pressure of the Arab conquest (651 ce) acted as a catalyzing agent in the Maz-
dayasnian community, and the need was felt to continue the Avestan canon in
a written form. The canons redaction in writing may have materialized during
the so-called Pahlavi Renaissance (9th c.), which attempted to set up a canon-
ical book indirect oppositiontothose that the other great religions offered, and
to the Quran in particular.
Despite its obvious filiation, the Avestan writing system distinguishes itself
sharply from the imprecision that characterizes the Pahlavi system, in which
the same signor ligature allows for various interpretive possibilities, andvowels
are not indicated(a practice inheritedfromAramaic). Avestan, fromthe outset,
assigned a specific value to each sign and marked the vowels precisely. In this
regard, the Greek writing system, which was well known throughout the East,
may have servedas a point of reference for the creators of the Avestanalphabet.
4 The Transmission
The enterprise that sought to commit the Avestan corpus to writing must have
been carried out shortly after the invention of the alphabet, and would have
culminated with the production of a sort of editio princeps of the Avesta, which
is usually given the name Sasanian Archetype. This unpreserved archetype
establishes the beginning of Avestans history of textual transmission. The
transmission of the corpus, however, obviously begins much earlier, with the
very moment of its composition. From here, it is possible to distinguish be-
tween several stages up until the time that the corpus took written form.
4.1. With regard to OAv., the stages are as follows (according to Hoff-
mann1989:51, with slight modification): 1. the original language of the gs of
Zarathustra, the Yasna haptaha iti, and the three sacred prayers (between 1000
and 900bce). The region where the Old Av. texts originated is usually located
in Northeastern Iran (Herat?); 2. changes due to slow recitation appear (here,
the numerous non-metrical anaptytic vowels are introduced); 3. changes due
to transmission in the hands of YAv. priests, who introduced phonetically YAv.
forms into OAv. (the so-called Young Avestanisms); 4. an intentional alter-
ation of the text through the orthoepic diaskeuasis, the aim of which was to
establish a canonical text.
4.2. Once the text of Old Av. was established, the transmission of the Old
andYoung Av. texts was carriedout incommon. The stages that affect the whole
of the Avesta are the following: 1. the original language of the Young Avestan
redactors; 2. the movement of the Avestan tradition around 500ce to Persis
[Fa rs] in Southwestern Iran; 3. the transmission of Avestan in a theological
school inSouthwesternIran(Es t akhr), which is reflected through the influence
of Old Persian and Median, in fanciful pronunciations that are presumably
the work of semi-erudite teachers, in the composition of late (grammatically
incorrect) Avestan texts, and in the incorporation of portions of texts that were
transmitted in other geographical areas.
4. With this stage, the purely oral transmission comes to an end. It is tradi-
tionally thought that, in the 4th c. ce, the alphabet, with phonetic notation for
Avestan, was created, and that the corpus obtained written form (the Sasanian
Archetype). This chronology is presently undergoing revision, andit seems that
this process took place rather in the 6th c. 5. Throughout the Sasanian period
(ca. 224651), the Avesta suffered serious deterioration because of incorrect
pronunciation (the period of the vulgate). 6. Starting from the first archetype
(11th c.?), hyparchetypes, which were not free of errors (cf. the regular inter-
change between s s
and s ), emerged. 7. Since 1288ce, the recent manuscripts

have been copied with a large number of errors and obvious corruptions; the
presently existing manuscripts reflect the outcome of this work.
chapter 2
Phonology
5 Alphabet and Phoneme Inventory
The Avestan alphabet consists of 16 vowel signs and 37 consonant signs. The
fact that a considerable number of the languages phonemes are representedby
multiple graphemes demonstrates that the writing system is effectively more
phonetic than phonological.
5.1. In Table 1, the letters of the Avestan alphabet are presented according
to their place of articulation, which is the traditional practice in Indo-Iranian
philology. The accompanying transliteration below each letter is that which is
usually employed at present. This transliterationscheme is based ona notation
established by K. Hoffmann (1971; cf. 1975: 316ff.), which the scientific com-
munity has accepted as the norm. Previously, different signs were used for the
transliteration of some characters, such as for x , c for c, j for j, w for , n for n
and n , for , s
, and s , y for y, , and ii, and v for v and uu. Recently, other char-
acters have been introduced into the modern transliteration system: , , v, ,
, m , , and s .
5.2. Certain letters are found only in the manuscripts and are usually not
present in the textual editions, a fact which should not take away from their
importance, because some of these letters may have been part of the original
alphabet of the archetype.
1. appears only in ms. Pd, where it is used instead of a preceding h.
2. , which scarcely occurs in the manuscripts, belongs to the original alpha-
bet. Inthe manuscripts that do use it, it most oftenappears inthe frequent final
sequence - n g (cf. esp. the mss. S1 and J3; 7.9.3), from which it is possible to
deduce that could have been an unreleased consonant like t (11.10.2); both
are the only final occlusives in Avestan.
3. v (-uh- [-h-] / -
u
h- < *-hu - < *-su -) likewise belonged to the archetype
and represents a labialized ; cf. 11.31. In the same fashion, represents a
palatal resulting from *-hi - < *-si -; cf. 11.29.
4. (cf. 11.4), in the archetype, represented a palatal n produced before *i ;
it is thus usually found in the manuscripts preceding ii (< *i ), but also before
i.
5. m was likewise a constituent letter of the original alphabet, used for the
representation of a voiceless m. Occasionally, one finds it replaced by hm.
8 chapter 2 phonology
table 1 The Avestan alphabet

a a

e e o i i u

k x x xv g

c j

t d t

p f b

v n n m m

y v r

s z s z s
h
6. The manuscripts that come from Iran usually employ instead of y, which is
used much more frequently in the Indian manuscripts. The substantial formal
difference between the two signs excludes the possibility that they might have
been mere variants; it may be that was originally employed for initial #i
(11.1.1). The two distinct graphemes could be explained by supposing that
original initial #i was written with , while y would have, in principle, served
to represent a palatal z. With the passage of time, the sounds z and z
were
no longer differentiated. Both sounds came to be written in inlaut with z,
consequently leaving the letter for z
free; this letter then must have been used

6 historical phonology 9
instead of by a certain Parsi community in India, whence the usage of y was
generalized in the Indian manuscripts.
5.3. The alphabet is written from right to left, and the letters are not
usually conjoined, but rather are written separately. Ligatures are not normally
employed, while those that do appear are sporadic and secondary by nature.
The most common ones are a, c, and t.
5.4. The Avestan writing system also makes use of an interpunct: the dot
(), which occurs at the end of each word or separates, in an inconsistent way,
the members of compounds, without observing any difference from the first
function: cf. r.vaca
who tells the truth (Y 31.12), fromthe adv. r right, cor-

rectly and vacah- word; vsp.voh who has all that is good; para.kauuistma
the best observer (Yt. 12.7) and its positive par.kauum (Yt 10.102), etc. In
some instances, the dot also separates the stem and the suffix/ending in a
word; cf. the superl. adj. spt.tma- the most beneficient (Y 37.3), the inst.pl.
drguu.db (Y 29.2), or the dat./abl.pl. drguu.dbii (Y 30.11), fromthe stem
drguuat- deceiver, follower of the Lie ( Skt. drhvan- deceptive, from the
root Av. druj, cf. PIE *dhreu gh deceive). Note that, in these cases, it is typical
to substitute the vowel a of the first element with , the typical composition
vowel, i.e., < drguua, etc.1
The sign , the function of which is to separate the Avestan text from the
accompanying interlinear translation, also appears inthe manuscripts (cf. ch. 1,
fn. 2). One can observe some examples of punctuation in the illustration on pg.
iv, extracted from ms. J2.
6 Historical Phonology
Avestan in its two varieties, when taken together with Old Persian, allows for
the reconstruction of a Proto-Iranian language (Ir.); the stage preceding Proto-
Iranian is Proto-Indo-Iranian (also called Proto-Aryan), which is reconstructed
through the comparison of Proto-Iranian and Sanskrit (especially the oldest
phase of the language: Vedic), which is a language closely related to Iranian.
1 The original composition vowel is a, which is found in some isolated cases: ustna-zasta-
with outspread hands, as a-cinah- seeking as a. In some instances, a is also found, justifiably,
to judge from parallel Skt. examples: kamna nar- with few men; vi spa .vohu having all that
is good (Skt. vivvasu-). o , of secondary origin, is the composition vowel of YAv., but it
was successfully introduced into OAv. An also occurs as a composition vowel in a few
rare instances: OAv. man.vista- found in thought, and, consequently, is found before some
endings as well; cf. 7.9.4.
Comparison with other Indo-European languages, such as Greek, Latin, and
Hittite ultimately allows for the reconstruction of Proto-Indo-European (PIE).
Likewise, the history of each one of the features of the language does not con-
sist solely of the description of the stages that the language goes through, but
also of their relative chronologies. Changes, the product of the internal history
of each language, are what determine the particular differences of each group,
and, at the same time, of the languages/dialects that make up that group.
For the purposes of deriving the Avestanphonological system, we proceedfrom
the following stage of Proto-Indo-European:
Vowels: e o i u
Consonants: p t k k
u
s h
1
h
2
h
3
i u l r m n
b d g g
u
bh dh gh h g
u h
In addition, the following combinatorial variants, or allophones, occur:
l r m n z
7 Vowels
The Indo-European vowel system was simplified in Proto-Indo-Iranian, a con-
dition that still holds in Proto-Iranian: the vowels e o (and their respective long
vowels) all become a (and a respectively), while the vowels i and u (and their
respective long vowels) remain unchanged. In general, the following develop-
ments can be stated:
table 2 The Avestan vowels
PIE *e *o > IIr. *a > Av. a PIE * * > IIr. * > Av.
PIE *i *u > IIr. *i *u > Av. i u PIE * * > IIr. * * > Av.
For Avestan, a vowel system very similar to that of Proto-Iranian is assumed.
In addition to these vowels, Avestan also has a vowel , which sometimes
functions as an anaptyctic vowel (9), though other vowels, such as , a, and o
are found filling the same role (cf. 9.2.5).
7 vowels 11
7.1. Comparison allows for the reconstruction of a vocalic system for Avestan,
which one would hope to find faithfully reflected in the texts; however, the
vowels in the text of the Avesta have undergone some additional changes.
For instance, it is not uncommon to encounter phenomena that result from
the peculiarities of liturgical pronunciation (slow or quick chanting), intro-
duced throughout the transmission. Some of these features may already have
been present in the original language (cf. 9).
7.2. Between Old and Young Avestan, one primary difference that emerges
is in the representation of final vowels. The vowels a, a , , , i, i , u, u , e, e,
o, o , are always written as long vowels in word-final position in Old Avestan,
while in Young Avestan, they are always written as short vowels, except for -
and -o , and in monosyllabic words (cf. 7.11.3, 7.13.1). Before the enclitics ca
and cit , in Old Avestan, i and u usually shorten, while a usually remains as
such.
The vowels found in the text of the Avesta have direct correlates in the
proto-language, or else are the products of specific phonetic developments. In
the following sections, we present a list of the vowels that appear in the text of
the Avesta, with the details of their respective histories.
7.3. a derives from IIr. *a < PIE *e, *o, *n , *m .
1. Sometimes, an a in the antepenultimate syllable is the product of shorten-
ing froma : OAv. caarasca four versus car id. (Skt. catva
ras); dtaras-ca
versus dtr (nom.pl. of dtar- creator; Skt. dta
ras); cf. also the ending of

the thematic gen.pl. -anm (versus Skt. nm, OP -nm; the final -m is bisyl-
labic). The same phenomenon is seen in YAv. (but not OAv.) in the abl. ending
-t preceding the preposition haca: ahmat haca.
2. The IIr. sequence *a i a sometimes becomes ai a in Avestan (just as in the
other Eastern Iranian dialects): YAv. asaiia- without shadow (cf. Skt. cha ya
-);
mazdaiiasna- Mazdayasnian, of Mazda ; raiia with wealth (inst.sg.; Skt. rya
;
but cf. gen.sg. OAv. ra iio ; Skt. rys). Shortening of prevocalic *i is quite
sporadic: YAv. vaiiu- wind (Skt. vy-).
3. The IIr. seqence *a u a occasionally develops to au a in Avestan (and in
Eastern Iranian): nauuza- navigator (Skt. nvj-); Av. as auuan- truthful,
following as a [truth] (Skt. r ta
van-).
7.4. a derives from IIr. *a < PIE *e, *o (and the corresponding short vowels
preceding laryngeals), *m H, *n H(YAv. za ta- born < PIE *n h
1
-t-; Skt. jt-, Lat.
gntus). Occasionally, is found instead of the expected a: 1. after a labial con-
sonant and before s (< *-rt-): YAv. vs m wagon < *varta-; 2. in initial syllables
(when followed by several other light syllables), one may find instead of
a: rmaiti- right-mindedness (Skt. armati-), kuuaiias-ca princes (Skt.
kavyas); 3. often, in OAv., following ii or uu (i.e., following Cii / Cuu ): vii-
darsm(1.sg.aor.act.ind. of dars see; cf. Skt. adaram); vrziitm(3.sg.pres.
mid.impv. of varz carry out, do; cf. Gr. -);
+
h miisait (3.sg.pres.mid.ind.
of yam hold); h-duurt (3.pl.pres.mid.inj. of duuar run); xv nuut
(inst.sg. of xvanuuat- sunny).
4. Finally, mention should be made of the development PIE *o > IIr. *
> Av. when the PIE vowel *o stood in an open syllable: cf. Gr. , Av.
pitarm, Skt. pitram, as opposed to Gr. , Av. dtrm, Skt. dta
ram. This
phenomenon, which is frequently subject to intraparadigmatic regularization
owing to the tendency to generalize but a single inflectional stem, is given the
name of Brugmanns Law or simply Brugmann; cf. 18.1.2, 18.5, 26.15, 32.1.c
(but only possibly in the case of 32.1.c).
7.5. a
reflects 1. an preceding h (< *-sa
-; cf. 5.2.3) or (i.e. NC; cf. 11.3.2);

OAv. ya
hm(gen.pl.f. of the rel.pron. ya-; cf. Skt. ya
sm); YAv. ma
hm(acc.sg.
of mh- moon; cf. Skt. ma
sam); da
h (2.sg.aor.mid.sbj. of d give / put <

PIE *deh
3
/ *dheh
1
, resp.); da
t (3.pl.aor.mid.sbj. of d); YAv. maza
tm (acc.sg.
of OAv. mazat- great; cf. Skt. maha
ntam); haca
t (3.pl.pres.mid.sbj. of hac
follow).
In inflection are found: 2. final -a
#, which derives from the IIr. sequence

*-s#:2 cf. OAv. ata ha
(gen.sg.f. of dem.pron. ata- this); OAv. dana
(gen.sg./
nom.acc.pl. of dan- religion; cf. the Skt. ending -s of gen.sg./nom.acc.pl. of
-stems.); 3. the ending of the gen.du. is also -a
< *-s (cf. Skt. -o; the loc.du.

has -uuo < *-au ; cf. 10.3.4).
7.6. is the product of 1. the IIr. sequence a
n before a fricative: x f s z :
mra- formulation (Skt. mntra-); YAv. zah- tightness (Skt. m has-; Lat.
angus-tus; < PIE *h
2
mhes-); dhita- most expert (Skt. dm siha-) and dids
(3.sg.pres.act.inj. of dh teach, Skt. dam s; cf. Gr. ); xnah- rein (cf. Gr.
-; from PIE *h
2
enk bend); YAv. fr forwards (Skt. pra
< *pra
k-);
OAv. psnu- dust (Skt. pm s-); vs (3.sg.aor.act.inj. < IIr. *u nst fromvan win,
prevail).
2 When the enclitic ca is added, the vowel a
is maintained by analogy: gen.sg. (mazd-) mazda
andmazda
sca; nom.pl. (dana-) dana
anddana
sca; acc.pl. (ssn-) ssna
andssna
sca, etc.
7 vowels 13
An becomes : 2. before a final nasal in all cases (-m -n): loc.sg. dm at
home; Av. m(acc.sg. 2.pers.pron.; Skt. tva
m); OAv. mm(acc.sg. 1.pers.pron.;

OP mm, Skt. ma
m); camm (loc.sg. of caman- eye); nmm, nmanm

(acc.pl. [cf. 11.3.3] and gen.pl. resp. of nman- name; cf. Skt. na
man-, Lat.
nmen); 3. in an open syllable before a nasal in some cases: YAv. nma
(nom./acc.neut.; cf. above; Skt. na
ma); YAv. dadmi (1.sg.pres.act.ind. of d give

/ put; Skt. dd
(h)
mi); but hunmi; uruun versus uruun (both nom.pl. of
uruuan- spirit). 4. On for , cf. 5.2.1. 5. On as an acc.pl. ending in YAv., cf.
7.9.3.
7.7. is found in some manuscripts instead of . Although it may initially
seem that we are dealing with two graphic variants for the same phoneme, it is
quite probable that, in the archetype, both signs corresponded to two distinct
phonemes: would reflect a long, nazalized (cf. nma or the ending -m),
while would reflect a short, nazalized .
7.8. represents *a 1. before a nasal: ht (3.pl.pres.act.ind. of ah be; Skt.
snti, Lat. sunt); vazti (3.pl.pres.act.ind. of vaz lead; cf. Skt. vhanti < PIE
*u eh); YAv. barn (3.pl.pres.act.inj. of bar carry; cf. OP abaran; PIE *bher);
ahurm (acc.sg. of ahura- lord; Skt. sura-); and 2. also preceding -uui- (i.e.,
-u i-): uuduua
ignorant (nom.sg.; Skt. vidvm s-); PIE *u ei d- > IIr. 1*u ai d-

find, 2*u ai d- know; suuita- (superl. of sra- poweful; Skt. viha-); tuu-
might (Skt. tvi-); kuuna- PNversus nom.sg. kauu (cf. Skt. kav- seer); YAv.
ruu- swift (< *rau -; cf. Skt. ragh--, Gr. ; PIE *h
1
leng
u h
-).
3. a is, however, almost always preserved if (a) ii (i ) or uu (u ) precedes it:3
xaiiamn (nom.sg. pres.part.mid. of x possess, be owner of; cf Skt. kyati);
YAv. auruuantm runner (acc.sg.; Skt. rvant-); YAv. bauuatm-ca (acc.sg.
pres.part of b; Skt. bh); a, though, is never preserved as such before a final
nasal -aN#, even when ii or uu precedes it: OAv. bnaiin (3.pl.pres.act.inj. of
ban be ill); ra
haiin (3.pl.pres.act.inj. of rah move away from); or (b) before

nasal followed by ii: OAv. kainibii, YAv. kainii (dat.pl. and acc.pl., resp., of
kainn- girl; Skt. kanya
; Gr. ?) OAv. spaniia
, YAv. spainiiahm (nom.sg.

and acc.sg., resp., of the comparative spainiiah- to spta- sacred); maiiu
spirit (nom.sg.; Skt. many-). Even though a before a nasal always became
, YAv. has sometimes restored original am and mostly restored original an

3 The pres.part.mid. saiianm lying, whose short vowel is unexpected, is not to be included
here, cf. Skt. yna-.
in word-internal position, whence it was introduced into OAv.: cf. Av. nmah-
homage (Skt. nmas-), but Av. manah- thought (Skt. mnas-). 4. On the
original development of a in YAv., cf. 7.14.4 (a > > i
; cf. also fn. 6 below),

7.16.1 (a > > u). 5. On as an anaptyctic vowel, cf. 9.2.5.
7.9. is found 1. often in OAv. passages as the outcome of a before a nasal
(but cf. 7.8): anii n (acc.sg.m. of aniia- other; cf. Skt. any-), vrz na- com-
munity, clan (Skt. vr jna-), xv nuuat- sunny (xvan- sun; cf. Skt. svr- id.),
hac na- fellowship (Skt. sacan-); hac mn (nom.sg.f. pres.part.mid. of hac;
Skt. scate); 2. in OAv., it reflects an a preceding the sequence hm: hm
(acc.pl. 1.pers.pron.; YAv. ahma); in YAv., a is preserved (or perhaps underwent
a reversion > a),4 and from there it was extended into OAv.: OAv. mahmi
(dat.sg.neut. of ma- my); ahmat (abl.pl. of az m I, Skt. asmt), etc.; 3. before
the cluster gh < IIr. *ns: m ghi (1.sg.aor.mid.sbj. of man < *mansi ); OAv.
v ghat , v ghait (2. and 3.sg.aor.act.sbj., resp., of van; cf. Skt. vm sat); s gha-
explanation (YAv. saha-; Skt. m sa-); but not before -h- < *-a
sa
-: man-
ah (inst.sg. of manah-; YAv. manaha; Skt. mnas); finally, cf. t g (acc.pl.m.
dem.pron.; Skt. ta
n); ams g (acc.pl.m of ams a- immortal); vsp g (acc.pl.

of vspa- all; Skt. vva-). The ending *-ans > OAv. - g / YAv. - : YAv. ams ,
vsp , pur (acc.pl.; the YAv. ending - results from the preservation of nasal-
ization when m, n, ii, or h preceded the vowel, while was the denasalized
outcome elsewhere: YAv. haom, asm, im, and from there was extended to
other cases: mazit am mazit am).
4. Likewise, , in OAv., develops from IIr. final *-as (Skt. -as > -a): tar
across, over (prev.; YAv. tar, Skt. tirs); cir (nom.sg.m. of cira- brilliant;
Skt. citr-); m (nom.sg. of ma-); par beyond (adv.; YAv. par, Skt. pars);
sar (abl.sg. of sar- union); haz (nom.sg.neut. of hazah- power, dominion;
Skt. shas-); v (gen.pl. encl. of 2.pers.pron.; Skt. vas); in YAv., the same final
sequence also resulted in - , but was replaced by -, which was in turn intro-
duced into OAv. (cf. 7.13.1).
In some forms, YAv. preserves the old final sequence when it occurs word-
internally: cf. YAv. raoc bii, (OAv.) YAv. raoc b (dat./abl.pl. and inst.pl., resp.,
of raocah- light, from a nom. *raoc ); vac b (inst.pl. of vacah- word; Skt.
vcas-) from an OAv. nom. vac (cf. 19.1.1). In these examples, can be inter-
preted as a composition vowel (cf. fn. 1 above).
4 Called Rckverwandlung [reversion] by analogy to some phenomena concerning Attic Greek
vocalism.
7 vowels 15
7.10. e comes from 1. an a after i , followed by a palatal consonant or a syl-
lable that contains i, ii, or e
(i.e., in a palatal environment): cf. iiiejah- aban-

donment (Skt. tyjas-); yesn (loc.sg. of yasna- sacrifice; Skt. yaj-) and its
derivative yesniia- (Skt. yajya-); OAv. xaiieh, YAv. xaiieite (2.sg.pres.act.ind.
and3.sg.pres.mid.ind., resp., of x; Skt. kyasi), sruuahiieit (3.sg.pres.act.ind.
of sruuahiia- seek fame).5 This change does not occur before r, uu, or hm:
cf. YAv. fraiiaire morning; mainiiauue (dat.sg. of maiiu- spirit); OAv. yahm
(loc.sg.m. rel.pron. ya-; contrast the gen.sg.m yehii).
2. In YAv., e may also derive from the sequence i a that is so frequent in the
gen.sg.m. of the thematic inflection: PIE *-osi o > *-ahi a > (OAv. -ahiia) YAv. -ahe
(cf. further 11.1.1). 3. On the YAv. development -e# < *-ai #, cf. 10.2.2.
7.11. 1. results from the gathacization of the YAv. sequence -e# < *-ai (cf.
10.2.2), which, in OAv., should have been -i (and indeed is, almost always). 2.
It is encountered in OAv. in some outcomes of the IIr. diphthong *ai : vada
,
mruii (< *mruu ai ; cf. 10.2.3); 3. it also appears in monosyllables ending in -e in
OAv.: t (nom.pl.m. dem.pron. ta-; contrast ate; Skt. t, Gr. ).
7.12. o 1. comes from an a after m, p, or u , followed by a syllable that contains
u (not u ): mou soon (Skt. maku
); pouru- much (Skt. pur-, OP paruv); vohu-

good (Skt. vsu-). If the intervening consonant is palatal, dental, or labiodental,
there is no rounding: pasu- livestock (Skt. pau-); YAv. mau- wine (Skt.
mdhu-, Gr. ). 2. o also occurs in some outcomes of the IIr. diphthong *au ;
cf. 10.3.1.
7.13. 1. corresponds to IIr. final sequence *-as#. This final sequence became
- inOAv. and- >- inYAv. (cf. 7.9.4), whence it was introducedintoOAv., sup-
planting the original - almost everywhere: YAv. vac (nom.sg.neut. of vacah-);
k (nom.sg.m. interr.pron. ka- who?); v (dat.gen.pl. encl. of the pers.pron.2.
tuu m you; Skt. vas); contrast OAv. vac (also vac), k , v . 2. appears as
the usual composition vowel (cf. 5.4. and fn. 1 above): par.kauum,
drguu.db. 3. In OAv., sometimes represents an a
or before the sequence

rC: OAv. crt (3.sg.aor.act.inj. of kar make, do; << *car-t; cf. Skt. -kar); OAv.
rtar- creator (Skt. tvar-; IIr. *tu ar- shape, create); OAv. drt
(3.sg.aor.act.inj. of dar hold; also OAv. drt; < *dhr--t). 4. also appears in
some outcomes of IIr. *-au : gtuu, hat; cf. 10.3.4.
5 In the course of transmission, forms such as as .yec (dat.sg. < as i a-c) also appear. On the
thematic dat.sg., cf. 19.1.7.
7.14. On the whole, i and are preserved as such in most cases, though, in a
number of contexts, i is lengthened to and can be shortened to i. i usually
appears 1. as the reflex of IIr. *i; Av. hi.haxti (3.sg.pres. act.ind. of sac follow;
Skt. sakti); pitum(acc.sg. of pitu- food; Skt. pit-); 2. as the shortenedoutcome
of IIr. * before uu (Av. piuuah- fat, Skt. pi
vas-); 3. as the YAv. outcome of

the development of (< *a; cf. 7.8) after i , c, j: YAv. yim (acc.sg.m. rel.pron.
ya-; < *i m < *i am; OAv. y m); YAv. yima- PN (OAv. y ma-, Skt. yam-); YAv.
drujim (acc.sg. of druj- Lie; OAv. druj m); YAv. hacite (3.pl.pres.mid.ind.; Skt.
scante);6 for the conditions on the preservation of a, cf. 7.8.3.
4. In isolated cases, and in a close relationship with the position of the
accent, i can also come from the vocalization * of a Proto-Iranian laryngeal
phoneme *H, in turn deriving fromthe PIE laryngeals *h
13
: cf. YAv. nom.sg. pita
<< Proto-Av. *pHta
r, influenced by the accent of the voc. p tar, versus the OAv.

dat.sg. fri < Proto-Av. *pHtrai . 6. On i as a product of epenthesis, cf. 8.1; on
anaptyctic i, cf. 9.2.5.
7.15. usually reflects IIr. *. It renders lengthening of short *i in the following
environments: 1. after u , uu, vh, andxv inanopensyllable: xvti- well accessible,
accessibility (hu+i-ti-); uuiia- apparent (fromthe adv. uui; Skt. [*a
viiya-]
a
viiya-); tuu- strength (Skt. tvi-); contrast uuisti- not finding (from
vid find [cf. 7.8.2]; Skt. vitti-); tuui-c brutality; suuita- most power-
ful (superl.; Skt. sviha-), YAv. stuuita- biggest (superl.; Skt. sthvia-).
also reflects *i and *i a (the latter only in YAv.) 2. before final -m#: YAv. axtm
(acc.sg. of axti- pain); dmm (acc.sg. of dmi- founder); YAv. paitm (acc.sg.
of paiti- lord), YAv. haim(acc.sg. of haiiia- real; Skt. satym); 3. as part of the
outcome of the sequence *ins: gair (acc.pl. of gairi- mountain: < *iNs); OAv.
cmah, YAv. cmaide (< *ci-n-s; 1.pl.pres.act.ind. and mid., resp., of ci gather
< PIE *k
u
ei s; Lat. crre); 4. in general, in the ending of the inst.pl., -bi
; 5. con-
sistently in the preverb v separately, dis- < *vi. 6. Sometimes, *i is lengthened
in open, mainly word-initial, syllables: OAv. jgrzat complains < *ji-gr z-a-t,
YAv. zzana- beget < *zi-zan-a-; 7. *i is often lengthened before and : mda-
prize < *mida-. 8. On the YAv. development of a > > i
, cf. 7.14.3 and fn. 6

below.
6 *i in the sequence *Ci N developed to *i i and later to YAv. i
. From there, it was introduced

into OAv.: YAv. aini
m(acc.sg. of aniia- other; Skt. anym) beside OAv. anii m, ainm; YAv. frm
(acc.sg. of friia- dear; Skt. priym); OAv., YAv. haim (acc.sg.of haiiia- truth; Skt. satym)
beside OAv. haiii m.
8 epenthetic vowels 17
7.16. u and also largely retain their etymological distribution, though *u is
affected by lengthening more regularly than*i. Av. short u canbe: 1. the reflex of
*u in a closed syllable (Av. uxan- bull, Skt. ukn-, supti- shoulder, Skt. pti-)
or in a non-initial open syllable (ahura- lord, tauruna- young); 2. the reflex
of * before ii: apuiiat- not deteriorating < *apuHi ant- (Skt. pyati stinks). 3.
Onepenthetic u, cf. 8.2; onanaptyctic u, cf. 9.2.3. 4. Onthe YAv. development
of a > > u
, cf. 7.17.5.
7.17. usually reflects: 1. IIr. *; 2. IIr. *u in an open initial syllable (bna- bot-
tom < *buna- < *budna-; stta- praised < *stuta-); 3. *u after *i ( ykta- yoked,
Skt. yukt-); 4. *u when affected by i-epenthesis (hiri- ahuric < *huri-,
vrnii cover! < *vr nudi); 5. *u and *u a before final -m#: (a) gtm(acc.sg. of
gtu- road; Skt. gt-); dax iim (acc.sg. of dax iiu- country; Skt. dsyu-); (b) as
the YAv. outcome of the development of (< *a; cf. 7.8) after u : YAv. tm you
(nom.sg. of the 2.pers.pron.; OAv. tuu m, Skt. tvm); YAv. tanm(acc.sg. of tan-
body; OAv. tanuu m, Skt. tanvm); YAv. riu
m(acc.sg. of riuua- third); YAv.

haurm (acc.sg. hauruua- all, whole; Skt. srvam); 6. as part of the outcome
of the sequence *uns: aidii (acc.pl. of aidiiu- harmless; Skt. dyu-); xrat
(acc.pl. of xratu- intelligence; Skt. krtu-); 7. also represents the YAv. outcome
of final *-ans after u by way of *uu h > *uu > *uu > : YAv. zr (gen.sg. of
zruuan- time); YAv. h (gen.sg. of xvan-).
8 Epenthetic Vowels
The insertion of the epenthetic vowels i and u in specific contexts is a very
common phenomenon in the Avestan texts. The epenthesis of i indicates the
palatalization of the affected phonemes, while that of u indicates labialization.
It seems that epenthesis occurs only once per word, and can coexist with
anaptyctic vowels (9).
8.1. i is inserted before consonants followed by i
, ii, or e
: irixta- (adj. from ric

leave; Skt. rikt- < *lei k
u
-); iiiejah- abandonment (Skt. tyjas-); Av. airiiaman-
tribe (Skt. aryamn-); aibii (dat.pl.m. of the dem.pron. ta- this; Skt. ebhys);
kainn girl (Skt. kan ); xvpaiiia- fecund (Skt. svapaty-); baraiti
(3.sg.pres.
act.ind. of bar carry; Skt. bhrati); kriti- pattern, arrangement (Skt. a
kr ti-);
mriiiu- death (Skt. mr ty-); daibit (adv.; Skt. dvita
); OAv. daibiaiiat-
enemy (YAv. t biaiiat- < *du ei s-; cf. Skt. dvi Gr. ); hacaite
(3.sg.pres.
mid.ind. of hac; Skt. scate); haxmain (dat.sg. of haxman- retinue; cf. the same
ending in Skt. -mane); zairimiia- house (Skt. harm
i
y-).
Epenthesis of i is not foundbefore , , st, t, m, or hm, nor after
(<*aN). The
enclititc caprevents the epenthesis that endings ini
ande
induce: drguuata-
c (dat.sg.m.) versus drguuite; j ghati-c (3.sg.) versus s ghait.
8.2. u appears exclusively before ru and ru : pour (acc.pl. of pouru- much);
Skt. pur-; YAv. duru- wood (Skt. da
ru-); hauruuatt- wholeness (Skt. sarv-

tti-). In some cases, metathesis together with a later epenthesis is at work:
uruuata- law (
u
ruu /ru / < *u r; Skt. vrat-); uruudah- happiness (cf. Skt.
vrdh).
The sequence *ru i
has a special development in YAv., namely, the u was

lost, though it left a trace through epenthesis, which affected the preceding
vowel or created a diphthong, and thus permitted the later epenthesis with
i: YAv. paoiriia- first < *pau rii a- < *paru ii a- > OAv. pa(o)uruiia-; brtu
iriia-
cousin <*brturi a- <*brtru ii a- (Skt. bhra
tr vya-); tiriia- uncle <*p()turi a-

< *(p)tru ii a- (Skt. pitr vya-).
9 Anaptyctic Vowels
Frequent instances of anaptyctic vowels, also referred to as svarabhakti, occur
inthe text. Anaptyctic vowels arise inclusters of occlusives, thus facilitating the
liturgical recitation. In addition, most of the time, they have been introduced
by different schools in the course of transmission (cf., for example, 4.1.2,
4.1.3). The vowel after r is purely phonetic (i.e., it is neither present in a words
underlying form, nor inserted through any phonological process). Anaptyctic
vowels are metrically irrelevant.
9.1. One may speak of several different anaptyctic vowels ( , a, o, i), though
the usual one, which appears in the majority of cases, is : YAv. dmna- house
(Skt. ma
na-); xvaf na- dream (Skt. svpna-); f ri (dat.sg. of ptar- father;

< *pHtrai > Skt. pitr); haxm (nom./acc.sg. of haxman-); rtar- (Skt.
tvar-; cf. 7.13.3); nr (acc.pl. of nar- man < *nrN); rzu- straight (Skt.
r j-); darra- support (Skt. dhartr-); darsm(1.sg.aor.act.inj. of dars see; cf.
Skt. dram); two cases that have already been mentioned in which anaptyxis
combines with epenthesis are kriti- (8.1) and mriiiu- (8.1).
Final -r usually appears as OAv. -r and YAv. -r: vadar
weapon (neut.; Skt.

vdhar-); huuar
sun (stem xvan- neut.; Skt. svr-).

10 diphthongs 19
9.2. Ineffect, the rest of the vowels are usually employedmuchless frequently:
sometimes seen are 1. an a: star m-ca (gen.pl. of star- star; v.l. str mca; Skt.
star-); s
iiaoana- act, deed (v.l. s
iiaona-; YAv. s
iiaona-; cf. further 11.23);

varat (3.sg.aor.mid.inj. of var choose; metrically disyllabic; but fra-uuart).
Anaptyxis combines withepenthesis indaibit (8.1). Insome evenrarer cases,
the following vowels appear: 2. : OAv. d jmspa- PN (YAv. jmaspa-); 3. u:
brdubii (dat.pl.); YAv. surunaoiti (3.sg.pres.act.ind. of sru); 4. : ardm
(2.pl.aor.mid.inj. of ars shape; versus rzdm from r); garb (inst.pl. of
gar- song); 5. i: azdib (inst.pl. of ast- bone; v.l. azdb); mazib (inst.pl. of
maz- great); YAv. ni-sirinaoiti (3.sg.pres.act.ind. of sri lean).
10 Diphthongs
As a result of the changes of PIE *e and *o to IIr. *a, and of PIE * and * to IIr.
*, Indo-Iranian possessed only four diphthongs: two with a short vowel, *ai
and *au , and two with a long vowel, *i and *u . In the Avestan writing system,
these diphthongs are not represented by their own graphemes, but their vari-
ous outcomes are reflected, rather, by the combinations of vowel signs. In this
way, the phonetic details are rendered.
The IIr. diphthong *ai occasions two distinct outcomes, depending upon the
position in the word in which it is found. Although it is possible to systematize
these outcomes, unexpected correspondences are found at times.
10.1. The spelling a represents 1. the outcome of the IIr. diphthong *ai in (a)
initial position: aa- search (Skt. a-); ama- ire (cf. Skt. yati, Gr. ,
Lat. ra); YAv. asma- firewood (< *ai zma- < *ai dzhma-; cf. Skt. idhm-); or (b)
in an open syllable: vada
1./3.sg.perf.act.ind. of vid know versus 2.sg. vist; cf.

another example below 10.2.1.7
2. a is likewise the outcome of the group *ai a before a nasal, by way of the
possible development > *ai > *a > a: OAv. am (nom.sg. of ima- this; also
OAv. aii m, Skt. aym); vam (nom.pl. of azm I; Skt. vaym); gam (acc.sg.
7 i is frequently found instead of a: OAv. ciat (3.sg.aor.act.sbj. of cit notice); cim(1.sg.aor.
act.inj. of ci); di (1.sg.aor.act.sbj. of dis show); miat (3.sg.aor.act.sbj. of mi leave);
YAv. biiiimaie (1.pl.pres.mid.opt. of bud sense; cf. Skt. bhvemahi of bhavi; versus 3.sg.
biiiata; cf. Skt. bhveta); OAv. vaocim-ca (1.pl.aor.act.opt. of vac; cf. Skt. bhvema; versus
OAv. apam to p, or hanamc to han).
of gaiia- life; Skt. gyam); YAv. brjam (1.sg.pres.act.inj. of barj greet); YAv.
v-dram (1.sg.pres.act.inj. of dar hold).8 The YAv. form ram (acc.sg. of ra
ii-
wealth; < *reh
1
i-; cf. Skt. raym) falls under the preceding development: *rai im
> *raim > ram.
10.2. The spelling i reflects the outcome of the IIr. diphthong *ai 1. in a closed
syllable: YAv. vin- wave versus vaa- strike (Skt. vga-); 2. in final position
in OAv.: nari (dat.sg. of nar- man; YAv. naire, Skt. nre); fri (dat.sg. of ptar-
father; YAv. pire, Skt. pitr);9 in this position, YAv. shows -e (except in yi and
maiiii), which was in turn (re)introduced into OAv. (7.11.1).
3. Some cases, such as OAv. mruii (1.sg.pres.mid.ind. of mr), OAv. tanuii
(dat.sg. of tan-), and YAv. uiie (nom.du.f. of uba- both; OAv. ub), exhibit a
phonetic development *-u ai > *-uu ai > OAv. -uu e > YAv. -uiie
, which was in turn

introduced into OAv.
10.3. The spelling ao represents 1. the outcome of the IIr. diphthong *au : OAv.
aojah- strength (Skt. jas-; cf. Lat. augre); aoah- burn (cf. Skt. oati, Gr. ,
Lat. r); raocah- light (Skt. rocas-, Gr. , Lat. lx; < PIE *leu k); sraot
(3.sg.aor.act.impv. of sru); mraot (3.sg.pres.act.impv. of mr speak); 2. in some
cases, YAv. ao reflects a recent (non-original) sequence au: YAv. paoiriia- first
(8.2; OPparuviya-); YAv. aoi (variant of auui, alsoYAv. aii, OAv. aib, Skt. abh ).
3. ao is likewise the outcome of *au a before a nasal, by way of a possible pho-
netic development *au > *au u > *au > ao: naoma ninth (Skt. navam-); YAv.
abaom (1.sg.impf.act.ind. of b; Skt. bhavam); YAv. mraom (1.sg.pres.act.inj. of
mr; Skt. bravam).
4. In final position, the IIr. diphthong *-au becomes -uu in the majority of
cases: YAv. da huu (loc.sg. of da hu- country; < *dahi au -); OAv. huu that
(OP hauv < *hau ); YAv. huxratuu (voc.sg. of huxratu- intelligent; Skt. sukrato).
In some cases, *-au resulted in -: YAv. zastaii (loc.sg. of zasta- hand); YAv.
vaii (voc.sg. of vaiiu-; Skt. vyo); YAv. hat (loc.sg. of hatu- bridge). It is likely
that *-au underwent monophthongization to (parallel to the change of *-ai to
8 In contrast, the forms of the 3.pl. in *-ai an are transmitted as -aiin: brjaiin, vi-raiin,
etc.
9 a is found instead of i in some words: YAv. masma- urine (maza- id.; cf. Gr. ;
< *h
3
mei h-); OAv. nastar- slanderer (from nid: cf. 10.5.1); YAv. raa- clump (and its
denominative); raat- warrior (and rait-; Skt. rathea
-); YAv. pairi.uruuata- he

who best destroys; srata- most beautiful (Skt. rha-); OAv. hamastar- despoiler (from
mi).
11 consonants 21
-e, cf. 10.2.2), which later diphthongized in [u] (written -uu) in many forms,
though not after -ii-.
10.4. The spelling u represents, as a general rule, the outcome of the IIr.
diphthong *au before -#: cf. the gen.sg. forms g u, maii u, OAv. dax ii u,
YAv. da h u. In YAv., final -ao (diiao [from diiauu-], drao [from duru-],
ranao) is more frequent and probably results from restoration of *-au- in the
u-stems. Final -ao was also introduced in OAv.:
+
mriiiao; conversely, OAv.
final - u expanded into YAv.
10.5. The spelling i represents 1. the IIr. diphthong *i : OAv.
+
i (3.sg.aor.
act.ind. of i desire; cf. Skt. ait); OAv. di (2.sg.aor.act.inj. of dis point;
< *di--s < *di -s-s); YAv. nism, nist (1.sg.act.ind, 3.sg.aor.act.inj. of nid
insult, reproach);
2. the sequence i before a nasal canbe the result of *i a, by way of a possible
development > *i > *i i > i: YAv. gim (acc.sg. of giia- step; Skt. gy-);
OAv. humm (acc.sg. of humiia-; Skt. sumy-). On *i a > ai a, cf. 7.3.2.
3. Note that epenthesis with i after does not differ graphically from the
original diphthong i: drguuite.
10.6. The spelling u represents 1. the IIr. diphthong *u : YAv. gu cow (Skt.
gu); xnu (3.sg.aor.act.inj. of xnu- welcome); vahu (loc.sg. of vohu-);
+
xratu (loc.sg. of xratu-; Skt. krtau).
2. u is likewise the result of *u a before a nasal, by way of a possible
phonetic development > *u > *uu > u: YAv. nasum (acc.sg. of nasu-
cadaver); cf. further YAv. as um (voc.sg. of as auuan-; Skt. r ta
van-; cf. 11.3.3).

On *u a > *au a, cf. 7.3.3.
3. Epenthesis with u after does not differ graphically from the original
diphthong u: duru- (Skt. da
ru-, Gr. ).
11 Consonants
The system of consonants reconstructed for Indo-European (cf. 6), in its his-
tory leading up to Avestan, has undergone a considerable number of changes,
which have completely disfigured the original system. These changes are in
part due to the developments of the separate phonemes on their own, and in
part to the specific developments of groups of two or more consonants. The
comparison of the Iranian material with that of Indo-Aryan allows, again, the
reconstruction of an Indo-Iranian consonant system.
In consonantism, some notable differences exist between OAv. and YAv.,
for example, as regards the preservation of Bartholomae clusters (11.11.1),
fricativization (11.11), etc.
The consonants of Iranian may be classified into 1. sonorants (semivowels
and liquids); 2. nasals; 3. occlusives; 4. fricatives; 5. sibilants. The consonants
maintain, withsome variation, a similar organizationinAvestan(where further
new fricatives have arisen).
11.1 The Semivowels: i , u
1. The PIE sonorant *i is maintained as such in Indo-Iranian and in Avestan
as well, where it is represented by ( y) at the beginning of the word and by ii
word-internally. It has already been pointed out above that the letter was
probably, from the outset, intended to graphically represent initial #i , while in
the Indian manuscripts y , which would be the letter corresponding to the
palatal fricative z
(< Ir. *i ), was employed.

Medial i often disappears before e (cf. 7.10): Av. vaheh (nom.pl.fem. of the
comparative vahiiah- to vohu-; < *u ahi eh-; Skt. vsyas); the YAv. ending of the
dat.sg. of i-stems: - e < *ai ai > Skt. -aye, OAv. -iii; also in YAv., the ending of
the thematic gen.sg.masc. -ahe (but OAv. -ahii 7.10.3).
2. The PIE sonorant *u is maintained as such in Indo-Iranian and in Avestan
as well, where it is written as v at the beginning of the word and as uu word-
internally.10
3. The distinct graphic values of the semivowels according to their position
inthe word necessarily reflect distinct pronunciations. It is quite probable that,
in initial position, the semivowels could have begun to develop into voiced
fricatives (as in a prestage of Modern Persian: < *i and b < *u ), while in
medial position, i and u would have reflected ii and uu , respectively: Av. friia-
dear (Skt. priy-); YAv. druua- hale, steadfast (Skt. dhruv-). The orthographic
conventions of Avestan (cf. 7.2) further indicate that YAv. jiia bowstring
(Skt. jya
-) and kuua where (Skt. kv) must be disyllabic (in virtue of having
-#).
These changes to *i and *u must have taken place in western Iran, most
likely influenced by Old Persian, and are commonly considered to be present in
the archetype. In fact, in Old Persian, post-consonantal semivowels are written
with iy and uv: cf. Av. aiia-, OP aniya-, Skt. any- < *ani a-; Av. hauruua-
OP haruva-, Skt. srva- < *saru a-. However, this Old Persian phenomenon
10 ii and uu used to be regularly transcribed as y and v, respectively, since they mainly
represent ii and uu . More on this issue in section 11.1.3.
11 consonants 23
sometimes occurs in intervocalic position as well: cf. OP a-draiya, Av. draiia-
or OP bauvatiy, Av. bauuaiti.
Consequently, the spellings ii and uu in Avestan represent ii and uu , which
would in turn come from *i and *u , though they may also represent an original
sequence *ii /*uu : OAv. iit < *-ii t < PIE*h
1
i-i h
1
-t (prev. + 3.sg.pres.act.opt.
of i go; cf. Skt. iya
t); YAv. sraiiah- (comparative of srra- beautiful, excellent;

Skt. ryas- < IIr. *rai H-i as-); YAv. gauustriia- pertaining to the pasture, from
*gau -u strii a-.
4. The phonetic development of some phonemes and sequences of sounds
has also produced non-etymological instances of ii and uu. One such example
is the assimilation of u to i in the sequences *-uu and *uu ai; these become
-uii by way of *-uu i : ahuii (dat.sg. of ahu-, Skt. su-), which comes from
*ahuii < *ahuu i < *ahuu < *ahu < *ahu ai , following the view presented
under 10.2.3.
Another source for uu is the development -b- > -uu-, which we find in YAv.
alongside the expected --: YAv. auui (also written aoui and aoi) and aii to,
towards alongside OAv. aib, Skt. abh; YAv. uiie (< *uu ai < *uai ) alongside
OAv. ub; auuauuat (3.sg.impf.act.ind. of b < *aau at; also abauuat ), etc.;
guruuaiia- (to grab seize), Skt. gr bhy-, versus YAv. grniti or OAv. h -
grabm (1.sg.aor.act.inj.); cf. further the ending of the inst.pl. in -uu < -u-i,
or, likewise, some dat.pl. forms such as nruii (to nar-), OAv. nrbii, Skt.
nr bhyas; YAv. as auuaoii, OAv. as auuabii (to as auuan-). Careful examination
of this phenomenon allows for the establishment of a relative chronology: the
development -- > -u- took place in YAv. before i-epenthesis took place (thus
auui and not

aiuui), but after the IIr. sequence *au i had become uui (i.e., u i;
cf. 7.8.2). This newoutcome is not distinguished fromold uu inits subsequent
development, as the already cited YAv. uiie (cf. above and 10.2.3), versus OAv.
ub, demonstrates.
5. From the combination of semivowels with various consonants (especially
when the consonant precedes the semivowel), a variety of articulatory changes
has taken place: *ni > under 5.2.4, 11.4; Ir. *cu > sp under 11.10.3; Ir. *ju > zb
under 11.14.2; Ir. *i > OAv. s
ii / YAv. s
under 11.23.1; Ir. *hi - > x ii- under 11.28;

Ir. *-hi - > - h- under 11.29; Ir. *hu > xv under 11.30; *-hu > vh under 11.31; etc.
On the metathesis of u r, cf. 8.2.
11.2 The Liquid: r
1. The two realizations of the IIr. phoneme *r < PIE *r (together with that
originating fromIIr. *l < PIE *l) have produced the same result in Avestan: both
come out as an r, though that deriving from syllabic *r is written as r (with
graphic variations).
For the realization as *r, cf., for example, Av. raa- chariot, Skt. rath-; Av.
vouru wide, Skt. ur- (cf. Gr. ); Av. raocah- light (cf. Skt. rcate shine
andLat. lx); Av.
+
rarizaite (3.sg.pres.act.intens.ind. of riz lick), Skt. rrihat (cf.
1.sg.pres. Skt. rhmi and the younger lhmi), cf. further Modern Persian litan,
Gr. , and Lat. ling.
For the realization as *r , cf., for example, YAv. krta- (ppp. of kar make,
do), Skt. kr t-, OP k-r-t- (i.e., *kr ta- [krta-]); Av. rzu- straight (Skt. r j-);
Av. pru- broad, Skt. pr th-, Gr. < PIE *pl th
2
-. The is not usually
written following a t: trm (acc.sg. of tar- fire).
2. When the clusters rk and rp were immediately preceded by the Iranian
accent, the outcomes hrk and hrp, without anaptyxis, are found: YAv. mahrka-
destruction (OAv. marka-, Skt. mrka- PN); YAv. vhrka- wolf (< *u rka-; cf.
Skt. vr ka-); YAv. khrp- figure, form, body (< *k rp-). In contrast, under other
accentual conditions, the developments rk and rp, respectively, are found:
OAv. marka-c, Skt. mark- death, destruction. One commonly acceptedthe-
ory proposes that a voiceless r underlies the graphic sequence hr of these forms.
3. Under the same conditions as in the preceding groups, the sequence rt
does not produce the expected sequence

hrt, but rather s : Av. mas iia- man,
Skt. mrtiya-; OAv. ams a- immortal, Skt. amr ta-; ps an- battle, Skt. pr tan-.
This s was probably a voiceless lateral fricative, as some (Middle) Persian
borrowings from Avestan, in which hr/hl are written for s , seem to show.
4. The result of PIE *r H > Ir. *ar > Av. ar is not distinguished from the
sequence ar < PIE vowel + *r: OAv. darga- long, Skt. drgh- < PIE *dl h
1
gh-
(cf. Gr. ); fra-uuart (3.sg.aor.mid.inj. of var choose <*u l h
1
-t-; cf. Skt.
vr t).
11.3 Nasals: n, m
Nasals found in Avestan may be either etymological or may have emerged
from particular sound changes (cf. below 11.13). Depending upon the envi-
ronment in which they are encountered, all the nasals also undergo minor
articulatory changes that are represented in writing through specific letters of
the Avestan alphabet. At times, the nasal phonemes disappear, though they
leave behind nasalization as a trace.
1. As a general rule, the PIE nasals *m and *n are preserved in Proto-Indo-
Iranian: nmanm(gen.pl. of nman- name, Skt. na
man-, Lat. nmen); nmah-

homage, Skt. nmas-, cf. the Gr. neut. wooded pasture and Lat. nemus
forest; mraoiti (3.sg.pres.act.ind. of mr; Skt. brvti < *mlu H-ti; cf. Russian
molv remark, rumor).
2. Preceding an occlusive (t, d, k, g, c, j, [p], b), the nasals are usually repre-
sented with the letter : atar within (OP an-ta-ra, Skt. antr); spta- ben-
11 consonants 25
eficient; jat (3.sg.aor.act.impv. of gam go, come); pardi- fecundity (Skt.
prandhi-); badaiieiti (3.sg.pres.act.ind. of bad bind; Skt. bandh); h kriti-
(derived from ham+kar); j ghati-c (3.sg.aor.act.subj. of gam); s gha- expla-
nation (Skt. am sa-); paca
five (Skt. pca, Gr. ); YAv. rj-ita- (superl.

< PIE *h
1
leng
u h
-; cf. YAv. ruu- swift and Skt. r/lghyas-); upaskabm pillar;
frascibana- beam.
Besides numerous graphic variants (especially with the sequence mb), ex-
ceptions to this neutralization are seen before the enclitic ca
, which usually
maintains the original nasal: uzuxiinca, vspmca, as mc, dargmc.
3. In word-final position, and if the same syllable began with a labial, we
find m for original n: OAv. camm (loc.sg. of caman-); OAv. nmm (acc.pl.
of nman-); YAv. as um < *as u n (voc.). On the nasalization of the vowel
in the final syllable, cf. 7.6. On the voiceless nasal m , also written as hm, cf.
5.2.5.
11.4. The palatal nasal is encountered in the oldest manuscripts, fromwhich
it could be deduced that this sign was already present in the archetype. is
written before ii (i ), and i; cf. 5.2.4 as well. The graphic restoration of n for
occasionally occurs in the manuscripts.
11.5. The (dorsal) nasal represents, on the one hand, an etymological sound
that derives from * < *k < *nk, such as in the YAv. word patavhum a
fifth < *paktahu m (from PIE *penk
u
to-). On the other hand, has arisen
from a development of Ir. *h < *s, which, as a function of its surroundings,
also produced other outcomes, namely, and v. The details concerning these
phonemes are found in the paragraph dedicated to the sibilant (11.19ff.).
11.6 The Occlusives
As a general rule, the inherited occlusives have been well preserved from
Proto-Indo-European into Avestan, though some changes in place of articu-
lation, conditioned by environment, are already to be seen in Proto-Iranian
(Ir.); these changes have produced new sounds and articulatory series that did
not exist in Indo-European or Indo-Iranian (cf. 11.9, 11.12.2, and esp. 11.11 and
11.15).
11.7. The phonemes that underwent the most changes in the subsequent
development of the Indo-European phonological system(cf. 6) are those that
belong to the dorsal series. In the course of their development into Avestan,
the velar and labiovelar series fell together in a single velar series. In virtue of
this characteristic, Avestan is classified as a satm language (as opposed to the
languages referred to as centum, such as Latin or Greek, which exhibit distinct
velar and labiovelar series, and in which the palatal series has been subsumed
under the velar series). In order to see the development of the three dorsal
series at distinct stages of the Indo-Iranian languages, a synoptic chart with
the outcomes of the primary (I) and secondary (II) palatals is given here. For
details, see 11.9, 11.12.2, 11.20.3, 11.24.
table 3 The PIE dorsals in Indo-Iranian
Indo-European Indo-Iranian Iranian Avestan Sanskrit
I * * *h * *j *j h *c *j s z j h
II *k *g *gh * * *h * * c j c j h
*k
u
*g
u
*g
u h
In Common Indo-European, a difference still would have existed between the
palatovelars and plain velars or labiovelars preceding the front vowels *e
, *i
. In
the stage common to Indo-Iranian, the Indo-European palatovelars (I) became
dental affricates, while the plain velars and labiovelars (II), already having
fallen together in a single velar series, underwent a conditioned palatalization.
Subsequently, the development diverges:
(a) in the Iranian branch, the dental affricates (I) lost their (pre)palataliza-
tion and remained as dental affricates [ts] and [dz], resp. In Avestan, these
affricates further developed to simple sibilants (cf. the parallel development
of Skt. mtsya- and Av. masiia- fish). The palatalized plain velars and labiove-
lars (II) became affricates. In Iranian, the loss of aspiration left only two voiced
segments: *j and *.
(b) Inthe Indo-Aryanbranch(Skt.), I andII ultimately give similar outcomes
(thoughcf. c): the palatalizationis preservedinthe voiceless andvoicedseries,
but the aspirate is depalatalized.
11.8 The Voiceless Occlusives
The Avestan series of voiceless consonants k, c, t, p has, as its immedi-
ate predecessor, the Iranian series *k, *, *t, *p, resp. In preconsonantal posi-
tion, the voiceless occlusives k, t, p become the fricatives x, , f, resp., though
some regular exceptions, which will be appropriately detailed (11.15ff.), are
detectable.
11 consonants 27
11.9. Av. k continues both PIE *k and *k
u
: Av. k-ma- desire, Skt. ka
ma-, cf. Lat.

crus, from the PIE root *keh
2
; the Av. neg.interr. kat , Skt. kd, Lat. quod < PIE
interr./indef. *k
u
od.
IE *k and *k
u
, when preceding a front vowel (e
, i
), developed to Ir. * > Av. c:

the Av. encl. particle ca
, Skt. ca, Lat. que, Gr. (Mycenaean -qe) < PIE *k

u
e;
YAv. caxra- wheel, Skt. cakr-, Old English hweol, Gr. < PIE *k
u
k
u
lo-s;
YAv. pacata(3.sg.pres.mid.inj. of pac cook), Skt. pcata, cf. Lat. coqu (<*k
u
ek
u
< *pek
u
), Gr. (< *pek
u
-i e/o-); cf. the synoptic chart under 11.7.
Within a paradigm where there was variation in vocalism on account of
ablaut (13.2) (such that the contact of velars would alternate between the PIE
vowels *e
, *o
(> Ir. a
) and the PIE resonants *n , *m , *r , *l ), a single variant of the

velar is sometimes generalized throughout the entire paradigm. For example,
the root *kar is found inthe Avestanparadigmas kar/car, but Skt., however, has
generalized the variant kar throughout; cf. further 11.12.2.
11.10. Av. t comes fromPIE*t, and Av. pcomes fromPIE*p: Av. ptar-, Skt. pitr-,
Lat. pater < PIE *ph
2
tr-.
1. t is lost before s (cf. 11.11.2), but is usually preserved following s/ (cf.
11.17.4). Note further the IIr. final sequences *-nt > IIr. *-n > Av. -n, and IIr. *-st
> -s: OAv. dadn (3.pl.pres.act.subj. of d; Skt. ddan); OAv. rriin (3.pl.pres.
act.subj. of rah); YAv. barn (3.pl.pres.act.inj. of bar < *bar-a-nt > Skt. bhran);
YAv. vidn (3.pl.pres.act.inj. of vid); OAv. dids (3.sg.pres.act.inj. of dh <
*didns-t); OAv. vs (3.sg.aor.act.inj. of van < IIr. *u n-s-t); Av. a
s (3.sg.impf.
act.ind. of ah < IIr. *s-t > Skt. s a
st).
2. t appears instead of t in word-final position after a vowel or r, as well as
in the YAv. sequence t b- = OAv. db- (11.14.1). The Avestan alphabet must have
beenrepresenting anunreleasedocclusive using this t : prsat (3.sg.pres.act.inj.
of fras ask); buuat (3.sg.aor.act.subj. of b); conj. yt (Skt. ya
d); ymat (abl.pl.

of the 2.pers.pron., Skt. yumd); xart (abl.sg.); OAv. crt (3.sg.aor.act.inj. of
kar < *kar-t). In the form t kaa-, the t k has arisen through the missegmenta-
tion of *aniiat .kaa- who has something different from the doctrine in YAv.
anii.t kaa-.
In forms such as OAv. yaogt < (3.sg.aor.act.inj. of yuj join < *i eu K-t),11 the
sequence -gt is not easy to interpret. It has been pointed out that in the cases
with the final sequence -gt , the t could serve as a sort of diacritic to indicate
word-final implosion. The other occlusive that is permitted word-finally is ,
11 In word-internal position, cf. OAv. hxta- (from hu + ppp. of vac; Skt. skt-); YAv. baxta-
(ppp. of baj apportion).
which should also be considered as unreleased; cf. 5.2.2.
3. p does not become a fricative, but rather remains as an occlusive preced-
ing t: YAv. hapta seven; OAv. iiapta- help (p+ta-); naptiia- (derived from
napt- grandson), but not before the cluster tr (cf. 11.10.3): fri (dat.sg. of
ptar-); YAv. nafrr- (gen.sg. of naptar-/[napt-]; Skt. npt- and nptar-); OAv.
rafra- aid (from rap).
A p of secondary origin derives from the PIE sequence *u > IIr. *u > Ir. *cu
that develops in Avestan (and some other Iranian dialects) to sp: YAv. aspa-
horse (Skt. va- < PIE *h
1
eu o-).
11.11 The Voiced Occlusives
The Iranian series of voiced occlusives, *g, *, *d, *b comes from the voiced
andvoicedaspirate series of Indo-European, andhas developedinto OAv. g, j, d,
b, resp., while in Young Avestan, these consonants remain as voiced stops only
in initial position. Otherwise, in YAv., they spirantize and become the voiced
fricatives , , , , resp., except following a nasal or a sibilant. The labial
is further lenited to uu, except preceding i or ii, where it remains in most
words.
1. A special cluster treatment dating to the Indo-Iranian period (which may
even be of Indo-European antiquity) is Bartholomaes Law, or simply Bartho-
lomae, in honor of its discoverer. In effect, Bartholomae observed that, in a
sequence of a voiced aspirate and a voiceless stop or sibilant in Indo-Iranian,
the voiceless stop was voiced, and the aspiration passed to the end of the
cluster: PIE DhT, DhS > IIr. DDh, DZh, respectively. This change can be seen
throughout all of Indo-Iranian, and in Avestan it is interesting to note the dif-
ferences that emerge between the dialects. Old Avestan preserves the results of
Bartholomae exceptionlessly, while Young Avestan has all too often analog-
ically remade the outcomes of Bartholomae: OAv. aogd versus YAv. aoxta,
both from IIr. *au g-dha < *au gh-ta (3.sg.pres.mid.inj. of aoj say); cf. also OAv.
aog (2.sg. < *au g-ha <*au gh-sa). YAv. does preserve some exceptional cases,
such as ubdana- made of fabric, which contains the ppp. of vaf weave: in
ubda, Skt. ubdh- < *(H)ubh-t-, cf. Gr. .
2. When two dental stops coincided, an epenthetic sibilant arose (see
11.20.2), which was also susceptible to Bartholomae: Av. vrzda- (ppp.
to vard grow, Skt. vrddh- < *vrdh-t-); Av. dazde
(3.sg.pres.mid.ind. of d <
*dhadh-tai , Skt. dhatt, versus daste
< *dad-tai ). These forms show the way in

which the sibilant developed in the cluster *-dd- > *-dzd-: it lost the first dental
of the cluster and became -zd-.
11 consonants 29
11.12. Iranian *g < PIE *g, *gh, *g
u
, and *g
u h
is found in (a) OAv. baga- portion,
YAv. baa-, Skt. bhga-; OAv. ugra- mighty, YAv. ura-, Skt. ugr-; YAv. aguta-
toe, Skt. aguh-; (b) OAv. darga- long, YAv. dara-, Skt. drgh-, Gr.
< PIE *dl h
1
gh-; YAv. mazga- marrow Skt. majjn-, OHG marg < PIE *mosgh-;
YAv. zaga- ankle, Skt. jgh-; (c) Av. gu (nom.sg. of gauu- cow), Skt. gu,
Gr. < PIE *g
u
eh
3
us; (d) Av. garma- heat, Skt. gharm- < PIE *g
u h
or-m-,
cf. Lat. formus, Gr. .
In YAv., (< g) is lost before u / uu: cf. YAv. druuat- versus OAv. drguuat-;
YAv. raom swift (acc.sg. < *raum), cf. Skt. raghm; YAv. Mourum, OP Margum
the country Margu. However, could be reintroduced through intraparadig-
matic analogy: YAv. drim (acc.sg. of driu- poor) after driao (gen.sg.).
1. appears in final position in place of g, and is thus probably unreleased;
cf. 5.2.2. Besides t , is the only stop permitted in absolute auslaut. On the
relation to the other unreleased consonant, t , cf. 11.10.2 (end).
2. The same conditioning factors that caused Ir. * > Av. c (11.9) induce
Av. j (< Ir. * < PIE *g, *gh, *g
u
, and *g
u h
before the front vowels e
, i
): YAv. jiia
(11.1.3), Skt. jiya
< PIE *g
u
iHeh
2
(cf. Gr. ); Av. jaiti (3.sg.pres.act.ind. of jan
beat, kill), Skt. hnti, cf. Gr. < PIE *g
u h
en; and especially OAv. aog versus
aojah from the s-stem aojah-, cf. Lat. gen-us (< *-os): gen-er-is (< *-es-es); cf.
the synoptic Table 3 under 11.7.
j is frequently extended throughout a paradigm, rather thanalternating with
g: cf. the formations to the root gam < *g
u
em-, which usually appear in Avestan
with jam, whereas Sanskrit has generalized gam: cf. the Av. pres. stem ja-sa-
versus Skt. g-cha-, both from *g
u
m -ske- (cf. impv. Gr. ).
11.13. Av. d comes from Ir. *d < PIE *d/dh: Av. dasa
ten, Skt. da < PIE

*dem ; Av. dar hold: OAv. draiiat , Skt. dhar: dhryati. The merger of the
two Indo-European phonemes *d and *dh in Iranian resulted in an important
homonymy: Av. d give, place corresponds to both Skt. d give, Gr. , Lat.
d < PIE *deh
3
and also Skt. dh place, Gr. , Lat. fc < PIE *dheh
1
(though
they can be distinguished in a few instances, cf. 11.11.2).
Apart fromthe conditions formulated above (11.11), some further instances
in which the YAv. fricativization of d to is not found may be mentioned:
in reduplicated formations (dada, etc.), in compounds with preverbs
(vdauua-, etc.) on account of influence fromthe simplex; likewise, it does not
usually occur when the d is in contact with an r (arduu-, rduua-, xudra-,
varduua-), though lenited forms (ara-, ra-, bura-, vare, etc.) are not
rare.
In word-internal position, Ir. -dn- simplifies to Av. -n-: Av. bu
na- depth,
ground, < *budna-, Skt. budhn-; Av. asna- near < *n sdno- (cf. the Skt.
comparative ndyas-, YAv. nazdii); xvana- melted < *su ai dna-, cf. Skt.
svinn-.
In YAv., d in the word-initial cluster #dm becomes #nm: YAv. nmna- house,
but OAv. dmna- (Skt. ma
na-).
11.14. PIE *b was a marginal phoneme, so much so that the correspondences
(and the words) that contain it are very scarce and controversial. The usual
source for Avestan b was the PIE voiced aspirate *bh (> Ir. *b > Av. b): YAv.
barn (3.pl.pres.act.inj. of bar), Skt. bhran, cf. OP abaran, Gr. , from the
PIE root *bher bear; OAv. aib to, towards, Skt. abh; YAv. upa.skabm pillar,
fra-scibana- beam, cf. Skt. skambh- support.
The phonetic development of certain clusters is a source of new, non-
etymological instances of b: 1. in the word-intial cluster OAv. db- / YAv. t b-, orig-
inating from the sequence < *du i-: cf. OAv. daibi hate (with anaptyxis and
epenthesis; cf. daibit 3.pl.pres.act.ind.), YAv. t bi, Skt. dvi (cf. dvi), cf. Gr.
fearful. However, *du > OAv. duu: Av. duuaah- hostility, Skt. dvas-.
In YAv., one also finds b- as an outcome of the cluster *du i-: bitiia- second ver-
sus OAv. daibitiia- and Skt. dviti
ya-. 2. The Ir. cluster *ju (< PIE *

(h)
u ) likewise
becomes zb in Av. (and other Ir. dialects): zbaiieiti (3.sg.pres.act.ind. of z call;
Skt. hvyati < PIE *heu H); zbarmna- (mid.part. to zbar move crookedly; cf.
Skt. hvrate < PIE *hu er).
3. In YAv., b fricativizes to and further to uu, following the formula-
tion in 11.1.4, and the exceptions mentioned under 11.13 also apply: redu-
plicated formations (bbuuar, etc.) and compounds with preverbs (brtm,
etc.).
4. On the change of *u to after a dental stop, forms such as Av. rtar-
creator versus Skt. tvar- may be adduced; see two further examples under
11.17.1.
11.15 Fricatives
Avestanpossesses a series of voiceless fricative consonants, x, , and f, which
derive from Ir. *x, *, *f, resp. This series did not exist in Indo-Iranian, as it
developed during the Iranian period due to changes in certain sequences of
consonants. On the one hand, these fricatives come from voiceless stops when
they preceded another consonant, thus TCsequences; on the other hand, these
fricatives come out of the voiceless aspirates of Indo-Iranian, whose origin lies
in sequences of voiceless stop + laryngeal, thus *TH.
11.16. Av. x thus comes from of Ir. *x < PIE *k, *k
u
before C: Av. xratu-, Skt.
krtu-, cf. Gr. adj. ; YAv. huxratu-, Skt. sukrtu-; YAv. baxta- (ppp. to baj
11 consonants 31
apportion, with IIr. devoicing of *g to *k before a voiceless stop; but cf. 11.10.2.
and fn. 11 above), Skt. bhakt-; Av. haxman-, Skt. sakman-, from PIE *sek
u
follow. The same root may serve to demonstrate the development of Ir. *x from
the sequence *kH / *k
u
H: YAv. haxa follower (dat.sg. has
e), Skt. skh (dat.sg.

skhye) < PIE *sek
u
h
2
-oi ; cf. Lat. socius as well.
Some cases of non-etymological x, which have sporadically developed be-
fore -, may be noted: xn- < *n-: Av. xn- know, OP xn, Skt. j < IIr. *j n- <
PIE *n-; xuua six, Skt. .
On x and xv, cf. the paragraphs corresponding to the particular develop-
ments of the sibilant in 11.28 and 11.30, resp.
11.17. 1. Av. comes from Ir. * < PIE *t before C: YAv. raiias-ca three, Skt.
tryas, Lat. trs; car four, Skt. catva
ras, Doric Gr. < PIE *k

u
tu or-es;
Av. m you (acc.sg.), Skt. tva
m; Av. mriiiu-, Skt. mr ty-.

2. After x and f, Iranian has undergone voicing, becoming : OAv. vaxra-
mouth (Skt. vaktr-); Av. uxa- word (Skt. ukth-); YAv. vauuaxa (2.sg.perf.
act.ind.; Skt. uvktha); fri (from ptar-); OAv. rafra- (from rap). On p, cf.
11.10.3.
3. In the same fashion, Av. comes out of Ir. * as the outcome of the
PIE sequence *tH: Av. pru-, Skt. pr th-, Gr. < PIE *pl th
2
-; Av. raa-
chariot, Skt. rtha- <*rth
2
-o- (cf. Lat. rota<*roteh
2
). The inflectionof the word
for path, YAv. pata
beside Av. pa, is illustrative; cf. ch. 3, fn. 2.

4. In spite of regular fricativization, the original t is preserved in some cases:
following s (thoughthe t is preservedinword-final positiononly if i or uprecede
the st-clusterelsewise, the t disappears; cf. 11.10.1): OAv. astuuat- bone, cf.
Skt. asthanvnt-; YAv. gauustriia- < *u strii a-; YAv. hastra- gathering, meet-
ing from had sit, Skt. sattr-. In final position, cf. OAv. vs (3.sg.aor.act.inj.
< IIr. u nst to van); OAv. ss (to sad resemble), etc., but YAv. nist, mist
(3.sg.aor.act.inj. to nid insult and mi), OAv. uraost (3.sg.pluperf.act.inj. to rud
moan, wail).
t is also preserved after : fraaotra- PN (which contains the word utra-
camel, Skt. ra-); YAv. pitra- bruise (cf. the Skt. root pe crush, Lat. pns);
YAv. xtuua- sixth. In contrast to st-clusters, t also remains in word-final posi-
tion after regardless of the preceding segment: OAv. tt, drt, cit (3.sg.aor.
act.inj. to ta fashion, dar, and ci, resp.). On the development of the cluster *ts
> Av. s, see 11.11.2.
11.18. Av. f comes out of Ir. *f < PIE *p before C: cf. the Av. preverb fra, Skt.
pr, Lat. pro; YAv. afnavhat- landowner, cf. Skt. pnas-, pnasvant-, PIE
*h
2
ep-nes-. p is preserved before t, but not in the sequence ptr; cf. 11.10.3.
One unexpected source of YAv. f is IIr. *bh in the word nfa- navel, cf. the
gathicism in YAv. nab-nazdita- close relative [i.e., the closest with respect to
the navel], Skt. na
bh-ndiha-.
11.19 Sibilants: s, z, , s , s
,
The PIE sibilant *s, in the course of its history going into Avestan, has
undergone various changes, which, in some cases, have profoundly altered its
nature. To the inherited s (along with the allophonic variants z and perhaps
?), other sibilants that have been created later are added. A list of the Avestan
sibilants with indications of their possible origins is given below.
11.20. Av. s arises from various sources. Firstly, 1. it comes from PIE *s > Ir.
*s when it occurred before a stop (or before the nasal n, but not before m, cf.
11.26.2): Av. skda- (probably Skt. skndha-); OAv. skiti- community (from
+hac; cf. Skt. a
sk-ra-); OAv. scat(3.pl.aor.act.impv. tohac); Av. ststand (Skt.

sth, Gr. , Lat. st); Av. asti
(3.sg.pres.act.ind. to ah; Skt. sti, Gr. ); OAv.

vast (3.sg.pres.mid.ind. to vah; Skt. vste); OAv. kas-n (nom.sg. of ka- + particle
n; see further OAv. k , Skt. ks); Av. ssn- lesson (cf. sh teach, Skt. s); Av.
sparz aspire to (Skt. spr h).
2. Av. s can also derive from an excrescent s inserted between two dentals
in the sequence *-tt- > *-tst- > -st- (cf. further 11.24.3): Av. ustna- stretched
(out) (Skt. uttn-); OAv. vist (2.sg.perf.act.ind. of vid; Skt. vttha; Gr. );
and, from the same root, vista- (ppp.; Skt. vitt-). For the change of *-tst- to -st-,
compare the outcome in other clusters such as *matsia - fish > YAv. masiia-
versus Skt. mtsya-; cf. further Av. hs (pres.part.act. of ah to be) < IIr. *Hsa-
nt-s). An outcome -zd- arose in combination with a voiced aspirate plosive, cf.
11.11.2.
3. PIE * > IIr. * > Av. s: Av. dasa
ten (Skt. da, Lat. decem, Lith. deim-ts

< PIE *dem ); YAv. satm hundred (Skt. atm, Lat. centum, Lith. im tas <
PIE *dm tm); Av. aspa- (Skt. va-; cf. 11.10.3); YAv. asman- stone, sky (Skt.
man-, Gr. ); YAv. sraiiah-, OAv. srata- (comparative and superl., resp.,
to srra-); Av. sru hear (Skt. rav, Gr. , Lat. clure); Av. sstar- master (Skt.
star-); Av. sh teach (Skt. s).
4. The PIE sequence *s developed to *s in IIr. (and in ruki environments
[11.20.5] in Ir. to *c), in Ir. to *sc, and finally, in Av., to s: YAv. jasaiti (3.sg.pres.
act.ind. to gam; Skt. gchati); Av. prsaite
(3.sg.pres.mid.ind. of fras; Skt.

prchte < PIE *pr -se/o-; Lat. posc, precor).
5. A group of segments that affected PIE *s, inducing phonetic changes
probably already in the Indo-European period, makes up the so-called ruki
environment. This term refers to the group of phonemes i
r k (as well as
11 consonants 33
their respective variants i u r , K) before *s, and the consequent articulatory
change of *s to *. The effects of ruki may be observed in all of the satm
languages (11.7), such as the Indo-Iranian and Balto-Slavic subgroups of Indo-
European.
Examples include: the ending of the inst.pl. -bi
; Av. as i prize; YAv. pitra-

(Skt. pe, Lat. pistus); YAv. va- poison (Skt. vi-); YAv. zuta- enjoyed (Skt.
ju-; cf. Lat. gust); YAv. arti- spear (Skt. r -); Av. ratu span, time; Av.
tanu body; OAv. drt (3.sg.aor.act.ind. to dar); pn- heel (cf. Skt. pa
ri-,
Gr. , Lat. perna); OAv. nr, YAv. nar (gen.sg. of nar- man); Av. vaxt
(3.sg.aor.act.inj. to vax grow; cf. Skt. impf. aukat; Gr. < *h
2
u eg[-s]);
vaxii (1.sg.pres.act.ind. to vac speak; Skt. vac < PIE *u ek
u
); Av. vai
(2.sg.pres.
act.ind. to vas desire; Skt. vaki to va < *u -si); YAv. mda- reward (Skt.
mh-, Gr. < PIE *mizdh-); OAv. aog (2.sg.pres.mid.ind. < *au gh-sa;
cf. 11.11.1). From these examples, it may be inferred that ruki affected Indo-
Iranian *z in addition to *s.
11.21. Av. comes out of 1. IIr. * > Ir. *c preceding t, dh, and bh: YAv. ata eight
(Skt. aa
, PIE *h
3
eteh
3
); YAv. nata- (ppp. to nas disappear; Skt. na); YAv.
vat (3.sg.pres.act.ind. to vas; Skt. vi; compare the 1.sg. vasm; Skt. vmi);
drta- (ppp. to dars; Skt. dr - to dar). 2. also comes out of IIr. * or *j > Ir.
*c, *j before n word-internally: YAv. frana- question (from fras; Skt. pran-
; cf. OAv. fras- id.); YAv. frnaoiti (from fra+nas carry; Skt. anti); YAv.
franu- ( fra+nu- with the knees forward; compare, for example, nubiias-
cit ; cf. 11.25.4). Some notable exceptions to this development do exist, though
they may be explained by analogy to other forms without the : OAv. vasn
according to my will (inst.sg. of vasna-?) instead of

vana (cf. OP van)
by analogy to the (verbal) stem vas- and vasah-; Av. yasna- (Skt. yaj-) by
analogy to the stem yaz. 3. From *s affected by ruki, cf. 11.20.5 with exam-
ples.
4. The IIr. sequence *s > * becomes Av. : mou-c soon (Skt. maku
, Lat.
mox); Av. vai
(2.sg.pres.act.ind. to vas << PIE*u -si; cf. 3.sg. vat ). 5. Incontact

with a preceding labial, (and cf. 11.25.2) are also found: nafu-c (loc.pl. of
napt- grandson); drafa- banner (Skt. draps- drop) fu livestock < *pu-
< *pu- (cf. pasu-, Skt. pau-, Lat. pecus). 6. also arises from the sequence *t:
taan- shaper, carpenter (Skt. tkan- < *teton-), at (3.sg.pres.act.ind. to i
dwell; Skt. kti, Gr. ) and, fromthe same root, Av. ira- region, dwelling
(Skt. ktra-).
11.22. s is originally an Avestan development of the sequence rt under certain
accentual conditions; however, later in the transmission and manuscript tradi-
tion, s has frequently become confused with , and at times with s
as well. The
origin of and examples for s are laid out in 11.2.3.
11.23. s
is a palatal phoneme. 1. In the manuscript tradition, the interchange of

and s with s
is often encountered, despite the fact that all of these phonemes

were distinct inthe archetype. s
comes out of Ir. *i , whichbecame OAv. i >YAv.

s
/: OAv. s
iiaoana-, YAv. (gathicism) s
iiaona- (< *i au tna- Skt. cyautn-; cf.

PIE *ki eu , Skt. cyav, Gr. ). In YAv., s
does not occur followed by ii (i ), which

is to say that the merger of and ii in a single phoneme s
had already been

completed. However, inYAv., the manuscripts usually donot write the expected
s
, but use the two other sibilant letters: OAv. s
iit (ppp. to s
ii, cf. Lat. quitus)

versus YAv. s
t (vv.ll. iit, t, s t) at ease.

2. See 11.16 for the vacillation between initial x- and - in some words.
11.24. The origin of Av. z is also heterogenous. 1. Firstly, it is the outcome of
the PIE palatovelars *, *h, by way of an IIr. stage *j , *j h > Ir. *j [dz]: Av. zaoa-
pleased (Skt. ja-; cf. Gr. , Lat. gust; PIE *eu s); YAv. zraiiah- sea (Skt.
jryas-); Av. zta- (ppp. to zan generate, Skt. jt- to jani); YAv. zazmi (to z
leave, Skt. jhmi to h); Av. zaotar- priest (Skt. hotr-; cf. Gr. , PIE *heu
pour); YAv. ziia
winter (nom.sg.; cf. Skt. him-, Gr. , Lat. hiems).

2. z is also found as the product of the voicing of *s preceding a voiced stop:
Av. nazdita- (superl. of asna- near, Skt. ndiha- < *nasd); Av. mazd- wis-
dom (Skt. medha
- < IIr. *mas-dhaH- < PIE *mn s-dheh

1
-); OAv. zd (2.sg.pres.act.
impv. toah; Skt. edhi <*azdh; cf. Gr. ); YAv. mazga- (cf. Skt. majjn-; 11.12.b);
OAv. rzdm (2.pl.aor.mid.inj. to r protect).
3. In addition, z arises secondarily from sequences of voiced dental stops
*-dd- > *-dzd-, which result in Av. -zd- (cf. further 11.20.2): OAv. fra-uuiz-
dm (2.pl.aor.mid.inj. of vid); OAv. sazdiii (mid.inf. to sad or sh), dazdi-
ii (mid.inf. to d); or in certain Bartholomae sequences (cf. 11.11.2): Av.
vrzda- (Skt. vrddh- < *vrdh-t-); OAv. azd certainly (OP azd, Skt. addha
).
11.25. Av. also has several possible origins: 1. from the allophone of PIE
*z > IIr. *z affected by ruki (11.20.5, with examples), including compounds
with the pejorative prefix du (Skt. du, Gr. ) plus a second element
with a voiced first consonant: duuuacah- with a bad word (Skt. durvacas-);
dudh- niggardly.12 2. In contact with Iranian labials, is also found (and ; cf.
12 Curiously, du before m was preserved as such, thus the tradition gives du-manah-. The
11 consonants 35
11.21.5): dia- deception; diaidiii (pres.desiderative.inf. to dab deceive
< *di-dbh-sa; cf. Skt. dpsati to dabh); vaaka- wasp (< *u abzha- < PIE
*u obhso-).
3. can also come out of IIr. *j , *j h > Ir. *j preceding t, dh, bh: OAv. vadra-
driver (Skt. vhar-; cf. Lat. ueh; PIE*u eh); OAv. grd (<*grh-ta, 3.sg.pres.
mid.inj. to garz lament; cf. Skt. garh). IIr. *j , *j h also give Av. 4. before word-
initial n: ntar- knower (Skt. jtr-; cf. Gr. ; PIE *neh
3
); nubiias-cit
(abl.pl. of znu- knee, Gr. ); the outcome in word-internal position, how-
ever, is n: YAv. barna (inst.sg. of barzan- height); cf. further 11.21.2.
5. The IIr. sequences *j +s and *j h+s become Ir. *j > Av. :13 YAv. uz-uuaat
(3.sg.aor.act.subj. to vaz; Skt. vakat < PIE *u eh-se-); OAv. ddr (2.sg.pres.
desiderative.act.inj. to darz strengthen < *didr j h-sa-; Skt. darh).
6. The Av. sequence -ji - becomes in YAv.: druaiti (to druj; Skt. drhyati),
but OAv. a-drujiiat-; baat (3.sg.pres.pass. in -iia to baj apportion, Skt. bhaj)
may also be included here, as well as the YAv. stem daa- burn (i.e., a stem in
-iia, like in the YAv. type jaiiia- pray; cf. Skt. dhati).
A YAv. phenomenon that is usually interpreted as a dialectal feature may
be detected in the change of intervocalic j to : snat and the part. snait-
(to sna snow). Contrary to what was formerly believed, this development
usually presents itself in nominal forms (cf. ai-, Skt. hi-; etc.), but is much less
common in verbal forms (only three): sna- (versus Skt. snihyati), nania-
(wash, Skt. nenikt), a- wave; only these three may indeed be explained
as exhibiting this dialectal feature.
11.26 The Fricative h and its Derived Phonemes: x, xv, h, h, vh
1. A first change, which must have taken place in the Iranian period at the
latest, is the development IIr. *s > Ir. *h in initial position and in word-internal
position before sonorants and i , u , m, r.
Av. h comes from Ir. *h < PIE *s in both word-initial and word-internal
position: YAv. hapta seven (Skt. sapt, Gr. ); hac (Skt. sac, Gr. , Lat.
sequor); hauruua- (Skt. srva-, Lat. saluus, Gr. (Ionic) ); Av. ahu- (Skt. su-);
ahura (Skt. sura-); Av. ahi
(2.sg.pres.act.ind. to ah; Skt. si, Gr. ); OAv. kahii

(gen.sg. of the interr. ka-; Skt. ksya); OAv. manahi-c (loc.sg. of manah-; Skt.
single example of OAv.
+
dumanah- ill-minded (Y 49.11; Gr. ) could originate in
the (Persian) vulgate pronunciation.
13 The same development would take place in a potential PIE sequence *d
(h)
h. The YAv.
example ar flow, cf. Skt. kar (krati) and Prkrit jhara to drip, must derive from PIE
*dhg
u h
er-.
mnasi). As the examples make evident, his preservedonly before i andu, while
between other vowels it was nasalized and became h; 11.27.
2. The change to h may also be observed before m: Av. ahmi
(1.sg.pres.act.ind.
to ah; Skt. smi, Gr. ); OAv. hm, YAv. ahma (acc.pl. 1.pers.pron.; Skt. asma
n,
Aeol. Gr. ); Av. dahma- wonderful (Skt. dasm-); gr hma- PN. The graphic
sequence hm in fact represents a voiceless m, which is also written as m in
the manuscripts; cf. 5.2.5. In initial position, h is lost before m: cf. mahi
(Skt.
smsi), or the preposition mat with (Skt. smd).
3. Depending upon phonetic environment, the new h was preserved as such
(see above), or became the new phonemes h (11.27), x (11.28), h (11.29),
xv (11.30), vh (11.31), or simply .
The change of h to is exceptional and occurs solely before r word-medially.
In effect, the PIE sequence *-sr/l- becomes Avestan -r-: ara- evil (Skt. asr-);
dara- experienced (Skt. dasr-). Note that r also takes on other spellings in
the manuscript tradition: hr, gr (a hypergathicism; cf. 11.3.2). Word-initial
*sr- simplifies in OAv. to r-: rma- brutality (cf. Skt. srma- paralytic), but in
YAv. it seems to have become r-: YAv. raot (Skt. srtas-, but OP rautah-).
11.27. h arises from the sequence a
ha
> Av. aha: ahat (3.sg.pres.act.subj.

to ah; Skt. sat); OAv. manah (inst.sg. of manah-); OAv. nmah (inst.sg.
of nmah-); OAv. a
har (3.pl.perf.ind. to ah Skt. sr). Due to intraparadig-

matic pressure, it is not uncommon to find h extended to positions where it
would normally be impermissible: for example, from Ir. *vahu-, the following
paradigm results: nom.neut. voh, gen.sg. vah u (< *u ahau ), and a dat.sg.
vahauu, but also nom.sg.masc. vahu and acc.pl. vah-c. In YAv., h can
also come from IIr. *ns before a
, and is equivalent to OAv. gh in this environ-

ment; cf. 7.9.3.
11.28. x is a secondary product of the new Iranian h. x is found in the word-
initial sequence #x iias the Av. reflex of Ir. *hi -: OAv. x ii m, x iia
, x iit (opt. of
ah; Skt. sya
t), YAv. x iiaona- Name of a people. Inword-internal position, Ir. *hi is

preserved in OAv., but yielded h in YAv. before a-vowels. Before u
, we find x ii in
both varieties of Avestan. Examples: OAv. vahiiah- (comparative of vohu-; Skt.
vsyas-); ahii (2.sg.pres.act.impv. to 2ah throw; Skt. sya); the ending of the
thematic gen.sg. as in madahii (of mada- intoxicating drink); acc.sg. dax iim
country. YAv. hii- as, e.g., in mhiia- PN (Skt. msya-; cf. Av. mh- moon,
month), is found before a disyllabic suffix -i.i a- < IIr. *-i.Ha- belonging to.
Word-internally, OAv. has a variant -x ii- before an a-vowel, found mainly if
the following syllable is non-final or is a word-final closed syllable. The condi-
tions are not completely clear, however. Examples include OAv. nmax iimah
11 consonants 37
(1.pl.pres.act.ind. to nmax iia- do homage, denom. to neut. namah- homage;
cf. Skt. nmas-), xvax iii (dat.sg.f. of the refl.pron. xva- own; Skt. sv-).
11.29. h properly belongs to YAv., and represents the development of -hi - >
hi (i.e., once h had become h): YAv. va h versus OAv. vahii (nom.neut. of
the comparative of vohu-; Skt. vsyas-); da h u versus OAv. dax ii u (gen.sg.);14
a hi versus OAv. ax iii (cf. above). To find h instead of h in the manuscripts
is commonplace.
11.30. xv is a secondary product of Ir. h, and represents the Ir. sequence *hu
in a single grapheme (contrast x ii < *hi ; cf. above): Av. xvafna- / xvafna- sleep
(Skt. svpna-); xva- (reflexive pron.; Skt. sv-).
Word-internally, the treatment of *hu varied, given that OAv. preserved it as
huu, while in YAv. it becomes vh (cf. below). -xv- appears sporadically: OAv.
nmaxvait (nom.pl.fem. devotees). This reflex alternates with the usual OAv.
-huu- in the same way that -x ii- alternates with -hii- (see 11.28). Word-initially,
the regular reflex of *hu - was xv- in OAv. and YAv. Initial huu- is found where
hu- good or huu- sun were restored. Compare OAv. xv g (i.e., /huu h/,
gen.sg. of xvan-; but nom.acc. huuar /hu ar/; Skt. svr), and OAv. xvra-
well-being (hu-ra-); xvti- of good access (hu-ti-); YAv. xvta- (hu-ta-; Skt.
suvit-); xvstra- of good pastures (hu-u stra-; cf. the hydronym xvstr-).
11.31. vh properly belongs to YAv. and reflects -hu -. In Indian manuscripts,
it also appears as uh or h: ysauha, ysavha (as well as ysaha); vauh,
vav (as well as vah); prsauha, prsava(as well as prsaha). YAv. -hu-
reflects the later addition of a vowel to a word ending in loc.pl. -hu: yhuua,
dmahuua (loc.pl. of the rel.pron ya- and of dman- creature, resp., + the
postposition ; cf. 16.8). The sequence -hu - (-huu -) appears as -huu- in OAv.:
OAv. gahuu, dhuu (2.sg.aor.mid.impv. to gu listen and d, resp.; Skt.
-sva).
14 In YAv., a case of preventive dissimilation in the acc.sg. and the gen.pl. of the stem da hu-
is discernable: dax iim (also da haom) and dax iiunm. The presence of m has impeded
the development of the secondary nasal . Asimilar case presents itself in the gen. ah u,
nom. ahu (*ahu), but acc. ahmor invahu (analogical, cf. neut. vohu
): vah u, but
vohm, vohunm (Skt. vasnm).
chapter 3
Morphology
12 Introduction
In this large chapter, we will present the facts concerning nominal inflec-
tion, including sections dedicated to adjectives and their degrees of com-
parison, pronouns, and indeclinable words (prepositions and preverbs). Fur-
thermore, verbal inflection and the formation of verbal stems will be dis-
cussed.
13 Nominal Inflection
The nominal inflection of Avestan is inherited from the inflectional processes
of Indo-European, by way of Indo-Iranian. On the one hand, Avestan exhibits
an archaic system, as it retains many fossilized features that were at one time
productive, but which have already fallen into disuse by the Avestan period.
On the other hand, Avestan nevertheless shows some clear innovations with
respect to earlier linguistic stages, though these innovations are based upon
inherited processes.
As has already been explained above, one of the aims of Comparative Indo-
European Linguistics is to linguistically delimit the Iranian languages and to
establish, using reconstruction, a common linguistic stage shared between
these languages and the Indic languages: Proto-Indo-Iranian (IIr.). Avestan
morphology immediately reveals a close relationship to Vedic, the oldest at-
tested stage of Indic. Vedic thus furnishes indispensable aid in the assessment
and classification of Avestan forms, which are often very difficult to interpret,
given the phonetic nature of the alphabet and the convoluted transmission of
the texts. Moreover, the distinction between OAv. and YAv. further complicates
linguistic and philological work.
13.1 Components of the Word
The linguistic analysis of a word focuses, first of all, on the elements that it
contains, and the objective of the analysis is the classification and definition
of those elements. The first general division of formal elements allows one
to distinguish a root, affix, and ending in each nominal form. An affix may
precede or follow the root, on which basis it is called a prefix or suffix,
40 chapter 3 morphology
respectively.1 At times, the root and suffix have a very tight bond, and the
whole consisting of root and suffix is then termed stem. Where the stem is
coextensive with the root, one speaks of a root noun. Finally, the word, as
is the usual practice for old Indo-European languages, receives endings that
situate that precise formwithin a case/number system(i.e., they place it within
a paradigm); likewise, the endings serve to establish syntactic and functional
relations with respect to other words within the sentence. On many occasions,
it is not possible to divide stemand ending clearly, in which case it is necessary
to speak of a non-segmentable ending, rather than a simple ending.
The usual manner of writing a root is to show it in the FG form with PIE *e,
which, in Avestan, is equivalent to the full grade a. That is to say, for example,
the root *men (Av. man) means think, or the root *u ek
u
(Av. vac) means speak.
Meanwhile, stems are given with a following dash. Thus, for example, we have
the stem *u ek
u
- (root noun) word (Av. vac- voice, Skt. va
k-), or the present

stem *mn i e/o- think (Av. maiia-, Skt. mnya-) to the root *men; etc.
13.2 Accent and Ablaut
Attentive observation of the means of inflection in particular words reveals
two further morphologically relevant elements: the accent, and patterns of
vowel alternation (ablaut), which may have originally been bound up with
one another.
Little can be said about the place of the accent in Avestan, though some
phenomena due to the effects of the accent have been noted in the preceding
chapter. All things considered, those phenomena do not have significant impli-
cations. However, one characteristic feature of Indo-European (nominal and
verbal) morphology that has left a profound trace on the daughter languages
is the system of ablaut. Ablaut consists in the regulated change of particular
vowels (including ) within morphological elements, i.e., in roots, affixes, and
endings. Eachone of these elements cantake onvarious forms depending upon
the value of the ablauting vowel. The system of ablaut, as may be seen in the
Table 4 below, sets up an opposition, on the one hand, between the zero grade,
full grade, and lengthened grade, and on the other hand, in the corresponding
gradations, between e grade and o grade.
The relation between FG, LG, and ZG is termed quantitative ablaut, while
the relation betwen e grade and o grade is termed qualitative ablaut.
1 Cases in which an affix is contained within the root itself are referred to as infixes. Indo-
European languages normally employ infixes only in the domain of verbal inflection.
13 nominal inflection 41
table 4 Ablaut in Indo-European
e Grade o Grade
Full Grade (FG) e o
Lengthened Grade (LG)
Zero Grade (ZG)
The function of the ablaut system within a single paradigm is regulated in
part by the accent, and in part by morphological category. As was already
mentioned, a strict relation between accent and ablaut might have held in the
proto-language. A syllable bearing the accent would then have e grade, while
an unaccented syllable would have o grade or ZG. Later, during a second phase,
the tightly bound relation between accent and ablaut could have been broken
by accent displacement or by changes in allomorphy, such that forms with
multiple e FGs or with ZG throughout arose.
Out of this system, Indo-Iranian, and ultimately Avestan, has maintained
only quantitative oppositions, as a consequence of the developments in the
Indo-European vowel system. However, remnants of the old qualitative oppo-
sitions may still be observed in the effects that certain ablaut grades produced
on the preceding consonant (i.e., palatalization, cf. 11.7, and further 11.9.,
11.12.2), or in the outcomes stemming from Brugmann (7.4.4). Nonetheless,
it must not be forgotten that, in general, these elements are mere traces of a
very archaic older stage, and therefore must be treated with caution.
13.3 Paradigmatic Nominal Categories
In the noun, Avestan distinguishes three genders: masculine, feminine, and
neuter; three numbers: singular, dual, and plural; and eight cases: nominative,
accusative, genitive, ablative, dative, instrumental, locative, and finally, voca-
tive.
In Avestan, these categories are expressed within a paradigm by means of
endings, which simultaneously carry information about case and number, and
sometimes gender. In contradistinction to adjectives and pronouns, the noun
lacks gender as a category in itself, but gender instead derives from the specific
stem (lexeme). In this regard, the distinction of gender does not arise directly
from the endings, save partially in the singular and in the plural of neuter
nouns, but rather is evident in virtue of concord between nouns and adjectives
or pronouns.
In Table 5, the fact that the endings in Avestan express more than one cate-
gory, a characteristic feature of the Indo-European languages, is self-evident.
The table aims to encompass all inflectional types, and is based on the for-
table 5 The nominal endings in their Proto-Avestan form
Singular Dual Plural
m./f. neut. m./f. neut. m./f. neut.
nom. *-S, *-h, *- *-ah
voc. *- *- *-aH *-iH *-H, *-
acc. *-(a)m *-ah, *-nS, *-nh
gen. *-S, *-h, *-ah *-h *-m
abl. *-(a)t
*-bi ah
dat. *-ai *-bi(m)
inst. *-(a)H *-bi
loc. *-i *-au *-Su, *-hu

mal evaluation of each distinct place in the paradigm. Keep in mind that the
endings given there are normally sufficient to distinguish a paradigmatic slot,
thoughpossible variations inthe stemor (infrequent) cases of heteroclisis must
be taken into account.
13.4 Endings
In the process of inflection, the stem is unified with the various endings, each
of which occupies a specific place in the paradigm. The number of endings
is limited and fairly small. In Avestan, the task of enumerating the endings is
rendered more difficult by the tradition, which has transmitted the text with a
markedly phonetic character, and as a consequence it is commonplace to find
that the endings have several allomorphs.
Giventhe impossibility of presenting a table that encompasses all of the allo-
morphs of every ending for the two dialects without losing the synoptic effect,
we have instead opted to present a general table with the Proto-Avestan end-
ings, i.e., those endings that must necessarily underlie each one of the histori-
cally attested endings. Only the standard endings are included in Table 5, and
account is not taken of sandhi phenomena, which will be explained separately.
Likewise, the possible alternations in the stem itself, which will be detailed
below under each particular class, are not considered here. In some cases, the
use of archiphonemes is necessary: S or M, as well as the single laryngeal H,
which probably still existed in Proto-Avestan. Finally, for practical purposes,
realize that differences between Old and Young Avestan also appear in inflec-
tion, as, above all, in the quantity of word-final vowels, already mentioned
above (cf. 7.2): Old Avestan always has -, -, -, -, whereas Young Avestan
always has short vowels: -a, -i, -u, -e.
14 case endings of the singular 43
14 Case Endings of the Singular
14.1. The nominative masc./fem., sometimes called animate, has two inflec-
tional types. 1. The sigmatic type is more common and is characterized by the
addition of an -s to the stem, with allomorphs that occur in virtue of the pho-
netic environment (i.e., ruki, etc.). Root nouns, stems in -t, -t, -i, -u, -a, as
well as a few stems in -n, and isolated cases of r-stems, are all found in this
inflectional type. 2. The asigmatic type is less commonthanthe preceding type,
and is characterized by a zero ending: -. This type includes some r-stems and
n-stems, as well as stems in - and -. Lengthened grade of the suffix occurs in
some of these classes.
14.2. The acc. masc./fem. has the ending -m (or -m).
14.3. The nominative/accusative neuter has, in contrast to nouns of animate
gender, the same form, with a zero ending. In a-stems (thematic stems), the
ending is -m.
14.4. The vocative of the singular is a case form with a zero ending. The voc.
of ablauting stems shows full grade of the suffix.
14.5. The genitive and ablative have identical forms in IIr. (and PIE) in all
inflectional types, with the exception of the thematic inflection, which has two
distinct forms.
1. OAv. preserves the older state, and exhibits an ending *-as (< PIE *-es),
which can undergo ablaut and thus be merely *-s (cf. esp. n-stems, 18.4). YAv.
has slightly modified this situation, and, taking the thematic inflection as its
model, formally differentiates the genitive and the ablative in all classes. The
YAv. gen. follows the old gen. in *-(a)s, while the YAv. abl. takes the ending
-t , adopted from the thematic inflection (19.1.6). To this new ending, the
postposition can be added, thus resulting in the YAv. ending -a
a. On the
development and graphic representation of the ending *-as, cf. 7.9.4.
2. The thematic inflection has, in the gen., an ending OAv. -hii, YAv. -he,
deriving from *-si a (7.10.3), while an ending -at is employed in the ablative.
14.6. The dative has, in OAv., an ending -i, which corresponds to YAv. -e,
both deriving from Ir. *-ai < PIE *-ei . In some cases, the YAv. ending has been
introduced into OAv. (and gathicized) as -; cf. 7.11.1.
14.7. The instrumental has anending -a
, but insome inflectional classes seems

no longer to have any ending. This apparent lack of ending is due to the fact
that the old PIE ending *-h
1
became *-H in Ir., and in word-final position was
lost without leaving any trace after a consonant or lengthening the preceding
vowel (but be aware of the situation regarding word-final vowels in Av. 7.2
and 13.4).
14.8. The locative is another case that has a zero ending (together withlength-
ened grade of the suffix). Other types of formations show an ending -i
.
15 Case Endings of the Dual
15.1. The nominative/accusative/vocative masc./fem. makes for a motley col-
lection, which must be examined under each inflectional type. The PIE ending
that underlies the forms is usually reconstructedas *-h
1
. The Av. feminine forms
are set up as *-ih
1
.
15.2. The nominative/accusative/vocative neut. has an ending -i
, which is
reconstructed as PIE *-ih
1
.
15.3. The dative/ablative/instrumental has -biia
< Ir. *-bi as its ending. On

just a single stem, the ending -biim, identical to the Skt. -bhym, is found: YAv.
bruuat .biim eyebrows.
15.4. The genitive and locative have distinct endings in Iranian, in contrast
to the common endings that they share in Sanskrit. In effect, compared to the
usual ending in Skt., -o, Avestan exhibits, on the one hand, 1. a genitive ending
-a
(< IIr. *-s), and on the other hand, 2. a locative ending -(uu) (< IIr. *-au ).
It is possible that the Sanskrit ending may have arisen as a blending of the two
IIr. case forms as continued in Avestan.
16 Case Endings of the Plural
16.1. The nominative masc./fem. has -, whose reconstructiongoes backtoPIE
*-es, as its ending. Two other endings also exist. 1. An ending that is the product
of morphological recharacterization: -a
h, which must have its origin in the

Indo-Iranian period: cf. Skt. -sas, OP -ha. 2. The thematic inflection has an
ending -a
(in YAv., -a in polysyllables, but -- in the auslaut of disyllables before

16 case endings of the plural 45
the enclitic -ca), which must have been long in IIr., as is evident from -a
in Skt.:
yuga
yokes. This form could derive from the old ending of the collective, PIE
*-h
2
. The existence of an ending that results from the PIE contraction *-o+-es,
as is found in -s in Sanskrit, is not assured. One should note the fem.pl. of the
-stems: -a
< IIr. *-s < PIE *-eh

2
-es.
16.2. The accusative masc./fem. has an ending -, deriving from Ir. *-as < PIE
*-n s. The thematic inflection takes the same ending, but according to the rules
of syllabification, the ending is realized after a vowel as PIE *-ns. The Ir. ending
*-a-ns becomes - g in OAv. and - or - in YAv.; cf. 7.9.3. In some cases,
particular phonetic developments have substantially disfigured the ending, cf.
7.17.7.
16.3. The nominative/accusative neuter has both 1. an ending -i
and 2. a zero
ending with lengthened grade of the element preceding the ending, as a formal
characterization (cf., for example, 18.4). 3. The thematic inflection has an
ending -a
in Av., which can be taken back to Ir. *-a-H. The underlying ending
is PIE *-h
2
> IIr. *-H, which has either become -i, lengthened the vowel, or
disappeared.
16.4. The vocative pl. is formally identical to the nom.acc.pl., for both animate
and the neuter nouns.
16.5. The genitive has an ending -m, coming from *-m. The inflectional
classes ending in a vowel showan innovative ending -nm, made by analogy to
the n-stems. In contrast to the other Indo-Iranian languages, Avestan exhibits
a short stem vowel in these forms.
16.6. The dative/ablative has anending -bii, whichappears as -biias-cabefore
an enclitic. This ending is equivalent to the Skt. -bhyas.
16.7. The instrumental has an ending -b, Skt. -bhi. In some very good
manuscripts, -bi is also found. In certain cases, special phonetic develop-
ments have substantially altered this ending, cf. 11.1.4. In contrast to the
other inflectional types, the thematic inflection shows an ending *-i, Skt.
-ais.
16.8. The locative exhibits assorted allomorphs for its ending, which all devel-
oped from the ending *-su, depending upon the environment in which it falls
(i.e., ruki, etc.): -u, -su, and -hu. Furthermore, the extension of the loc.pl.
through the addition of the postposition (just as in the loc.sg. and the abl.sg.)
is commonplace.
17 Inflectional Classes
Every noun in Avestan can be morphologically classified according to the stem
formant that it possesses. Internal to each inflectional class, further types may
sometimes be distinguished on the basis of particular inflectional behaviors,
the origin of which is normally to be sought in ablaut patterns.
Not every inflectional class has subdivisions, but those considered to be
the oldest from the point of view of Indo-European inflection indeed do; such
classes must therefore be regardedas archaisms inAvestan. The enormous vari-
ation entailed not only by the shifting of the accent, but also by the alternation
of vowels in the constituent elements of the word, could have already been reg-
ularized very early infavor of a single invariable stemwith fixed accent. Of all of
the types that have beenverifiedas part of Proto-Indo-Europeaninflectionwith
greater or lesser reliability, Avestan has preserved but a tiny remnant, which is
clearly disappearing due to regularization.
The two most important types that have been preserved are the proterody-
namic and the hysterodynamic. In the proterodynamic type, the root takes the
accent in the direct cases, while the suffix receives the accent in the oblique
cases. Inthe hysterodynamic type, the suffix takes the accent inthe direct cases,
while the ending receives the accent in the oblique cases. As a general rule,
the direct cases (also referred to as strong) are the nom. of all numbers, the
acc.sg. and du., and the loc.sg.; the remaining cases are oblique (also referred
to as weak). Keep in mind that the accent and full grade are usually linked (cf.
13.2). Other types of inflectional patterns will be commented on under the
corresponding entries.
In general, inflection can be divided into stems ending in a consonant and
stems ending in a vowel, which in some cases are the continuants of old
consonant stems (i.e., laryngeal stems).
The forms given in the tables below belong solely to Young Avestan, except
where expressly indicated to the contrary using italics.
18 Consonant Stems
Consonant stems include an enormous number of stems of diverse nature, but
the endings are added to a consonantal element in all of them. The subgroups
18 consonant stems 47
table 6 The inflection of root nouns
nom. drux p spas
voc. nar as-ca pa asta-ca asti
acc. pm ap vs
gen. drj nara
apm vsm
abl. drujat
aiii vibii
dat. sari ape nrbiia
inst. vrd apa mazib vaibi
loc. sair vsiia n

into which consonant stems may be divided are a. root nouns (with assorted
final consonants) and b. derived stems, which, in turn, include stems in -t
and -at, stems in -an, -man, and -uuan, stems in -ar, and stems in -ah and
-uuh/-u.
18.1 Root Nouns
Root nouns constitute an important and archaic class of nouns, characterized
by the fact that the stem is equal to the root. The final element of the stem is
a consonant (-c, -j, -t, -d, -, -p, -m, -n, -r, -z, -, -h), though some other stems
having a final vowel (-, -, -), which were originally consonantal (< *H), may
be included in this class as well.
The majority of these nouns belongs to the animate class (above all fem-
inines), though some neuters are met with as well. Note further that a good
number of root nouns are the second members of compounds.
Words belonging to this class are: vac- m. voice, hac- following; druj- f.
lie; ast- neut. bone, brt- carrier, stt- f. praise; pa- m. path; pad- m.
foot, vrd- f. growth, zrd- neut. heart; ap- f. water, khrp- f. body; dam- m.
house, zam- f. earth, zim- m. winter, ham- m. summer; span- m. dog, jan-
slayer; kar- who does X, -er, gar- f. greeting, nar- m. man, sar- f. union, star-
m. star; ns- f. wait, vs- f. tribe, village, spas- spy; drz- f. tether, binding,
brz- high, maz- great; - f. vigor; h- neut. mouth, mh- m. moon, month.
Some examples of root nouns terminating in a vowel are: d- f. delivery,
x- f. well, maz-d- m. Mazd, st standing; x- f. lament, j- with life, fr-
pleasing; s- thriving.
N.B. The inflection of these stems is very similar, taken as a set, but at times
it shows some deviations, which may be the result of a special phonetic
treatments or of morphological processes (ablaut). The details of the actually
occurring divergences and possible anomalies will be given in the following
overview.
Singular: 1. The nom.sg. has the ending -s in animate stems, with the condi-
tioned variant -: Av. vx (with lengthened grade; Skt. va
k f.), drux (11.20.5);

YAv. br; YAv. krf (11.21.5). Root nouns ending in a vowel are also sig-
matic: xa
, sta
, mazda
, pata
2 (< *--s < *-aH-s; 7.5.2); r-j, ratu-fr. Root

nouns in a sibilant are sigmatic: YAv. v (Skt. v), spa (11.21.4; Skt. sp), OAv.
dr-c, YAv. bar (11.20.5); ma
(Skt. ma
s). The nom.sg. of root nouns in -n

is usually considered to be asigmatic, in spite of some evidence to the con-
trary: YAv. ja (OAv. hapax vrraj [with vv.ll.]; Skt. ha
) versus ja
, YAv. sp
(Skt. va
). Root nouns in -r are asigmatic: OAv. n.2. The acc.sg. has the end-
ing -m in animate stems: Av. vcm and YAv. vcim (with LG from a FG *o
[cf. the gen., coming from a FG *e]; Skt. va
cam, Lat. ucem), OAv. druj m (Av.

drujm and YAv. drujim; 7.14.4); YAv. brtm, pm (LG, cf. Lat. ps, ver-
sus the FG *o in the pl.; Skt. pa
dam); pm (LG), OAv. khrp m (YAv. khrpm).

Root nouns in -m present some problems, because they seem to lack a formal
marker of the acc. due to the fact that acc. *-mmerged with the *-m- of the stem
before the IIr. sequence *-m-m could have become *-m-am: Av. zm (cf. also
Skt. ka
m), YAv. ziim; janm(Skt. hnam), spnm(7.4.6; Skt. va
nam); Av.
nar
m(Skt. nram), sar m; vsm(Skt. vam), YAv. spasm, brzm, OAv. m;

YAv. ma
hm (cf. 7.5.1; Skt. ma
sam). Vowel stems: stm, Av. mazdm, YAv.

patm(Skt. pthm; also YAv. patnm); yauua-jim, sum.3. The voc.sg.
has a zero ending: YAv. nar (with FG); ending in a vowel: OAv. mazd.4. The
nom./acc.neut. has a zero ending: YAv. as-ca (hapax < *ast-ca; more frequent
is astm [!])5. The gen.sg. has the ending (< *-as, but cf. 7.13.1): YAv. vac,
drj; brt, ast, Av. pa (cf. fn. 2 below); a
p (Skt. aps). Among root nouns

in -m, a genitive ending with ZG may be found: OAv. d g (< *daNs; 7.9.3), in
a formation very similar to the proterodynamic genitives of n-stems (18.4);
cf. also OAv. ads (< ds?, abl.), versus Av. zm (Skt. jms) and YAv. zim
(andzm); n (Skt. ghns), sn (Skt. nas). Traces of ablaut remaininroot
nouns in -r: cf. gar versus nr (YAv. nar; Skt. nars); old and similar to the
r-stems (18.5) are OAv. sar (also sar ), YAv. vs (Skt. vis), OAv. maz (abl.),
2 The inflection of this noun in Avestan seems to be composed of two suspiciously similar
stems: pa- and pat-. Both stems are explicable through sound change. The stem that
underlies the Avestan stemwould be PIE *penth
1
- or *ponth
1
-, whose Av. nom.sg. pata
comes
fromIIr. *pntaHs; cf. Skt. pnths (where the th is secondary by analogy to the oblique cases),
while the gen.sg. pa comes fromIIr. *pn tHs; cf. Skt. paths. The exact PIEetymology of this
noun is disputed.
brz; YAv. ma
h, OAv. a
h (Skt. ss). Vowel stems: Av. ta
, mazda
.6.
The abl.sg. is, in Old Av., identical to the genitive, but in Young Avestan, it has
replaced*-s/hby -t following the thematic inflection(19.16.): pat , zma
t (also
zma); vsat ; nat .7. The dat.sg. has as its ending (< PIE *-ei ) OAv. -i and
YAv. -e, the latter reintroduced into OAv. as - (14.6): brte; ape (apa-cit );
ne (Skt. ghn), sne (Skt. ne); gair, nari (YAv. naire; Skt. nar), sari; vse
(Skt. vi), mazi. Vowel stems: ti, te; r-jiii.8. The inst.sg.: vaca (Skt.
vca
); zrd-c, paa; apa (Skt. apa
), khrpa; zm, hama; vsa (Skt. vi),

drz; OAv. a
h (YAv. a
ha; Skt. sa
).9. The loc.sg. has the ending -i
, to
which the postposition can be added. In some cases, a locative with a zero
ending, but formally marked through ablaut, is detectable. Examples: pai;
khrpiia; dm(< *dm, YAv. dmi), zm (< *-ai; Skt. jmay-a
, Gr. ) and zmi

(Skt. kmi); vsi (Skt. vii) and vsiia with postposition.
Dual: 10. The nom./acc./voc.du.: brta; pa, pa (Skt. a
p); spna (Skt.

va
n).11. The nom./acc./voc.neut.du.: asta-ca.12. The gen.du.: nara
(Skt.
nr).13. The abl./dat./inst.du.: vaibii-ca; nrbiia.
Plural: 14. The nom.pl. has the ending - < *-as < PIE *-es. Although the
nom.pl. is considered a direct case while the acc.pl. is considered oblique, it is
not infrequent to find an acc.pl. that is formally a nom.pl. However, both cases
are clearly differentiated in those stems which show ablaut: vc; brt; p
(Skt. a
pas); zm (Skt. ka
mas), zim; jan (Skt. hnas), spn (Skt. va
nas);
nar (Skt. nras); spas; ; ma
h; vowel stems: xa
, ta
, patn; ratu-frii.
15. The acc.pl. has the ending - < *-as < PIE *-n s: vac; pa, pa. Among
labial-stem root nouns, the OAv. hapax ap (with FG as in Skt. aps) stands
out in contrast to YAv. p (with generalized LG), khrpas-ca; zmas-ca; jan
(Skt. ghns); gar, nr (Skt. nr
n); vs (Skt. vas), , ma
h. Vowel stems:
yauuajii.16. The nom./acc./voc.neut.pl.: asti.17. The gen.pl.: astm,
pam; apm (Skt. apa
m), khrpm; zmm-ca; nm, sunm (Skt. nm);

narm (Skt. nara
m), str m-c (YAv. strm); vsm (Skt. visa
m); vowel stems:

xm.18. Theabl./dat.pl.: aiii(ap-; Skt. adbhys); nrbiias-c(Skt. nr bhyas)
as well as YAv. nrbii, nruii (11.1.4); vibii (Skt. vibhys), vibii (Skt.
vgbhys). Vowel stems: yauuajibii.19. The inst.pl.: azdib, padb (pa-),
garib, mazib, vaibi (Skt. vgbhi).20. The loc.pl.: vowel stems: ratu-
friu.
18.2 Stems in -t, -ac
Among the dental stems, there are some nouns with the suffixes -t- and -tt-,
used to create abstract nouns, such as amrtatt- f. immortality, hauruuatt-
f. wholeness, perfection, yauuatt- f. eternity. Ablaut is found in the inflec-
tion of some nouns: cf. napt- m. grandson, and the directional adjectives in
*-anc,3 derived from adverbs, such as apa
c- apart, fra
c- forward, niia
c-
downward.
The inflection of these invariable stems can be inferred fromTable 6 in 18.1.
The forms of napt- are nom.sg. napa
(*napts), acc.sg. naptm, gen.sg. napt,

and loc.pl. nafu (*naptsu).
For the *ac-stems, the following forms may be adduced: nom.sg.m. ap <
*apa-ank-, fr, acc.sg. m. niia
cim (with analogical *, cf. Skt. nycam), the

inst.sg. apca (Skt. apca), and the nom.pl.m. niia
c.
18.3 Stems in -nt
This formant is found in the present active participle in -at, with an invari-
able stem in the case of thematic verbs: jasat- coming, jaiiiat- praying.
Athematic verbs, meanwhile, exhibit ablaut: surunuuat-/surunuuat- listen-
ing, ht-/hat- being. Possessive adjectives in-uuat-/-uuat- (IE*-u ent-/*-u n t-)
and -mat-/-mat- (IE *-ment-/*-mn t-), and some adjectives in -at, are to be
included here as well: brzat- high, OAv. drguuat-, YAv. druuat- wicked,
fumat- having cattle.
With regard to inflection, note that the athematic nom.sg.m. ends in - in
YAv., which must be explained from an ending *-as < IIr. *-at-s. In thematic
stems, the ending *-ant-s is reflected in YAv. as - (- when denasalized) and in
OAv. as -s. While YAv. - is the outcome of the phonetic development of IIr.
*-ans (cf. 7.9.3, 7.17.3, 19.1.16), OAv. -s reflects the analogical restitution of
the stem -ant from the oblique cases. Keep in mind that, in these stems, the
masc. and fem. nom.acc.sg. are considered direct cases (cf. 17), but the neut.
nom.acc.sg. inflects as an oblique case, cf. cuuat how much. In *-mant- and
*u ant-stems, analogy to the stems in -uuh/-u has induced the substitution of
the nom.sg. in *-u ant-s (e.g., cuus howmuch) by *-u s and *-ms (e.g., astuua
bone ( material), xratuma
intelligent). Note similarly the voc.sg. druu <

*drugh-u as wicked.
18.4 Stems in -an, -man, -uuan, -in
The suffix -n-, originally individualizing in function, occurs directly affixed to
the verbal root (e.g., taan- shaper, creator, Skt. tkan-), but also in the com-
pound suffixes *-man, *-u an, *-Han, or -in. The last two have possessive value
and almost always form adjectives. Some examples are caman- sight (Skt.
cak see), nman- name, as auuan- truthful (fromas a- order, truth), mrn-
knowing the mras (*mra-Han-), kainn- girl (*kani-Hn-), parnin- with
3 With the PIE suffix *-h
3
k
u
- looking towards; the nasal was introduced in Indo-Iranian.
feathers (from parna- feather), zruuan- time (from PIE *r h
2
-u en-, cf. Gr.
old man).
With the exception of the in-stems, all of these stems have an ablauting
inflection, which can considerably alter the appearance of the word. The
nom.sg. regularly has LG and loses the final nasal: OAv. uruu, YAv. uruua
spirit, kaine (Skt. kan
i
ya
). The other direct cases show an alternation between

FG (acc.sg. asmanm sky, nom.pl. as auuan) or LG (acc.sg. tanm, nom.pl.
martn mortals) as the result of Brugmann (7.4.6) or of *H.
The acc.sg.neut. has ZG (nma name). In the gen.sg., the endings *-n-as
(> -n) and *-an-s (> - g, -): OAv. mrn, YAv. tan; OAv. cam g, YAv.
dmn(*dm), zr(7.17.3). The abl.sg., besides the forms that are identical
to the gen., shows other innovative forms, such as YAv. camanat or barsmn
sacrificial straw, which exhibits the YAv. substitution of original *-an-s by *-an-
t > *-nt > -n. There are also ablaut variants in the nom.acc.pl.; furthermore,
a hypercharacterizing *-i may be found after the ending. Thus, the following
endings occur: -m < *-m (after labial consonants) *-n (cf. Skt. krm):
OAv. anafmm, dmm creatures (cf. 11.3.3); -ni (cf. Skt. krm ): OAv.
afmn, YAv. cinmni desires (cf. the types such as OP taumani, Lat. nmina,
OCS imena); -an (originally dual? cf. Skt. du. krm ): OAv. nm n names.
Note also that as auuan- has a lengthened presuffixal vowel in some cases
(cf. Skt. r ta
van-, OP rtv < *r ta
u an-), especially in OAv.: gen.sg. OAv. as un

(*art-un-as) versus YAv. as aon (*arta-un-as; made by analogy to the short
vowel of the nom.), but voc.sg. YAv. as um (*art-un). As for stems in -in (cf.
Skt. hastn- with hands), the nom.sg. has - (from *-n), while the other case
forms show the regular stem in -in.
18.5 r-stems
This group is composed of kinship nouns in -(t)ar, nomina agentis in -tar, as
well as a few other isolated stems. Examples: pitar- father, mtar- mother,
brtar- brother, xvahar- sister, zaotar- priest (Skt. htar- pourer), nar-
man, star- m. star, tar- neut. fire (on root nouns in -r, cf. 18.1).
The nom.sg. has LG and is asigmatic. The final *-r seems to have been lost
already in the IIr. period: xvaha, Skt. svs. The nom.sg. tar constitutes an
exception, both because it is a sigmatic nominative and because it has an
acc.sg. with ZG of the suffix, YAv. trm. In the animate nouns, the direct cases
show the same alternation between FG and LG of the suffix that is seen in the
n-stems, which can be attributed to the effects of Brugmann (7.4.6), e.g.,
pitarm versus dtrm. The two distinct forms of the gen.sg. in *-r-as and in
*-ar- (< *-r- > Skt. -ur), alongside the abl.sg. in *-t , nrt from the man, are
also found here. On the acc.pl. in *-r-n (nr s, str ), cf. 7.7.
18.6 r/n-stems (Heteroclites)
These stems make up a small number of neuter nouns that are usually called
heteroclites because they alter the final segment of the stem throughout their
inflectional paradigm: a stem in -r in the nom.acc. and a stem in -n in the
oblique cases. This class goes back to Indo-European; cf. Gr. , (<
*-n tos) water, Lat. iecur, iecinoris liver. Nouns belonging to this class include
huuar sun, eye, aiiar day, yr year, yakar liver (though of the latter two,
no n-stemforms are attested inAvestan), but also nouns derivedusing the stem
*-u er/u en-, such as anuuar bow (Skt. dhnvan-), karuuar section of the
world (kar cut).
Some noteworthy forms of the paradigm follow here. The gen.sg. in *-an-s
gives different outcomes ineachvariety: OAv. xv g <*huu anh4 (cf. 11.30), YAv.
h < *huu anh (7.17.3), aiin (v.l. of aii) < *ai anh (7.9.3). Intraparadigmatic
analogy (cf. the nom.sg. IIr. *suHar) resulted in the YAv. gen.sg. hr. Various
formations occur in the nom.acc.pl.: 1. with LG of the stem terminating in -r
especially in OAv.: aiir , saxvr . This formation is very archaic, judging from
the Anatolian evidence: Hitt. uttr words; 2. in -n: aiin, karuun; and 3. in
*-ani (< PIE *-enh
2
): OAv. sxv n, YAv. bauuani.
18.7 h-stems
This class includes the PIE neuter *s-stems and adjectival compounds based
on s-stems, e.g., auuah- aid, manah- spirit, vacah- word, zraiiah- lake, sea,
humanah- benevolent, of good spirit, anaocah- unaccustomed to. This class
also includes comparative adjectives formed with the suffix *-i ah- (cf. 20.1):
vahiiah- comparative of vohu- good, as well as some isolated nouns: uah- f.
dawn, biiah- m. fear.
The suffix has FG in in all case forms, apart from the nom.sg. of animates
and the nom.acc.pl. of the neuter, which end in *-h, e.g., nom.sg.masc. ua
,
OAv. vax iia
, YAv. va ha
better (versus Skt. vsyn; cf. further 18.8), nom.acc.pl.

neut. raoca
lights. The nom.sg.neut. ends in *-ah > - /-: OAv. vac , OAv./YAv.
vac (Skt. vcas); OAv. nm , YAv. nm (Skt. nmas), etc. On the development
*-ah > - , cf. 7.9.4.
An ending *-hi is attested in the nom.acc.pl. neuter, e.g., OAv. varc.hc.
The animate acc.sg. shows various ablaut grades: *-ah-am and *-h-am, which
are once more explained by Brugmann (7.4.6): ua
hm but va hahm.
4 This noun comes from PIE *seh
2
-ul-/*sh
2
-u en-, i.e., a heteroclite in -l/n-.
19 vowel stems 53
18.8 Stems in -uuh/-u, -i, -u
The perfect active participle was formedwiththe ablauting Indo-Europeansuf-
fix *-u s/-u os/-u, fromwhich both the nom.sg.m. (IE*-u s) as well as the other
direct cases (IE *-u os-) come to have persistent LG in Avestan, due to Brug-
mann(7.4.6). The oblique cases have the ZG-u-. The motionfeminine shows
-u--; 19.3. Some examples are vduuh-/vdu- to vid know or vauuanuuh-
/vaonu- to van win, be superior. The nom.sg.m. does not have the nasal that
is characteristic of the perf.part.act. of Skt.: OAv. vduua
, YAv. vuua
(Gr. )
versus Skt. vidva
n (cf. also 18.7). Late forms of the nom. of the sort YAv. zazuua
which has left behind could have arisen by analogy to the n-stems (18.4). The
acc.sg.m. is YAv. vuua
hm < *-sam, in contrast to Skt. vidva
sam.
A small number of neuter nouns in -i and -u have an invariable stem
throughout the paradigm, e.g., tuui- force, hadi- seat, ardu- wound.
19 Vowel Stems
19.1 a-stems
This inflectional type is usually referred to as the thematic inflection, on
account of the fact that the stem ends with the thematic vowel -a- < PIE *-e/o-.
The thematic inflection is extremely common, and includes a large number of
masculine and neuter nouns, as well as adjectival forms in those genders. Some
masc. nouns belonging to this class are Av. ahura- lord, YAv. daxma- morgue,
Av. mas iia- mortal, Av. mra- formulation, yasna- sacrifice, YAv. vra- man,
Av. zasta- hand, YAv. haoma- haoma. Some neuters are Av. as a- truth (sg.
only), Av. uxa- word, Av. xara- dominion, power, Av. mira- contract; OAv.
s
iiaoana- (YAv. s
iiaona-) act, deed. Some adjectives are Av. aka- bad, Av.
ams a- immortal, Av. vspa- all, hauruua- whole.
The inflectionof thematic nouns is significantly different fromthe inflection
of consonant stems, and it is common to find endings of pronominal origin.
Recall that the inflection of neuters differs only in the nom./acc., and is identi-
cal to the masculine in all other case forms.
N.B.: Singular: 1. The nom.sg. was OAv. - (< *-ah < *-as; cf. 7.9.4, 7.13.1), but
in YAv. this ending further developed into . When followed by an enclitic,
the older form of the nom.sg. is preserved in both varieties: kas-ci
t , gaiias-c,
yas-ca
, yasnas-ca, etc.2. The voc.sg.m/n. has an ending -a
: OAv. ahur, YAv.

ahura (Skt. asura); neut. OAv. as .3. The acc.sg. has the ending -m in YAv.
The original form of the OAv. ending is - m, which came to be replaced by
-m in many instances: cf. doublets such as OAv. d m and dm-ca (to
table 7 The inflection of a-stems
nom. ahur
ahura
h mas iia
voc. ahura
iiaoani ux
as m ahura
acc. ahurm saite mas ii g xara

mas m zast haom
gen. ahurahii ahurahe vraiia
iiao(a)nanm
abl. as t yasnibii
dat. ahuri zastibii ahuraibiia ams aibii
inst. ahura
iiaoni
loc. yesne
zastaii vspa daxmauua

the adj. da- suitable), OAv. part m, YAv. partm (ppp. to fras ask). On the
other hand, the vowel of the ending could undergo complete assimilationand
contractionwhen incontact with a preceding i (7.14.4) or u (7.16.1), i.e., -iim
> -m: OAv. paouruum (YAv. paoirm; cf. 8.2.), Av. mas m, Av. yesnm (but cf.
anii m, acc.sg.m of aniia-); or in turn -uum > -u
m: YAv. haurm, YAv. riu
m,
YAv. drm. As part of this particular development, account must be taken
of the endings -ai am and -au am themselves (cf. 10.1.2, 10.3.3, resp.): gam
(acc.sg. of gaiia- life; Skt. gyam); YAv. graom (acc.sg. of grauua- cane); simi-
larly, of the endings -i am and -u am (cf. 10.5.2, 10.6.2, resp.): OAv. humm
(acc.sg. of humiia-), YAv. nasum (acc.sg. of nasu- cadaver < *nasu m).
4. The nom./acc.neut is formally identical to the acc.sg.masc., and thus is
unique within the paradigm, since it bears a formal mark of its gender.5.
The gen.sg. derives from a PIE ending *-osi o > IIr. *-asi a > Ir. -ahi a, which has
its regular outcome in OAv. as -ahii and in YAv. as -ahe (7.10.3). On the OAv.
change experienced by the gen. ending before an enclitic, cf. 11.28: as ahii,
sptahii versus as ax ii-c, spntax ii-c.6. The abl.sg. is distinct from the
gen.sg. only in the thematic inflection, while in all other inflectional types,
both cases exhibit the same form. OAv. has preserved this distinction, with
an abl.sg. ending -t , which is identical to Skt. -t: Av. drt from afar, Skt.
dra
t. YAv. also possesses the abl. ending -t , but note that this ending has
been analogically extended to other inflectional classes in YAv.; cf. 18.1.6,
19.2.5, 19.3.5, etc. Preceding the enclitic ca
, the abl.sg. ending appears as -at -ca
in the majority of cases. The fact that this development does not occur pre-
ceding the enclitic ci
t (as t ct , ahmt cit , atahmt cit ) suggests that -at -ca
arose from an effort to distinguish it in pronunciation from the ending --c.

In many cases, a shortening is observable in YAv. of -t preceding the preposi-
19 vowel stems 55
tion haca: the two would form a single word, and shortening of the antepenul-
timate syllable would result (cf. 7.3.1): nmnat haca, vhrkat haca, sptat
haca.7. The dat.sg. has the ending -i (< *-i < *-o+ei ) and exhibits slight
dialectal variation. For its part, the OAv. ending can have an enlargement in
(which is not the postposition that commonly appears in the loc.; cf. 16.8,
and below 9 and 21): ahuri. (from *ahurii, later separated during trans-
mission), ytii (to yta- prayer; cf. further ch. 2, fn. 5). Similarly, in Sanskrit,
the ending -ya occurs; it must be connected with the OAv. forms. Meanwhile,
YAv. shows only the ending -i.8. The inst.sg. has the appearance of a zero
ending, on account of the fact that the PIE ending *-h
1
becomes IIr. *-H, and
in Avestan the length of word-final vowels is subject to phonological condi-
tions depending upon the variety (7.2).9. The loc.sg. had, in the thematic
inflection, an ending -i, which fused with the thematic vowel: PIE *-oi > IIr. *-
ai : cf. Gr. versus nom.pl. . This ending has various outcomes in Av.:
OAv. shows -i and -: xari, s
iiaoani, xvar. The ending properly belong-

ing to YAv. is -e, which was introduced and adopted in OAv. (10.2.2): s
iiaoane.
Before enclitics, the outcome a is found (10.1.1): aspaca, vspaca
. The loc.
can take the postposition : OAv. xari., xvrii, YAv. zastaiia, nmnaiia-
ca.
Dual: 10. The nom./acc./voc.m. has an ending -a
, which comes from Ir. *- <

PIE *-o-h
1
: OAv. spad, YAv. zasta; cf. Skt. hst, deva
(but in Skt., an ending -au

also exists), Gr. .11. The nom./acc./voc.neut. has anending OAv. -i, YAv.
-e, which comes from IIr. *-ai < *-o-ih
1
: OAv. s
iiaoani, YAv. saite, hazare (Skt.

at, sahsre), (f.) uiie (Skt. fem./neut. ubh 10.2.3).12. The gen.du. has an
ending -aiia
< *-ai -s, similar to Skt. -ay-o < *-ai -au , but in both the gen.du.
andthe loc.du., Avestanvery likely has archaic endings whichhave beenleveled
into a single form in Skt.: OAv. saiia
(to sa- part; Skt. m sayo); rnaiia
(to
rna- ?); YAv. diraiia
, vraiia
; meanwhile13. the loc.du. has an ending -aii

< -ai -au : OAv. zastaii, ubii (Skt. hstayo, ubhyo), YAv.
+
uuaii.14. The
abl./dat./inst.du. comes out of an ending *-ai bi > OAv. -ibii, YAv. -aibiia:
zastibii (Skt. hstbhym), ubibii (Skt. ubha
bhym), rnibii, aspaibiia,

gaoaibiia (and gaoae), maaibiia. We also find the ending -biia in YAv.
pnbiia, dirbiia.
Plural: 15. The nom./voc.pl. of the thematic inflection in Avestan cannot,
as in other IE languages, be derived from an ending *-s < *-o+es, but rather
must be related to forms of the type Lat. locus (pl. loca), that is to say, reflect
an old collective ending *- < *-eh
2
: Av. mas iia
. The IIr. ending *-s < PIE

*-o+es would have given -a
in Av. (cf. 7.5.2), but its presence in the texts

cannot be ascertained. Nevertheless, a recharacterized ending -a
h < *-sas
(> OP -ha), similar to Skt. -sas, can be found in various instances: OAv.
zauutiia
h, vspa
h, (voc.) mas iia
h, Av. ahura
h.16. The acc.pl.

shows a number of endings, due to some specific sound changes. The OAv. end-
ing is - g < *-ans < PIE *-o-ns, while that of YAv. is - , but - after a nasal (cf.
7.9.3), though both forms of the ending are occasionally extended to other
environments. Before enclitics, an ending -s is found: OAv. mrs-ca, Av.
mas iis-ca, yasns-ca
, YAv. haoms-ca. In YAv., - s also occurs before encli-

tics: ams s-ca, cir s-ca, vsp s-ca. Curious is the YAv. form aiia
s-cit, with
-a
s for -s. In addition, it is common in YAv. to find forms of the nom.pl.

used in place of the forms of the acc.pl.17. The nom./acc./voc.neut.pl. has
an ending -a
< PIE *-eh

2
: OAv. s
iiaoan, YAv. s
iiaona, Av. yesnii-c. Forms

with an ending -a
, explicable as analogically borrowed from neuters in -ah,

also exist.18. The gen.pl. has the ending -anm, deriving from *-nm (with
analogical shortening of the first ). The expected ending is found in YAv.
mas iinm, probably preserved through an effect of the -i -. Finally, in con-
trast to the reconstructable ending *-m, IIr. shows an innovation in the form
of a nasal inserted into the ending, which is originally the form of the gen.pl.
in n-stems (19.1.18): yasna-n-m.19. The abl./dat.pl. has as an ending OAv.
-ibii, YAv. -aibii: OAv. vspibii, uxibii (Skt. ukthbhyas), miribii, YAv.
ams aibii sptaibii, but forms with the ending -aibii are also to be seen
in OAv.: daibii, martaibii. Note that an -i- of pronominal origin (already
in IIr.) is inserted between the thematic vowel and the ending. Before en-
clitics, the ending appears as -aibiias, which exactly reflects the IIr. ending
*bhi as > Skt. -bhyas: dtibiias-c, vspaibiias-ca (Skt. vvebhyas).20. The
inst.pl. of the thematic inflection diverges from the rest of the case forms, in
that it shows an ending -i, deriving from PIE *-is: OAv. zasti (Skt. hs-
tai), mri, Av. mas iii-ca
21. The loc.pl. shows an ending -au < *-

ai u: nmnau, mas iia (Skt. mrtyeu), vspa. Note the frequent addi-
tion of the postposition after the loc. (16.8): daxmauua, stnauua sta-
ble.
19.2 -stems
The -stems make up a class of numerous feminine nouns, whose origin goes
solidly back to the parent language. In effect, the - of the stem goes back
to PIE *- < *-eh
2
, a reconstruction that allows for the -stems to be histor-
ically connected to the ablauting -stems (< *-ih
2
; devi
-type), whose formal

similarity is more than evident (cf. 19.3). The adjectives of the thematic inflec-
tion model their feminine forms after this inflection class. Some words that
belong to this class are Av. uruuar- plant, ga- living being, dan- reli-
gion, ssn- teaching, YAv. gruu- neck, v- herd, flock, zaor- liba-
tion.
19 vowel stems 57
table 8 The inflection of -stems
nom. dana
ub
voc. dane gaa
uruuaire
acc. danm
gen. danaiia
uruuaraiia
gaanm
abl. danaiit
gabii gauuii
dat. danaiii vbiia
inst. dana
danaiia
gab
loc. gruuaiia (= gen.du.) gah gahuua
N.B.: Singular: 1. The nom.sg. has a zero ending and thus terminates in -a
,
depending on the dialect (cf. 7.2). Comparison with other languages (Skt., OP,
Gr.) yields the reconstruction *- < *-eh
2
.2. The acc.sg. has an ending -m <
*-m; cf. 7.6.2.3. The voc.sg. of this inflectional class reflects an IIr. ending
*-ai (Skt. -e): uruuaire (cf. Skt. ave). However, in contrast to this diphthongal
ending in IIr., other languages show forms in -: Gr. , , Umbr. tursa
(nom. -o), which should not be compared to certain Av. forms in -a
in which
the nom. performs the functions of the voc. Originally, the voc. was probably
a case with a zero ending; in some languages (cf. above), it may be seen that,
in this case form, the laryngeal disappeared, leaving behind coloration as its
only trace; cf. further stems in - (19.3.3).4. The gen.sg. has an ending -aiia
< *-i s, similar to the ending of other Iranian languages that also share the
--, versus other endings such as those of OP in -y and those of Skt. in
-ys. We suppose that the -- has been extended from the inst.sg. in -aii (cf.
below and also 7.3.2). Other languages present an ending that comes from
PIE *-eh
2
-es: cf. Dor. Gr. , etc., and traces in Lat. familis, uis, Umbr. tutas
city.5. The abl.sg. is identical to the gen.sg., but YAv. has adopted, just as in
other inflectional classes, anending -aii-t formed after the thematic inflection
(19.1.6): uruuaraiit , zaoraiit .6. The dat.sg. has an ending -aiii < *-i i
(on the --, cf. the gen.sg.), comparable to -yai of Skt.: srya
yai, etc. Examples

from other languages, such as Gr. , Lat. uiae, or Oscan deva allow for the
reconstruction of an ending *-i < *-eh
2
-ei .7. The inst.sg. has two forms: one,
danaiia
, agrees with the corresponding Skt. forms in -ay (and which, as has
been mentioned, serves as a model for some oblique cases of the paradigm),
and the other, dana
, formed with just the pure stem plus the ending. Skt. also
possesses both endings: jihva
and jihvy; the forms in -ay are more common.

Dual: 8. The nom./acc./voc.du. has an ending -e
, equivalent to Skt. -e:

ubh both, pre cheeks. It is usually reconstructed as PIE *-eh
2
-ih
1
.9. The
gen./loc.du. shows an ending equivalent to Skt. -yo10. The abl./dat./inst.
du. has an ending that is only slightly different fromthat which Skt. presents in
-bhym (15.3).
Plural: 11. The nom./acc./voc.pl. exhibits an ending -a
from IIr. *-s (< PIE

*-eh
2
-es) > Skt. -s (priya
s, but also priya
sas). Before enclitics, this ending

appears as -a
s (cf. ch. 2, fn. 2): uruuara
sca
. The nom. andthe acc. were identical

in Indo-Iranian. Comparison reveals some languages with an equivalent end-
ing: Skt. pr tans, Goth. gibos (Germanic *-z), and (Balto-Slavic) Lith. mergs,
while others show an ending deriving from *-ns < *-eh
2
-ns: Cret. Gr. -
(Attic-Ionic
< *-ans), Lat. uis, Umbr. vitlaf (Lat. uituls), OCS eny.
One thinks of different dialectal (or already IE?) treatments of *-eh
2
-ns: in
one group, the nasal was lost in this sequence, while in the other group, it
was maintained (or restored?).512. The gen.pl. shows an ending in -anm
with short --, in contrast to the other Indo-Iranian languages, which have an
ending in -nm (cf. 7.3.1 and esp. 19.1.18). In the YAv. form nnm(ca),
the monosyllabic stem n- may have been restored.13. The abl./dat.pl.
shows an ending equivalent to Skt. -bhyas. Before enclitics, the Av. ending
appears as -biias: uruuarbiiasca, zaorbiiasca. The YAv. forms gauuii
and vinuii exhibit the regular lenition explained under 11.1.4.14. The
inst.pl. is formed in the usual fashion and is equivalent to the Skt. ending -
bhi.15. The loc.pl. is formed in the usual fashion with the ending -hu (Skt.
-su). Note the frequent addition of the postposition after the loc. ending
(16.8), which took place sufficiently late so as not to show the development
*-hu - > *-vh- between vowels: (11.31): gahuua (gah), ghuua.
19.3 -stems
The -stems make up a class of relatively numerous and very productive fem-
inine nouns, which were already present in the parent language. The stem
exhibits an ablaut --/-ii- in Avestan, which goes back to PIE *-- < *-ih
2
- /
*-i - < *-i eh
2
-. -stems are basically used to form motion feminines to athe-
matic forms, i.e., to provide the corresponding feminine form to athematic
nouns andadjectives (cf. 20): e.g., YAv. dauu- (evil) goddess, dr- (female)
giver, nir- woman (from nar-), sn- bitch (from span-). Cf. further adjecti-
val forms, especially stems in -u, -a, -u at, -mat, and participles in -at and
-u, e.g., Av. vavh- to vohu- good, as aon- to as auuan- truthful, bauuait- to
bauuat- becoming, ciciu- to ciciah- having perceived.
5 From an IE sequence *-eh
2
ns, a realization *-eh
2
n s is normally to be expected, but in this
sequence, the same development as in forms such as Gr. , (); Skt. dya
m, ga
m; Lat.
diem, Umbr. bum, may have occurred, where *-eu m *-m (Lex Stang), i.e. *-eh
2
ns *-ns.
19 vowel stems 59
table 9 The inflection of -stems
nom. as aoni
voc. as aoni
az as aon
acc. as aonm
gen. as aoniia
as aoninm
abl. (YAv. barriit )
as aonibii
dat. as aoniii
inst. vaviia
loc. xariu barriuua

N.B. Not all case forms are attested, for which reason there are gaps in the
chart above. Inthe following notes, correspondences withSkt. will be indicated
systematically.
Singular: 1. The nom.sg. has a zeroending, andsoterminates in-i
, depending
uponthe dialect (7.2). Comparisonwithother languages (Skt. -, Gr. -) allows
for the reconstruction of an ending *- < PIE *-ih
2
; cf. Skt. devi
.2. The acc.sg.

has anending -m, cf. Skt. devi
m, whichmust goback to*-m<*-ih

2
-m.63. The
voc.sg. terminates in -i
, from which one cannot tell whether it is the original

voc. form or rather nom. pro voc. Skt. shows an ending -i: dvi << PIE *-ih
2
;
on the loss of the laryngeal, cf. 19.2.3.4. The gen.sg. has an ending -iia
<
*-i s < PIE *-i eh
2
-s; cf. Skt. devya
s.5. The abl.sg. has, in YAv., an ending -iit ,

remodeled by analogy after the thematic declension (19.1.6). In OAv. (and in
the rest of the IE languages), it is identical to the genitive (cf. above).6. The
dat.sg. has an ending -iii, which can be taken back to *-i < PIE *-i eh
2
-ei ; cf.
Skt. devyi.7. The inst.sg. has an ending -iia
; cf. Skt. devya
(< PIE *-i eh

2
-h
1
,
with ZG?)8. The loc.sg. is devya
m in Skt.
Dual: 9. The nom./acc./voc.du. has an ending in -i
; cf. Skt. devi
. The du. is
not well attested in Avestan. Skt. has dv for the voc.du., devi
bhym for the

dat./abl.du., and devys for the gen./loc.du.
Plural: 10. The nom./voc.pl. shows an ending -; cf. Skt. devi
. Its recon-
struction is disputed, because the expected ending, parallel to the ending
reconstructed for the --declension, would have been PIE *-i eh
2
-es (or *-ih
2
-es),
whereas - can only go back to IIr. **-iHs. The case ending here has probably
beensubject toinfluence fromthe corresponding ending of the --inflection.
6 In some other languages, however, there are also indications of a FG *-i eh
2
m, cf. Goth. mauja
to mawi.
11. The gen.pl. shows an ending -inm, with --, versus Skt. devi
nm; on the
vowel shortening, see the ending -anmof the -stems with -- (onthis change,
cf. 19.2.12). The form vavhnm owes its -- to the preceding labial glide, vh
(7.15.1). The same takes place in the12. abl./dat.pl. form vavhbii, which
normally has an ending -ibii; cf. Skt. devi
bhyas.13. The inst.pl. is devi
bhi
in Skt. 14. The loc.pl. has an ending -iu, which, with the usual postposition ,
becomes -iuua (16.8); cf. Skt. devi
u.
19.4 -stems
Just as with the -stems, it is presumed that -stems served to make derived
feminines, using the suffix *-h
2
, to u-stems. In Avestan, only three witnesses to
this formation are found. These forms are distinguished from u-stems through
their inflection, insofar as they are clearly hysterodynamic forms. Each one of
these stems has corresponding forms preserved in Sanskrit: tan- body (Skt.
tanu
-), ar- unmarried, virgin (Skt. agru
- < PIE *n -g
u
rh
2
-h
2
- not-pregnant)
and hiz- (along with hizuu-) tongue (Skt. juhu
- and jihva
-). It is likely that

these stems were already seen as simple stems in -u with hysterodynamic
inflection in YAv. times.
Singular: 1. The nom.sg. tanu has a short vowel -u-, in contrast to Skt.
tanu
, possibly due to the influence of u-stems.2. The acc.sg. YAv. tanm

can go back to *tanuu am from *tanuHm , which seems to be confirmed in OAv.
tanuu m, cf. Skt. tan
u
vm.3. The gen.sg. tanuu, hizuu has an ending *-uH-
as, proper to the hysterodynamic type.4. The abl.sg. YAv. tanuuat has anend-
ing remade by analogy after the thematic inflection (19.1.6).5. The dat.sg.
tanuii shows the development of the sequence -uu ai into -uii (10.2.3).6.
The inst.sg. tanuua is the equivalent of Skt. tan
u
va
, but Av. hizuua
couldequally
continue an original form in *-uH-aH or an inst.sg. *hizu from an -stem.7.
The loc.sg. tanuui represents *tan-uH-i.
Plural: 8. The nom./acc.pl. aruu goes back to an IIr. ending *-uH-as. OAv.
tanuu shows a trisyllabic scansion: /ta-nu-u ah/.9. The gen.pl. has taken
the ending -unm, found for the expected *-uH-nm, from the u-stems; cf.
Skt. tanu
nm.10. The abl./dat.pl. similarly has an ending with -u: tanu-

bii; cf. Skt. tanu
bhyas.11. The inst.pl. is attested only in OAv. hizub (Skt.

juhu
bhi), which implies, at least for OAv., that this whole class is remod-
eled after u-stems.This conclusion finds further confirmation in the OAv. 12.
loc.pl.
+
tanu-c, YAv. tanuu.
19.5 i- and u-stems
The paradigms of i- andu-stems exhibit suchsimilarity that the twoinflectional
types may be presented conjointly. The endings are the same as the endings
19 vowel stems 61
table 10 The inflection of i- and u-stems
nom. as i xratu
as aii xratauu
voc. paite maii paiti maii rzu
acc. as m xratm neut. ai vavhi gair xrat
nom.acc.neut. biri vohu r voh
gen. as i xrat u as iuua
maiuua
gairinm vohunm
abl. garit xrataot
gairibii da hubii
dat. axtiii vahauue
aibiia
ahubiia
inst. as xrat
loc. gara vahu ahuu vahuu
of the consonant stems. These two inflectional types include a considerable
number of feminine and masculine nouns, as well as adjectives.
Some words belonging to the i-stems are Av. f. as i- reward, m. axti- pain,
adj. biri- abundant, YAv. m. gairi- mountain, m. paiti- lord. Some words
belonging to the u-stems are Av. m. ahu- life, m. prtu- bridge, m./f. gauu-
bull, cow; adj. aidiiu- harmless, m. xratu- intelligence, m. gtu- road, way, f.
dax iiu- (YAv. da hu-) country, m. maiiu- spirit, adj. vohu- good. In addition,
some very interesting neuters are present here: iiu- life, duru- wood, znu-
knee. As for i-stems:
Singular: 1. The nom.sg. shows ZG of the suffix and of the ending in the
regular type, while for haxa companion and kauua Kavi, a hysterodynamic
nom.sg. with LG of the suffix and a zero ending must be reconstructed: *sakhi
and *kau i . Comparison with the nom.sg. skh of Skt. shows that the loss
of the final *-i already took place in IIr.2. The acc.sg. has *-i-m (which
always becomes -m), except in hysterodynamic stems, where the PIE ending
*-oi-m became IIr. *-i am (by Brugmann, 7.4.6), which is attested in OAv.
hu.haxim (Skt. skhyam). YAv. kauuam can be the result of a shortening of
*kau i am, i.e., *-i am > *-ai am. On the acc. ram, Skt. raym, cf. 10.1.3.3. The
nom./acc.neut. has a zeroending: biri-c(Skt. bhu
ri)4. Thegen.sg. of IIr. PD

type, *-ai-, has sometimes been provided, in YAv., with an analogical i on the
stem: janiii to *jani-.5. The abl.sg. (cf. 19.1.6) in -it is formed off the IIr.
ending *-ai, the - of which has been substituted by the -t of the abl.6. The
dat.sg. exhbits ablaut variants, following two stem types. One type comes from
*-ai -ai , and is found as -iii in OAv. axtiii (to *axti i ), and in YAv. as - e, with
contraction of the final diphthong and loss of the intervocalic -i -: frauuas e for
Fravai. The original sequence is more clearly reflected before an enclitic: nip-
taiiaca to protect. The other type of dat.sg. comes from *-i -ai : paie (to pati-)
or has
e (to haxai-), with absorption of the *i by the preceding fricative.7. The

inst.sg., in general, comes out of *-i-H; the HD ending, *-i-aH, is found only in
has
a from *hax-i -.8. The loc.sg. has a zero ending LG of the suffix; its *-
corresponds to the Skt. forms in -, and again attests to the loss of *i in the PIE
sequence *-i #/-i #.
Dual: 9. The nom./acc./voc.du in -i (arapaiti) goes back to IIr. *-, though
haxaiiashows FGinthe suffix, whichcouldbe basedona LG; cf. Skt. skhy.
10. The nom./acc./voc.neut.du. likewise shows *-: ui two ears, ai two
eyes.11. The gen.du. shows the ending *-s in OAv. haxtiia
.12. The
abl./dat./inst.du. shows the usual *bhi : uibiia and aibiia.
Plural: 13. The nom./voc.pl. in -aii has FG of the suffix and the proper
ending for the case, inherited from PIE.14. The acc.pl., coming from PIE
*-in-s, preserves a trace of nasalization of the *i in the ending -.15. The
nom./acc./voc.neut.pl. has an ending -i and comes from PIE *-i-h
2
.16. The
gen.pl. has taken -nm as its ending from the n-stems, just as was done in Skt.,
but without lengthening the stem vowel (or rather, shortening it, following
19.1.18?): YAv. gairinmfromthe mountains versus Skt. -nm. However, there
are some forms that reflect the expected PIE ending *-i-om: Av. kaoiim from
*kau -i-m, has
m from *saxi m (cf. Lat. marium, Gr. ). In Av., the long

vowel proper to the a-stems has been introduced.17. The abl./dat.pl. ends in
the familiar -bii.
As for the u-stems: Avestan has regularized the paradigms in favor of the
proterodynamic type (cf. 17), though some traces of the old distribution of
HD and PD stems, which will be indicated immediately below, still remain. On
the neuters, cf. 19.5.39ff.
Singular: 18. The nom.sg. has the ending *-s: xratu, prtu; cf. Skt. krtu.
The nom.sg. bzu (in compounds) shows a nom. with lengthened grade
(HD) < PIE *u -, in contrast to the simplex bzu. The nom.sg. of gauu- is
gu (Skt. gu).19. The voc.sg. has an ending -uu < IIr. *-au : rzuu
straight, but maii preserves, in all probability, the intermediate stage -, in
virtue of the preceding ii (10.3.4).20. The acc.sg. lengthens the *u before
the -m: ahm, dax iim (cf. Skt. dsyum). Forms with FG and LG of the suffix
also exist: da haom to *dasi au am, nasum < *nasu am. In YAv. texts, dax iim
and da haom are used without any apparent semantic distinction.21. The
nom./acc.neut. has a zero ending: vohu
(OAv. vohu-c); cf. Skt. vsu22. The

gen.sg. in IIr., *-au - (10.4), takes the form - u in OAv.: dax ii u; meanwhile,
in YAv., it takes the form -ao: ranao. In some cases, it appears that YAv.
forms have been introduced into OAv.: rzao, parao, mriiao. Opposite
these PD endings, some HD forms in -uu (< *-u -as) are found: YAv. rzuu
19 vowel stems 63
finger, xra, pasuu, ra (11.14.4). Forms ending in - u found in YAv.
must be considered Gathicisms: ah u. HD stems have gradually taken on
PD inflection, cf. Skt. krto and krtvas. In Skt. the PD type has also been
extended.23. The abl.sg. differs from the gen.sg. only in YAv. (cf. 19.1.6):
YAv. xrataot , da(i) haot 24. The dat.sg. has an ending -auue
(*-au -ai ): YAv.

da(i) hauue; remnants of HD forms, meanwhile, show an ending -uue
(*-u -ai ):
YAv. xrae, pre. Some sound changes have partially disfigured this ending
(11.1.4): OAv. ahuii (modernization of the YAv.) and YAv. avhe < *ahu ai ; cf.
Skt. vsave and krtve.25. The inst.sg. has an ending -u
< PIE *-uh

1
: ah,
xrat, voh (vohu-c). The HDforms have an ending -uua
(11.14.4) < IIr. *-u aH:

xra; cf. Skt. krtv, mdhv (and mdhun).26. The loc.sg. has a zero
ending and shows ablaut in the stem vowel (LG and FG): *-u and -, -uu
(both from IIr. *-au ; 10.3.4): Av. vahu, xrata
(read xratu; cf. 7.5.3), YAv.

dai h (and dai huu), gtuu; cf. Skt. vsau, krtau. In addition, forms with
the postposition (16.8) are found: YAv. dai hauua, ahauua (< *-a
u -a).
Dual: 27. The nom./acc./voc.du. of animates has an ending -u
< IIr. *-uH:

YAv. dai hu; cf. Skt. mdh, bhu
. In YAv., the isolated HD ending -auua is

found: bzauua (Skt. bhv),
+
zanauua.28. The nom./acc./voc.neut.du.
terminates in -uui
: vavhi
; cf. Skt. urvi
( ja
nun ). An ending -u, as in the masc., is

not uncommon: pouru-ca, vohu-ca.29. The gen.du. comes from*-uu h: OAv.
ahuua
, YAv. pasuua
.30. The abl./dat./inst.du. comes from *-ubi : ahubii.

The YAv. formbzue exhibits fricativization of the -b- and the development of
final *-i to -e (7.10.3).31. The loc.du. is attested in OAv. ahuu (ahu- life)
< *ahu -au .
Plural: 32. The nom./voc.pl. continues PIE *-eu -es in the ending -auu,
-auuas-ca, while the YAv. forms da huu (OP dahyva) and nasuu seemto
reflect *ou -es.33. Theacc.pl. in- <*-un-s reflects nasalizationonthe vowel:
Av. vah, nas. The corresponding HD in *-u n -s appears in YAv. pasuu
(Skt. pav) and in pr crossing, gate, bridge (from *pr t-), versus the
acc.pl. of the PDtype in OAv. xrat, prt. Some forms seemto have adopted
the form of the nom.pl.: ps auu bridges (from *pr tu-), da huu.34. The
nom./acc./voc.neut.pl., seen in OAv. voh, comes from IIr. *u asu-H.35. The
gen.pl., as in dax iiunm, vohunm, could have retained the -u-, in contrast to
the ending -nm of Skt. and OP. The HD form *-u -m is attested in YAv. pasu-
um, vahuum.36. The abl./dat.pl. has restored the *-b- (e.g., da hubii),
with the exception of hiniii.In contrast, the 37. inst.pl. in *u-bhi shows
systematic lenitionof the *-b- (auuah < *a-vahu-, ytu < *ytu, both
in Y 12.4) or employs the form of the dat./abl.pl. (bzubii). A YAv. form in -
ub is unknown.38. The loc.pl. has *-u-u, which, with the postposition
in, gives YAv. pasu.huua < *pasuu among the sheep.
Afewneuter nouns, whose inflection retains interesting remnants of ablaut,
must be mentioned as well: iiu- life, duru- wood, znu- knee. 39. The
nom.sg. iiu, duru, znu shows LG by Brugmann (cf. 7.4.6; cf. Skt. da
ru,
ja
nu, Gr. , ). YAv. nm is a secondary form, but it illustrates the ZG

of the stem that is also used in compounds.40. The gen.sg. is seen in OAv.
yao, YAv. drao (Skt. dr < *dreu -s versus Hom. Gr. < *deru -os).41.
The dat.sg. is OAv. yauui, yauu.42. The inst.sg. is yauu.43. The abl.pl.
is YAv. nubiias-cit (cf. n- in 11.25.4).
20 The Adjective
Just as in Indo-European, adjectival inflection is not in any way distinct from
nominal inflection in Avestan. The Indo-European situation was preserved in
Indo-Iranian down into Avestan, with the exception of some innovations in
certain adjectives that have been pronominalized (20.2). The function of the
adjective is expressed through its position and agreement in the sentence, as
well as by the semantics of its formative suffix.
Regarding gender, it may be noted that the neuter differs fromthe masculine
only in the nom. and acc. cases. The athematic classes usually form the femi-
nine by way of the motionsuffix *- <PIE*-ih
2
, while the thematic adjectives are
served by -: pru- m., pr- broad, versus paana- neut., paan- f. dis-
tant, vast. Thematic adjectives with the motion suffix - are occasionally found.
In those cases, an individualizing value (with substantivization) is discernable:
YAv. apura- m., apur- without sons, versus hupur- she whohas goodsons.
20.1 Degrees of Comparison (Gradation)
Degrees of comparison in the adjective were expressed in the parent language
through special suffixes. In the same fashion, Avestan forms comparatives and
superlatives withinheritedsuffixes. Twotypes of suffixes usedtoexpress degree
of comparison on adjectives are found.
1. The more archaic type, used for adjectival comparison stricto sensu, is
formed using a pair of suffixes coming from IIr. *-i as-, for the comparative,
and IIr. *-ita-, for the superlative. These are primary suffixes: both are added
directly to the root with FG, i.e., the stem is stripped of its suffix: the compara-
tive of *vah-u- good is OAv. vah-iiah- (cf. Gr. : - -). This kind of
adjectival comparison ceased to be productive in the Avestan period.
Examples: ura- strong gives aojiiah- aojita-; kasu- small gives kasiiah-
kasita-; mazat- large gives maziiah- mazita-; vahu- good gives OAv.
vahiiah- (YAv. va hah-) vahita-; srra- beautiful gives sraiiah- srata-.
20 the adjective 65
2. The second type is formed using a pair of secondary suffixes coming from
IIr. *-tara- and*-tama- (cf. Gr. : - -), the originof
whichseems tolie informations having oppositional or contrastive function. In
effect, the PIE suffix *-tero- served to express the notion of a binary opposition,
first with adverbs and pronouns: cf. Gr. before, Skt. ttara- above,
Skt. katar- and Gr. which (of two)?. Later, the use of this suffix was
extendedtoother nominal classes as well. Inthe same fashion, the suffix*-tama
was also first used with adverbs, cf. Skt. ntama- that which is nearest, Lat.
intimus, Skt. katam- which (of several)?.
This second type of of adjectival comparison is the only one that has con-
tinued to be productive beyond Proto-Avestan. The suffixes are appended to
the stem(i.e., they are secondary), and are even found attached to compounds:
a.aojastara- stronger, froma.aojah- with much force; vrrajstma- most
victorious (from vrra-jan- destroying the obstacle).
20.2 Pronominalized Adjectives
The semantic proximity of some adjectives to the pronouns induced these
adjectives to partially adopt some endings from the pronouns. This phenome-
non is not unknown in other Indo-European languages: cf. Lat. gen.sg. ullus,
dat.sg. ull. This formal reshaping must have taken place already in part dur-
ing the Indo-Iranian period, judging from various forms of the m./n. plural of
a-stems (cf. 19.1), such as IIr. *-ai bhi as and *-ai u, with correspondents in San-
skrit, Avestan, and Old Persian. Beyond these forms, this phenomenon is found
to be much less extensive in Avestan than in Sanskrit.
The forms that we find are as follows: dat.sg.m/n. YAv. aiiahmi for the
other (Skt. anysmai); vspmi for all, and YAv.
+
vsp m i (5.2.5, 7.9.2; Skt.
vvasmai; the nominal formvspi is also found); the nom.pl.m. YAv. aiie (Skt.
any), vspe (Skt. vve), versus OAv. vspa
h; the gen.pl. YAv. aiiam (Skt.

anym), YAv. vspam (Skt. vvem), versus Av. vspanm. Some pronom-
inalized forms of the numeral auua- one are likewise evident: loc.sg.m. a-
uuahmi (see Skt. kasmin), gen.sg.f. auua ha
(see Skt. kasys).

table 11 Numerals
Cardinals Ordinals Cardinals Ordinals
auua- 1 fratma-; paoiriia- duua.dasa 12 duuadasa-
duua- 2 daibitiia-; bitiia- vsaiti- 20 vsstma-
ri-, tir- 3 ritiia- rissas, risat- 30
caar-, catar- 4 tiriia- caarsat- 40
paca 5 puxa- pacsat- 50
xuua 6 xtuua- xuuati- 60
hapta 7 haptaa- haptiti- 70
ata 8 atma- atiti- 80
nauua 9 naoma- nauuaiti- 90
dasa 10 dasma- sata- 100 sattma-
21 Numerals
21.1 Cardinal Numbers
The numbers 1 through 4 have a full nominal inflection in case and gender,
but inflection for number depends on the numeral involved: 1 inflects in the
singular, 2 in the dual, while 3 and 4 inflect in the plural. The numbers 5 to
19 are indeclinable, and all following numbers are inflected depending upon
their respective stems.
1 is auua- (OP aiva-, Gr. alone), and differs in suffix from Skt. ka-
(from *ai ka-; Mittani aika), though it is found in the Skt. adv. ev(m) indeed
< PIE *oi -u o just that, on which the Av. form rests. The numeral auua- is
declined as an a-stem, although some pronominal forms occur (20.2). One
form that calls for comment is the acc.sg. iim, which is the regular reflex of
IIr. *ai u am by way of *i u m > *i u um > *i um. In some texts, the simplified
variants im and m can be found.
2 is duua- (Skt. d
u
va
Gr. ) and also OAv. uba-, YAv. uua- the two, both
(Skt. ubha
, OP ub, OCS oba; with a nasal, Gr. , Lat. amb). Both stems
agree in their inflection with the dual of an a-stem(cf. 19.1): nom.acc.m. duua,
uua7 (OP ub, Skt. ubha
); nom.acc.neut./f. OAv. ub, YAv. duiie, duuaca, uiie

7 This is the only Av. form with initial uu-, because *u - always becomes v- (11.1.2). The manu-
scripts often show va, v, auua.
21 numerals 67
(Skt. ubh; cf. 10.2.3); inst.dat.abl.m. OAv. ubibii, YAv. duuaibiia (Skt.
dva
bhym), uuaibiia (Skt. ubha
bhym); gen.m. duuaiia
, uuaiia
; loc.m. OAv.
ubii, YAv.
+
uuaii (Skt. ubhyo).
3 has a stem ri- for the masculine and neuter, but tir- for the feminine,
just as with Skt. tri- and tisr-. This divison goes back to PIE, to judge from
OIr. toir < *tisres. In the other IE languages, the feminine stem is completely
lost. The forms attested in Avestan are: nom.m. rii, raiias-ca (Skt. tryas;
note the of the Av. versus the short a of the Skt. Perhaps Av. has taken this
from the number 4). The acc.m. r (Skt. tri
n) derives from PIE *trins. The

YAv. nom.acc.neut. r (cf. 7.2.; Skt. tri
) derives from PIE *tri-h

2
. The inst.m. is
ribii (Skt. inst. tribh, dat. tribhys). The gen.m. exhibits two forms, raiim
and riim, though it is difficult to know which is the original, given that Skt.
tra
m offers no help. The feminine possesses a nom. rii (originally masc.),

the acc. tir (Skt. tisrs) and the gen. tirm.
4 possesses two stems that depend upon gender. The masculine and neuter
are formed from a stem caar-/catur-, while the feminine is built to a stem
catar-. The nom.masc. is car, with resulting from Brugmann (7.4.6),
just like Skt. catva
ras; with an enclitic, a shortened vowel is found, thus

caaras-ca, following 7.3.1. The acc.m. catur shows an ending introduced
from the thematic nouns in -a (Skt. catras). The same has happened with the
nom.acc.neut. catura, but not with the gen. caturm. The feminine of 4 also
carried a formant *-sr-, like that of 3, whence YAv. acc. catar from IIr. *tas-
ras (Skt. ctasras). As was already mentioned above, these forms can go back
to PIE, to judge from the OIr. continuant cethoir < *k
u
etesr-.
The cardinal numbers 5 through 19 were originally indeclinable, but there
are nevertheless forms of the gen. attested for 5, 9, and 10: pacanm (Skt.
pacna
m), nauuanm (Skt. navna
m), dasanm. The form pac-ca in

pacca vsaiti- 25 has secondary (cf. 16.1.2) opposite the regular nauuaca
nauuaiti- 99. These forms correspond exactly to their Skt. equivalents. Note
that the initial x- in xuua 6 results from the secondary prothesis of x- onto
an initial cluster of plus consonant (cf. 11.16). Of the numbers from11 to 19,
only duuadasa 12 and pacadasa 15 are attested as cardinal numbers; the rest
appear only as ordinals.
The decads inflect as follows: the decads from 20 to 50 behave as -ant-
stems, while those from 60 to 90 are abstracts in -ti. 20 corresponds to Skt.
vim at- (with secondary nasal), from *vi-ati- < PIE *du i-dm t-i two decads.
For 30, we have the nom.m. riss, which continues the PIE form *tri(h
2
)-
domt- (Gr.
, OIr. tricho), the nom.acc.neut. risatm, and the gen.

risatanm, which point to a secondary thematization of the oblique stem
*risat-, retained in risaat- 30 times. Another athematic form occurs in
pacsat b-ca, the inst. of 50, the long of which calls for comment. Both
this form and Skt. pac-t show lengthening of an *e following the loss of
a preconsonantal *d, proceeding from PIE *penk
u
e-dm t-. The long vowel of
haptiti- and atiti-, which the Skt. forms saptat- and at- lack, may be due
to the form of 80, if it was built off the simplex 8 (Gr. ; PIE *h
3
etoh
x
), i.e.,
from IIr. *ataH-ti-.
The numbers sata- hundred and hazara- thousand inflect as a-stems, but
bauuar ten thousand inflects as a neuter r/n-stem. Av. sata- is equivalent to
Skt. ata- and to forms in other IE languages that allow for the reconstruction
of a PIE *m t- from *dm t-- (the) tenth (decad). It is notable that a form
for 1000 cannot be reconstructed for PIE, but Av. hazara-, as well as Skt.
sahsra-, do contain an IIr. element *-hasra- < PIE *sm -heslo- (cf. Lat. mlle <
*smih
2
-heslih
2
), whichcanbe comparedwithGr. (Ionic) , (Aeolic)
thousand < *heslii o-.
21.2 Ordinal Numbers
For first, fratma-, the original sense of which was the foremost (cf. Skt.
pratham-), and which was a superlative to fra forth, in front, is used. First
in the temporal sense was *pr Hu - in IIr. (derived from PIE *pr h
2
s before);
cf. Skt. pu
rva-, Av. pauruua- previous. The suffix *-i a- was added to this stem
already in (Proto)-Iranian: *pr Hu i a-, which results in Av. paoiriia- (cf. 8.2).
Second was IIr. *du i-tii a-, which gave OAv. /dbitiia-/, written daibitiia-
(11.13), and, with loss of d-, YAv. bitiia-; cf. Skt. dviti
ya- The same formation

is encountered in ritiia- third, starting from a stem rita- (cf. Gr. ; a dif-
ferent stem in Skt. tr ti
ya-), which is also found in names. The Av. form tiriia-

fourth < IIr. *ktur-()i a- < PIE *k
u
tur, with ZG of the root, had already lost the
*k- in IIr. times, judging from Skt. turi
ya-, but the k- is retained in Av. xtirm

four times < *-kturi am. Av. puxa- fifth seems to be a product of an ear-
lier *paxa- (Skt. pakth-), also with ZG in contrast to paca 5, which would
have taken the -u- from*turi a- fourth and froma supposed *xuta- sixth, the
actually attested form of which, xtuua-, lacks a definitive explanation.
The ordinals from seventh to tenth correspond exactly to their respec-
tive forms in Skt., and exhibit incipient generalization of the IIr. suffix *-m-
< PIE *-m-, which would become very productive in the later Iranian lan-
guages: 7th: Av. haptaa- (Skt. sapttha-); 8th: YAv. atma- (Skt. aam-);
9th: YAv. naoma- (Skt. navam-); 10th: Av. dasma- (Skt. daam-). Mean-
while, eleventh to nineteenth are formed with the same simple thematic
inflection of the cardinal numbers, thus auuadasa- 11th, duuadasa- 12th,
ridasa- 13th, carudasa- 14th, etc. The formvsstma- twentieth apparently
substitutes an earlier *vsastma- (from vsat-tama, cf. 11.20.2), which already
22 pronouns 69
shows the suffix *-tama-, used also in sattma- hundredth and hazartma-
thousandth.
The form of the acc.sg.neut. of ordinals can serve as a temporal adverb, thus
paoirm for the first time or firstly, bitm for the second time, though the
dat.sg.neut. is also used for this expression, thus paoiriii, bitiii.
21.3 Multiplicatives
For once, hakrt, a formidentical toSkt. sakr t, is used. Inbothforms, the prefix
sa- reflects *sm - (ZG of PIE *sem- one, united, cf. Gr. - once, one).
Other multiplicatives are built by adding a suffix -s after the stem in the ZG:
bi twice, ri thrice, caru four times, and xuua six times. For numbers
greater than six, adjectives with a suffix *-u ant- are formed, thus vsaitiuuat-
twenty times, risaat- thirty times, etc. It seems that this same suffix was
likewise used for the forms found in late texts, biuuat twice and riuuat
thrice, from bi and ri.
21.4 Fractions
Half is expressed by ara- (Skt. ardh-) or nama- (Skt. nma-). For the
formation of other fractions, a suffix *-su a-, which is attested in Avestan and
Old Persian, comes into play. This suffix developed from a metanalysis *s-u a-,
beginning from one third and one fourth: riuua- neut. one third (OP
*iuva-) < *tr-u a-, cf. Gr. fig leaf, and neut. riat- id. is also found;
caruuua- one fourth (OP *auuva-) < *catru-u a-; the form patahuua-
one fifth is unexpectedincomparisontoOP*panauva- <*panca-hu a-; we also
find haptahuua- one seventh and atahuua- one eighth (OP *atauva-).
22 Pronouns
Already in the reconstructed proto-language, the pronominal inflection was
characterized by a set of its own case endings, distinct from those belonging
to the inflection of nouns and adjectives. Suppletion is likewise very common
in the inflection of pronouns, i.e., distinct stems co-occur within one and the
same paradigm. In Avestan, like in the rest of the Indo-European languages,
the pronominal inflection has preserved some peculiar features, but analog-
ical interchange and mutual influence between the nominal and pronomi-
nal inflections are observable already from the oldest period (cf. 19.1.19,
20.2).
In Table 12, the (Proto-Ir.) endings from which one must proceed in order to
explain the majority of the Avestan forms are given. Where it is not possible to
table 12 The pronominal endings in their Proto-Iranian form
masc./neut. fem. masc./neut. fem. masc./neut. fem.
nom. *-, *-h, *- *-, *-H, *-ai *-ai *-ai
*-t *-H *-H *-h
acc. -(a)m *-m *-anh
gen. *-hi a *-hi h *-i h *-ai m *-hm
abl. *-hmt *-hi t
*-ai bi ah *-bi ah
dat. *-hmi *-hi i
inst. *-n, *- *-i a
*-ibi *-ai , *-ai bi *-bi

loc. *-hmi *-hi -ai u *-hu
abstract a concrete morpheme fromthe evidence, the box in the table has been
left empty. The personal pronouns must be considered separately.
22.1 Personal Pronouns
Personal pronouns are characterized by suppletion and lack of grammatical
gender from the Indo-European period forward. In Table 13, the attested forms
of both stressed and enclitic personal pronouns are given.
N.B. for accented pronouns:
1. The 1st.sg. has a nom.sg. az
m, which agrees with Skt. ahmand OP adam.

A hapax as-ct , of doubtful interpretation, is also known. The rest of the forms
are built froma stem*ma-: YAv. acc. mm(OPmm, Skt. ma
m); the abl. *-d (Skt.

md); the gen. *-na (OP man, OCS mene); the dat. *-bhi a (cf. Skt. mhyam);
YAv. muuiia<*maiia, cf. 7.4.2, 9.2.4, 11.1.4. The formmaibii, with- instead
of -, is due to the influence of pronominal dat.pl. forms in -bii, such as aibii
and yaibii (cf. 22.2, f.)
2. The 2nd.sg. uses the stems *ta- and *tu a-, but the nom. shows old *tuH
> Av. t alongside *tuH-am (OP tuvam, Skt. tvm), an innovation based on the
latter part of azm I. Cf. further vam we < *vai -am (cf. Goth. weis) and ym
you (pl.) (cf. Goth. js).
3. The 1st.du. v (Skt. va
m) is of the same stem as the nom.pl. vam. In the

1st.pl., a suppletive stem *asma- is used (22.1.5). A curious feature of this
inflection is that, except in the nom. and gen., the cases of the sg. and the
pl. are built with the same endings, availing themselves of suppletion in the
stem to convey the difference in number: cf. dat. maibii: ahmaibii (cf. Skt.
mhyam: asmbhyam), etc. For the 2nd.du., we have yauukm(cf Skt. yuva
ku).
Cf. further the Skt. forms nom. yuvm, acc. yuva
m, gen. yuv.
22 pronouns 71
table 13 Personal pronouns
1st Accented Enclitic 2nd Accented Enclitic
nom.sg. az m, azm asct tuu m, tm; t
acc.sg. mm m m
gen.sg. m.n, mana mi, m tauu, tauua ti, t
abl.sg. mat ? at
dat.sg. maibii, maibii,
muuiia
mi, m taibii, taibii ti, t
inst.sg.
nom.du. v
acc.du. uu?
gen.du n yauukm
nom.pl. vam y m, ym; y
acc.pl. hm, ahma na
, n va
, v
gen.pl. ahmkm n , n xmkm; ymkm v , v
abl.pl. ahmat xmat ; ymat
dat.pl. ahmaibii n , n xmaibii, xmuuiia v , v
inst.pl. hm xm
4. The 2nd.pl. OAv. y m, YAv. ym (cf. Skt. yym) has an unexpected
(from external sandhi of *y-am? 11.25.1). The oblique forms make use of the
stems xma- and yma-. An IIr. stem*uma- (cf. Aeolic Gr. acc. ) underlies
both of these stems. The variant xma- is explicable through loss of the u- >
*ma- and subsequent prothesis of x- (cf. 11.16). The variant yma- is due to
influence from the nom. y. Cf. dat. OAv. xmaibii and ymaibii versus Skt.
yumbhyam.
N.B. for enclitics (unaccented pronouns):
5. The 1st. Singular: inthe acc., mis identical toSkt. m. The oblique enclitic
(gen.dat.) is OAv. mi, YAv. m (OP -maiy, Skt. me).Plural: the stemna- is used
in the dual and plural, and it is also contained in the stem*asma- < PIE *n sme-.
In OAv., distinct forms exist for the acc.pl., na
(Lat. ns), and the oblique cases,

n (cf. Lat. ns-trum), in contrast to the single case form of YAv., n (and of Skt.
nas).
The 2nd. Singular: in the acc., Av. is identical to Skt. tv. The oblique
enclitic (gen.dat.) is OAv. ti, YAv. t (OP -taiy, Skt. te).Plural: In the plural,
the stem va- is used. In OAv., distinct forms exist for the acc.pl., va
(Lat. us),
and the oblique cases, v (cf. Lat. us-trum), in contrast to the single case form
of YAv., v (and of Skt. vas).
6. For the enclitics of the 3rd person, some forms of the old demonstra-
tive pronoun a-/i-, which were substituted by the stem ima (22.2), are used:
acc.sg.m.f. m(Skt. m, Cypriot Gr. ?); acc.sg.neut. OAv. t , YAv. it (cf. the Skt. par-
ticle t); acc.pl.m.f. ; acc.pl.neut. . Some YAv. forms with initial d- must have
originated in sandhi with a preceding dental (e.g., *yad im): acc.sg.m.f. dim(OP
-dim); acc.sg.neut. dit ; acc.pl.m.f. d (OP -d); acc.pl.neut. d.
Other forms took an *s- by analogy to the paradigm of sa-, ta-. Accented
froms are nom.sg.f. OAv. h and nom.du.neut. OAv. h. Enclitic forms are acc.sg.f.
Av. hm (OP -im, Skt. sm), acc.pl.m.f. Av. h (cf. OP -) and the gen.dat.sg.
OAv. hi, YAv. h, (cf. OP -aiy). At times, variation between h- and - occurs,
which is explicable as the result of ruki in external sandhi: pairi but at ca
h.8
7. The accented reflexive pronoun has adopted the stem *su a- from the
possessive pron. In Av., only the following forms are attested: gen. xvahe; dat.sg.
YAv. huuuuiia (from *hu ai a, cf. muuiia above 22.1.1) and YAv. xvi.
22.2 Demonstrative Pronouns
1. a-/i-/ima- this. This pronoun, of anaphoric origin, combines stems deriving
from PIE *h
1
ei -/*h
1
i- and *h
1
e- (cf. Lat. is, ea, id) in a suppletive pattern. The
nom.sg.m. comes from *ai am > Skt. aym (cf. OP iyam)The acc.sg.m. arises
from an IIr. innovation in which the old form *im (Lat. im) was accusativized
by adding the ending *-amThe nom.acc.sg.neut. imat (cf. OP ima) replaces
the *-m of the new form with the -t of the neuter pronouns. Starting from the
twopreceding case forms, it became possible toextract a stemima-, fromwhich
the rest of the direct cases of the masc., neut., and fem. with thematic endings
emerged.
The oblique cases are built off the stem a-: the gen.sg.m. comes from *asi a
> Skt. asya;9 the gen.du.m. a-iia
(cf. Skt. ay; 19.1.12); the inst.pl.m. i; the

abl.dat. and loc.sg.m. make use of a formant *-sm-. The inst.sg.m. from *an
(OP an; cf. Skt. ena
) is built with the stem ana- (cf. below 22.2.4).The

feminine forms of the gen.abl.sg. come from*asi s >Skt. asya
s; (the forma ht
8 In YAv., however, this old distribution is often disrupted. Old Persian, which maintains the
use of the 3rd person enclitic, offers only variants with --.
9 ahe is used attributively and anaphorically, while a he is only found in anaphoric use. ahe
probably represents the original development (cf. the gen.sg. of the a-stems), whereas a he
introduced - h- from the fem. pronouns.
22 pronouns 73
table 14 The demonstrative pronoun a-/i-/ima-
1. a-/i-/ima- this Masculine, Neuter Feminine
nom. sg. aii m, am m
acc. imm imm
nom.acc.neut. imat
gen. ahii, ax iica, ahe, a he a ha
, a ha
sca
abl. ahmt a ha
; a ht
dat. ahmi ax iii, a hi
inst. an, ana ii, aiia
loc. ahm, ahmi, ahmiia a he
nom.m. du. ima
gen. aiia
; a
sc
dat. biia bii
nom.m. pl. ime ima
acc.m. im ima
nom.acc.neut. im, ima, ima
gen. am a
hm
dat. aibii biiasc, bii, aiiiasca
inst. i; aibi b
loc. au, auua h, huua
shows the late substitution of *-s by *-t); the dat.sg. from *asi i > Skt. asyi;
the loc.sg. is from *asi (cf. Skt. asya
m). The inst.sg.f. from *ai (Skt. aya
) is an
exception in that it does not contain an -s-. There is also a form f. iia, which
owes its long vowel to the preceding *i in paiti iia zm on this earth.
In the plural, the m./n. oblique cases are built off a stem *ai -, which recalls
the nom.m. *ai -am, just like the stems ata- and auua- (see below). The
oblique cases of the fem.pl. are formed starting from an unexpected stem -,
which probably derives from the inflection of the pronouns ta-, ka-, and ya-.
Note the ending *-sm of the gen.pl., versus PIE *-som as reflected in Lat.
e-rum, e-rum, and OCS tx of these.
2. ta- that, 3. ata- that there. The first of these pronouns continues the
old demonstrative. Its paradigm conforms to the usual pattern of suppletion.
The nom.sg. was suppletive even in PIE, m. *s, neut. *td, f. *sh
2
> Av. h, tat ,
h (Skt. s, tt, sa
; Goth. sa, ata, so; Gr. , , ). The nom.sg.m. attests to a

sigmatic variant, OAv. h , YAv. h, has-cit . The stemtakes thematic inflection in
the direct cases, as well as inthe forms of the inst. Apart fromthe inst., there are
table 15 The demonstrative pronouns ta-, ata-
2. ta- that,
3. ata- that there Masculine, Neuter Feminine
nom. sg. h; h , h, hascit ; aa; a h; a
acc. t m, tm tm; atm
nom.acc.neut. tat ; atat
gen. atahe ata ha
; ataiia
abl. atahmt
dat. atahmi
inst. t; ata ataiia
loc. atahmi, atahmiia
nom.acc. du. t, tca
nom.acc.neut. t; ate
gen. ataiia
nom. pl. ti, tact , t; ate, ataca ta
, ta
scit ; ata
s.t
acc. t g, t t ; at, at ta
, ta
sc; ata
nom.acc.neut. t; ata, atcit

gen. atam atahm
dat. ataibii atbii
inst. ti
loc. atau, atauua
nooblique forms attestedtothe stemta-. Instead, the stemata- appears, which
has a complete inflection apart fromthe suppletive nom.sg., which comes from
m. *ai a (Skt. e), f. *ai (Skt. ea
).
The inflection of these pronouns follows that already seen under 22.2.1,
22.5.2, i.e., throughthe insertionof typically pronominal elements suchas *-sm-
in the oblique cases of the masc., or *-ai - and *- in the oblique cases of the
masc./fem.pl. There are some forms worthy of note: the gen.pl.fem. atahm
shows shortening of the *-- (Skt. eta
sm); the gen.sg.fem. ata ha
is equivalent
to Skt. etsyas, while ataiia
and ataiia
scit owe their endings to the influence

of feminines in - (19.2.4).
The fact that no forms of ata- are found attested in OAv. may be mere
coincidence, but the total absence of oblique forms of ta- in YAv. (except in
the inst.) is probably an innovation of YAv., in which ta- was replaced by ata-.
4. ana- this is a suppletive stem in Avestan that is found attested in other
languages: OCS on that there, Lith. ans that < PIE *h
1
ono-, which has left
22 pronouns 75
traces in IIr.: cf. the Skt. adv. ana
therefore (inst.sg.neut.), anay (inst.sg.f.).

Avestan attests to OAv. inst.sg.m. an (OP an; Skt. ana
); inst.pl.m.neut. ani;
gen.sg.m.
+
anahe; abl.sg.m. anahmt ; and the inst.sg.m.neut. ana.
5. The stem of the demonstrative auua- that there may be compared with
the Skt. gen.du. av (OP ava, and also OCS ov this < PIE *h
2
eu o-). The Aves-
tan paradigm possesses a nom.sg.m., OAv. huu (*hau ; cf. 10.3.4) and YAv.
hu, which is, in origin, a fem. form also attested as a nom.sg.fem. YAv. hu,
made froma supplementary stemcomparable to the Skt. nom.sg.m.f. asu that
there or OP hauv. The other case forms allow for derivation from the stem
auua-, with its already familiar pronominal endings: nom.acc.sg.neut. auuat
(OP ava); acc.sg.m. aom(OP avam); gen.sg.m.neut. auua he and auuaheca (OP
avahay); inst.sg.m. OAv. auu, YAv. auua, nom.pl.m. auue (OP avaiy); acc.pl.m.
+
auu ; nom.acc.pl.neut. auua and auua
(cf. 19.1.17); gen.pl.m. auuam (OP

avaim); inst.pl.m. auui; acc.sg.f. auum (OP avm); gen.sg.f. auua ha
;
abl.sg.f. auua ht ; dat.sg.f. auuahi; acc.pl.f. auua
andauuaha
(ina late text);

inst.pl.f. auuabii (*-bii).
22.3 Relative Pronoun
The relative pronoun ya- which, who (Skt. y-, Gr. < PIE *[h
1
]i -) follows the
same inflection as the pronominal stem ta-, ata-, etc. seen above, for which
reason the enumeration of all the forms here would be superfluous. Keep in
mindthat the initial y-, incombinationwith*-i - inthe following syllable results
in raising of the -a- (7.10), and one therefore finds gen.sg. OAv. yehii, yex iic,
YAv. ye he (Skt. ysya); gen.sg.f. ye ha
(Skt. ysys); abl.sg.f. ye ht , ye h-a;

loc.sg.f. ye he (v.l. with h). Curious are the nom.pl.m. YAv. yi (Skt. y), which
did not develop to

y (cf. 10.2.2), and the nom.acc.sg.neut. OAv. hiiat (versus
YAv. yat ; Skt. yt), which may be the result of external sandhi: h yat . On the
other hand, the acc.pl.m. formOAv. y gs-t which you apparently arose from
a blending of the regular forms, OAv. y g and ys (cf. Skt. ya
n, ya
ca).
22.4 Possessive Pronouns
The possessive pronouns for the 1st and 2nd persons are attested in OAv. alone,
while the gen. of the corresponding personal pronouns assumes the functions
of the possessive in YAv. Various formations co-occur in the possessive pro-
nouns, as will be seen below.
Singular: The 1st person is formed in OAv. with a stem ma- my (Gr. <
PIE *h
1
mo-): nom.sg.m. m (7.9.4); gen.sg.neut. mahii; dat.sg.neut. mahmi;
inst.sg.neut. m; loc.sg. mahm; nom.acc.pl.neut. m; inst.sg.f. maii; gen.sg.f.
max iia
(x cf. 11.28).
The 2nd person is formed in OAv. with a stem a- your (Skt. tv-, Lat.
tuus, Gr. ): nom.sg.m. ; gen.sg.m ahii; abl.sg.m. ahmt ; dat.sg.m.
ahmi; inst.sg.m.neut. ; loc.sg.m. ahm; nom.pl.m. i; nom.acc.pl.
neut. ; nom.sg.f. i; gen.sg.f. ax iia
; loc.pl.f. h.
Plural: The 1st person is formed with the stem ahmka- our (Skt. asma
ka-):
nom.acc.sg.neut. ahmkm; acc.pl.m. ahmk g; inst.pl.m. ahmki. For the
2nd person, two familiar variants (already commented upon in 22.1.4) are
found: xmka- (with the acc.sg.m. xmkm; nom.acc.sg.neut xmkm;
gen.sg.m xmkahii; dat.sg.m. xmki; inst.sg.m. xmk; acc.sg.f. xmkm)
and ymka- (with the acc.sg.m. ymkm; gen.sg.m. ymkahii; dat.sg.m.
ymki).
Concerning the pronominal stems OAv./YAv. xva- and YAv. hauua-, the FG
in hauua- may have been introduced from the pers. and dem.pron. in *ha-.
As regards its inflection, xva- follows the pronominal type (e.g., loc.sg.m. YAv.
xvahmi (cf. Skt. svsmin)), while hauua- follows the thematic type in all of
its forms (cf. 19.1), e.g., loc.sg.m. hauue (cf. Skt. sv); nom.pl.m. hauua
h;
gen.pl.m. hauuanm (Skt. sva
nm).
22.5 Interrogative and Indefinite Pronouns
The interrogative and indefinite pronouns in Av. are formed on the basis of the
stems ci-/ca- and ka- who?, some, the comparison of which with Skt. k-, k-
and Gr. , , and especially with OCS to, kto, allows for the reconstruction
of an IE pron. *k
u
id, *k
u
o-s. In principal, the accented pronouns functioned
as interrogatives, while the enclitics functioned as indefinites. However, the
indefinite function could also be indicated by postposing the IE particles *k
u
e
or *k
u
id; cf. Lat. quis-que, OP kas-ciy, Skt. ka-cit.
1. To ci-/ca-, the following forms are found: nom.sg.m. OAv. YAv. ci (cf.
Skt. k), ci-ca
someone, na-ci no one; acc.sg.m. YAv. ci
m, OAv. na-ci
m;
nom.acc.sg.neut. YAv. cit (cf. Skt. particle cit), na-cit , cm (cf. Skt. km), ci;
gen.sg.m.neut. OAv. cahii; dat.sg.m YAv. cahmi; loc.sg.m. YAv. cahmi;
nom.pl.m. OAv. caiiasc, YAv. caii; nom.acc.pl.neut. OAv. c-c, YAv. ci-ca.
2. The same inflection as in the pronouns a-/ima-, ta-, and ya- is found with
ka-, so there is no need to enumerate forms. Nevertheless, cf. the gen.sg.m.
OAv. kahii, YAv. kahe, ka he (Skt. ksya). Note further the concurrence of the
inst.sg.m.neut. forms Av. k and YAv. kana (cf. Skt. kna), the latter of which
seems to have been made after the inst.sg.m.neut. of the demonstrative ana-
(22.2.4; like the inst.sg.m. of a-/i-/ima- in 22.2.1).
23 prepositions and preverbs 77
23 Prepositions and Preverbs
A distinction between adnominal prepositions and preverbs exists only par-
tially, since the majority of the forms admit of bothuses. Prepositions canoccur
before the noun (i.e., as a preposition stricto sensu) or after it (as a postposi-
tion), though there is a small number of forms that always precede the noun.
The majority of preverbs immediately precede the verb.
In the following list, prepositions/postpositions are given, followed by the
case that they govern. Keep in mind that, in large part, these words may also
serve as adverbs, the sense of which can be derived without further difficulty
from the meaning given for the basic forms. Those forms that are exclusively
adverbial will be expressly indicated as such.
1. aiti (Skt. ti, Gr. ) adv. trans-, beyond; sideways.
2. aipi
(Skt. pi, Gr. ) + acc. above, during; on; + inst. after. [temporal]
3. aib, aii, auui, aoi (Skt. abh, Lat. amb-) + acc. towards; against; for; + loc.
around, about.
4. aairi (Goth. undar) + acc. below.
5. apa
(Skt. pa, Gr. , Lat. ab) adv. away, dis-.

6. ana (Gr. ) + acc. on, along.
7. n, anu (Skt. nu) + acc. towards, following
8. atar (Skt. antr, Lat. inter) + acc. between, inside.
9. (Skt. a
) + acc. to; in, on; + loc. on; + abl. since, (away) from, from; OAv.
+ inst. for, because; cf. further 36.
10. upa
(Skt. pa, Gr. , Lat. sub) + acc. about, towards; up to, into; + loc.
in.
11. upairi
(Skt. upri, Gr. , Lat. super) + acc. over, above; + inst. beyond
and in comparison.
12. us, uz- (Skt. d) adv. out, up.
13. OAv. tar , YAv. tar, tarasca (Skt. tirs) + acc. through, by way of; ex-
cept.
14. paiti
(Skt. prti, Gr. [, ]) + acc. against; by, with; through; +

loc. by, through.
15. pairi
(Skt. pri, Gr. ) + acc. around, about; + abl. from, (without).

16. para
(Skt. pura
) + acc., abl., or loc. before.

17. YAv. par (Skt. purs) + gen. infront of, before + abl. by; because of, since,
for.
18. pasca (Skt. paca
) + acc. behind; after; + gen. after, behind; + abl. or inst.

after.
19. fra
, fra- (Skt. pr, Gr. ) adv. along.

20. ni
, nii- (Skt. n ) adv. down.

21. ni
, ni- (Skt. ni-, nir-) adv. out.

22. mat (Skt. smt) + inst. with, together.
23. v, vi-, vii- (Skt. v ) adv. apart, separately, cf. Lat. dis-.
24. haca
(Skt. sc) + abl. (out) of, from; in relation to, according to, with the
agent of the passive also by; OAv. + acc. or inst. of; by (reason of).
25. had, haa (Skt. sah) + inst. together with.
26. hara
(Skt. satra
) + inst. together with

27. h m, h (n), ha-, hm(-), ham- (Skt. sm) adv. together, conjointly
24 The Verb
Just as is the case for nominal inflection (13), the study and description of the
Avestanverbal systemmust first be contextualized among the old Indo-Iranian
languages, thanks to whose comparative study we can reconstruct the pre-
history of the Indo-Iranian verb. The Avestan verb has retained some notable
archaisms, but at the same time also exhibits clear formal and functional inno-
vations with respect to earlier linguistic stages. The following sections will dis-
cuss both of these aspects, making constant reference to Sanskrit and, insofar
as possible, to Old Persian.
25 Component Elements
25.1 The Root. The Stem
The root is the element on which the verbal system is based, in that it deter-
mines the lexical content of the forms that are derived from it. The root is
usually subject to ablaut (13.2), thoughincertainverbal classes it is invariable,
such as in the case of denominative formations (e.g., nmax iia- do homage,
honor from nmah- homage, etc.; cf. 26.13). This method of morphological
characterizationis inheritedfromProto-Indo-European, andis of great interest
and help for comparison with other Indo-European languages, and especially
with equivalent Sanskrit forms, which are often very similar to the Avestan
forms.
1. Ablaut serves to characterize each inflectional form and to distinguish
verbal stems. The different types of ablaut patterns can show the root in: A.
FG/ZG, e.g., 3.sg. OAv. as-t, Skt. sti versus 3.pl. OAv. h-t, Skt. snti (< PIE
*h
1
s-ti : *h
1
s-nti). The FG usually appears in the root or in the infix/suffix of
the active singular indicative and injunctive, as well as throughout the entire
25 component elements 79
paradigmof the subjunctive, active and middle.B. LG/FG, e.g., in the s-aorist,
OAv. di-/di- to dis point (*-i-/*-ai-).
Some further examples of ablaut in the root are: bar bear (PIE *bher):
Av. pres.ind. bara-; caus.ind. braiia-; perf.opt. bariin;mr speak (PIE
*mleu H): Av. pres.act.ind. mraomi; pres.mid.ind. mruii;vax grow (PIE
*h
2
u es): Av. caus.ind. vaxaiia-; pres.ind. uxiia-; zan generate (PIE *enh
1
):
Av. pres.ind. zzana-; pass.ind. zaiia-; fut.ind. zhiia-.
2. The roots meaning canbe modifiedthroughthe use of preverbs (23): e.g.,
the root d- give; make forms several compounds: -d- accept, receive, take,
us-d- set up, install, paiti-d- confer, para-d- grant, fra-d- try; perform,
ni-d- gather, v-d- distribute. The preverb usually precedes its verb, whether
adjacent to or in tmesis from the verb. At times, the preverb appears to be
doubled. In late texts, a tendency toward fusion of the preverb and the verb
may be observed.
3. The combination of the root with certain affixes (infixes, suffixes, or some
combination of these) comprises a lexical unit called the stem (cf. 13.1).
Among the different verbal types, a fundamental divisioncanbe made between
thematic formations (i.e., formations built with the suffixal vowel -a- < PIE
*-e/o, known as the thematic vowel) and athematic formations (all the rest).
A characteristic feature of old Indo-European languages is the large number
of forms that can be built to a temporal stem, owing to the interaction of the
various verbal categories.
25.2 Diathesis
Avestan has inherited from Indo-European the distinction between the active
and middle voice, each characterized by a specific set of endings for all the
tenses and moods. The original function of the middle is to highlight the role
of the subject of the verbal action. Oftentimes, the middle can be translated
as a passive, e.g., act. a hat shot (an arrow), mid. a himnaiia
(arrows that
have been) shot. In other cases, the middle gives the verb reflexive (act. fra-
pinaoti fatten, mid. frapinuuata swell) or reciprocal (act. hacaiti accompa-
nies, mid. hacaite go together) value. It should be noted that, even in the
parent language, there are many verbs that occur only in a single diathesis, so-
called activa tantum or media tantum verbs: YAv. sate, Skt. te, Gr. , Hitt.
kitta.
The passive does not exist as a separate diathesis, but, as will be seen, the
present formation in -iia- has passive meaning. In addition, an isolated ending
of the 3rd person in -i, which goes back to IIr., is found in the aorist with
passive value: Av. vc was said (Skt. -vci), sruu was heard (Skt. ra
vi); cf.
32.1.c.
25.3 Tense
The expressionof tense is carried out throughthe usage of specific verbal stems
(25.1.3). Fromthe point of viewof tense, three mainstems are distinguishable:
present, aorist, and perfect. The future stem is, in fact, a present formation
(26.18).
The Avestan tense system, based on these three stems, derives from an
aspectual system that is still present in Greek and residual in Sanskrit. This
system opposed the imperfective aspect of the present stem (action seen in
its development) to the perfective aspect of the aorist stem (action seen as
complete in its entirety). The perfect was not part of this aspectual opposition,
and served to express a resultative (attained state). It is crucial to distinguish
between aspect, which is a binary opposition, and type of action (Aktionsart),
which is expressed by multiple formations that were always opposed to an
unmarked base. From a functional perspective, Old Avestan still retains a large
proportion of the old IE system of oppositions, while Young Av. has developed
more towards a temporal system based on the opposition present : injunctive
(impf.; 31, 37.2).
26 Present Stems
Athematic Formations
1. Root Presents: some roots have ablaut (25.1.1), e.g., ah- be, mr- speak,
i- go, while in others the stem is invariable: tac- run, hah- sleep, aoj-
speak.
2.a. Presents with partial reduplication (the root has FG/ZG ablaut): e.g.,
dad-/dad- give; put (the two verbs attested in Skt. as ddmi give
and ddhmi put fall together in Avestan), hi.hak-/hic- follow (Skt.
sak-/sc-).
2.b. Presents with total reduplication (intensives): reduplication with the en-
tire FG root, and ablaut FG/ZG in the root. Examples: carkr- exalt,
dadi- show, zaozao- call.
3. Presents in-n: the root contains aninfix withablauting -na-/-n- before the
final consonant of the root, e.g., vinad-/vid- find. Class 9 is a secondary
subtype of this type.
4. Presents in -nu: root (generally ani) in ZG + an ablauting suffix -nao-/
-nu- (< *-nau -/-nu-), e.g., surunauu-/surunu- hear (IE *leu ); krnauu-
/krnu- make, do.
5. Presents in -n: root (of se origin) in ZG + ablauting suffix -n-/-n- (<
*-naH-/-nH-): grn- seize, frn-/ frn- please.
26 present stems 81
These three infixed present formations can be traced back historically to a
single type that split apart onaccount of particular phonological developments
and synchronic reanalysis. Thus, class 4. in fact comes from infixed roots that
ended in -u-, whose stem final sequence *-na-u -/-n-u- became productive as a
independent suffix: e.g., to kar make, YAv. krnaoiti, Skt. kr ti. Meanwhile,
class 5. is usually limited to old se roots, i.e., *-na-H-/-n-H-: e.g., YAv. grniti,
Skt. gr bha
ti.
Thematic Formations
Thematic formations (characterized by suffixation of the thematic vowel -a-/
-- < PIE *-e-/-o-, which precedes the ending) were still productive in the
Avestan period. Fourteen different formations, which have precise semantic
functions only in part, can be distinguished.
6. Root in FG (root-accented; Skt. bhvati type), e.g., bara- bear, yaza- wor-
ship, haca- follow. This is the most productive type in Av.
7. Root in ZG (accent on the thematic vowel; Skt. tudti type), e.g., ia- set
in motion, rsa- cut.
8. Root in ZG and partial reduplication (Skt. pbati type), e.g., hita- stand,
nihia- sit down, jana- strike.
9. Root in ZG with -n- infix, e.g., krta- cut (Skt. kr ntti), hica- sprinkle
(Skt. sicti). This is a thematization of 26.3.
10. Root in FG + suffix *-i a-, e.g., jaiiia- pray, mainiia-/maiia- think (Skt.
mnya-).
11.a. Root in ZG + suffix *-i a- (root-accented, non-passive, Skt. type di
vyati),
e.g., OAv. drujiia-, YAv. drua- deceive (Skt. drhyati), vrziia- work
(Goth. waurkjan).
11.b. Root in ZG + suffix *-i - (passive, Skt. type nahyte), e.g., miriia- die (Skt.
mriyate). This type employs active or middle endings without distinction
in Av.
12. Root in ZG with reduplication + suffix *-i a-, e.g., iziia- desire (< i-iz-i a-),
yaiia- boil (< *i a-i -i a-).
13. Denominatives with suffix *-i a-, e.g., fuiia- rear livestock (Av. pasu-/ fu-
livestock), biaziia- heal (cf. Skt. bhij- healer), viixmainiia- make a
speech from viixman- assembly
14. Root in ZG + suffix -aiia-, e.g., guzaiia- hide, saaiia- seem, zbaiia- call.
This type includes some archaic OAv. causative forms with ZG: OAv.
urdiia- make cry, urpaiia- make sick.
15. Root in FG ( in an open syllable; 7.4.6) + suffix -aiia- (causative): rao-
caiia- light up, draiia- hold. This type exhibits a distinction between
roots with FG of the root, such as vaxaiia- make grow (Skt. vakya-),
and verbs with LG of the root, e.g., tpaiia- heat (Skt. tpya-). This dis-
tinction results from the fact that the root vowel of the causative was a
PIE *o, which has different outcomes depending on its position in the
syllable (i.e., Brugmann 7.4.6). It should be noted that the causative
(sub)type tpaiia- became very productive in Avestan. Cf., for example,
the causative of s enlarge: OAv. sauuaiia- (< PIE *ou h
1
-i e-), versus YAv.
suuaiia-, with LG introduced here.
16. Root in ZG + suffix -sa- (including the inchoative), e.g., prsa- ask (Skt.
prcchti), tafsa- become warm. The suffix -sa- continues the PIE suffix
*-ske/o-. Old present formations (non-inchoative, of the type jasa- to
gam), as well as Avestan inchoative formations, fall under this type.
17. Root inZG+ suffix -ha/sa/a- (desiderative). Two subtypes appear here: a.
simple: grfa- try to grab;b. with partial reduplication in i: dia- try
to deceive (cf. Skt. dpsati). The diffent forms of the suffix are historically
explicable from the IIr. suffix *-(H)sa- in different environments.
18. Root in FG + suffix -hiia/siia- (future), e.g., vaxiia- to vac speak (Skt.
vakyti), zhiia- to zan generate (Skt. janiyti).
27 Aorist Stems
1. Athematic root aorist (the root has FG/ZG ablaut). The root itself is the
stem of the aorist, e.g., jam-/gm-/ga- to gam go, come; d-/d- to the two
roots d give; put; srauu-/sru- to sru hear.
2.a. Thematic root aorist: ZG of the root + -a- (root-accented), e.g., vda- to vid
find, know, Skt. vda-; sa- teach to sh, Skt. ia-. In some formations,
FG of the root appears, but this generally occurs for phonetic reasons:
taa- (Skt. tka-), hana- (Skt. sna-).
2.b. Reduplicated thematic aorist: ZG of the root with partial reduplication in
a + suffix -a-, e.g., vaoca- (Skt. vca-) to vac speak, nsa- (*na-n-a-) to nas
disappear.
3. Sigmatic aorist (athematic): root with ablaut pattern LG (indicative and
active injunctive) / FG (other) + *-s-, e.g., ra
h- to r protect, bax- to
baj divide, raos- to urud- cry, m gh- to man think, di-/di- to dis
point.
30 personal endings 83
28 Perfect Stem
The perfect stem is an athematic formation with reduplication and ablaut (<
*o/zero) in the root. When the root contains or ends in i
, the reduplication
vowel will be i
; when the root contains or ends in u
, the reduplication vowel

will be u
.; in all other cases, the reduplication vowel is a (or rarely, ).

The root takes FG in the active singular (but in an originally open sylla-
ble, cf. Brugmann 7.4.6), FG throughout the entire subjunctive, but ZG in
all other forms. In contrast to the other two stems, the perfect is further char-
acterized by a set of special endings in the active and middle indicative.
Examples of perfect stems are diaii-/di- to d look, susru- to sru hear,
vauuac-/vaoc- to vac speak, bauuar-/bar- to bar bear, jam- to gamgo, come,
vuuarz-/vuurz- to varz work, dad-/d-/da- to d put; give, hazd- to
had sit.
The root vid find, know constitutes an exception in that it forms an (inher-
ited) perfect without reduplication: vad-/vid-/vd-.
29 Moods
Avestanpossesses five moods, each of whichhas its ownmeaning (cf. 37), and
which are expressed through special formations. The subjunctive and optative
are formed with suffixes on the three types of primary stems. The subjunctive
is formed with a suffix -a- and primary or secondary endings (without distinc-
tion?). The root always takes FG, eveninthe middle voice. The optative employs
the (ablauting) modal suffix -ii-/-- (< PIE *-i eh
1
/*ih
1
-) and secondary endings.
The present indicative uses primary endings, the indicative of the aorist uses
secondary endings, while the perfect indicative has its own special endings.
The injunctive does not carry the augment (31) and is served by secondary
endings (with the augment, the imperfect, which serves as the preterite for the
present, would be formed). The imperative has its own endings.
30 Personal Endings
Personal endings are an essential element for the expression of the various
verbal categories. Endings are added to the verbal stem and mark person (3),
number (3: singular, dual, plural), and diathesis (2). The endings are multi-
functional, in the sense that each one expresses several paradigmatic roles: for
example, -mahi
is a 1st.pers. plural active. InAvestan, just as inIE, there is a basic

table 16 The verbal endings of the active
Primary Secondary Imperative Perfect
1st sg. -mi
; -a
-m, -am -a
2nd -hi
, -i
*-h, -s, - -; -di
, -i -a
3rd -ti
-t , -t -tu
-a
1st du. -uuah -uu w.a.

2nd w.a. w.a. w.a. w.a.
3rd -t; - -tm w.a. -atar
1st pl. -mahi
-ma
-ma
2nd -a
-ta
-ta
w.a.
3rd -ti
, -ati
, -aiti -n, -at ; -r, -r -tu
-ar
, -r
distinction between primary endings (PE, used in the pres.ind. and, partly, the
subjunctive) and secondary endings (SE, used inthe pres.inj., the aorist and the
optative); in addition, there are other sets of endings for the imperative and the
perfect indicative. The subjunctive has a special ending just in the 1.sg.10
30.1 Active Endings
Be aware that some endings are not attested in Avestan; these are marked here
as w.a. (= without attestation). Those endings with variants conditioned by
phonetic environment are given with a comma separating the variants: for
example, the secondary ending of the 2.sg., which comes from IIr. *-s, appears
as - (mrao you said, cf. 11.20.5), as -s (before enclitic -ca, 11.20.1), and as
the outcome of *-h (in the thematic ending, e.g., jas you came, cf. 7.13.1). A
semicolon separates endings of different origin, such as those of the 3.du, of
which -t corresponds to the original 3.du. ending (cf. Skt. -tas), while - must
have been taken over from the 2.du. (cf. Skt. -thas). As regards the thematic
formations, it must be noted that some forms have been partially disfigured by
phonetic developments: e.g., the already cited 2.sg. - < *-a-h.
N.B.: a. Primary Endings: The agreement with the endings of Skt., sg. -mi, -si,
-ti and pl. -masi, -tha, -nti/-ati is evident. The 1.sg. thematic -mi exhibits an
10 The terms primary and secondary are traditional; descriptively, the primary endings
are derived from the secondary ones by adding an element -i, which in PIE was a marker
of current relevance (hic et nunc).
30 personal endings 85
< PIE *o (7.4.6); after ii, we do not find -mi but -emi; here * was analogically
replaced by *a; cf. 7.10.1. On the other hand, a 1.sg. ending - (Gr. -, Lat. -
< PIE *-oh
2
), exclusive to thematic formations, also occurs. This distinction
between the thematic and athematic conjugations is old, and ceases to be
productive in YAv., which adopts the ending -mi for both: barmi (and in
Sanskrit: bhrmi).The 1.du., in comparison with Skt. -vas, has taken the *-i
from the 1.pl.
b. Secondary Endings: The agreement of these endings with those of Skt., sg.
-(a)m, -s, -t, 1.du. -va, pl. -ma, -ta, -(a)n/-ur, is evident.The 3.du. -tm concurs
with OP -tam, but in Skt., -tam is the form of the 2.du., while Skt. -tm is the
form of the 3.du. In Iranian, the endings of the 2.du. were introduced into the
3.du.; the same happens in the middle (cf. 30.2.a).The endings -r and
-r are found only in the athematic optative: the -- is part of the suffix -ii-,
and the ending is -r(), which corresponds exactly to Skt. -ur (3.pl.aor.opt.
buiintogether withbuiir, Skt. bhyr). The 3.pl. thematic ending is -n: YAv.
pacaiin.
c. The subjunctive has a unique ending in the 1.sg., -ni
, which coexists with

-a
. The 3.du. and the 2.pl. have primary endings, the 1.du. and the 1.pl. have
secondary endings. The remaining persons take both PE or SE without dis-
tinction (but Skt. has the SE in the 3.pl.)The imperative has forms only for
the 2nd and 3rd persons of the sg. and pl. Just as in IE, the 2.sg. can take an
ending (thematic stems) or, in athematic stems, take the IE/IIr. ending *-
dhi > Gr. -i, Skt. -dhi/hi. The remaining Avestan endings also correspond to
those of Sanskrit, -tu, -ta, and -ntu. However, the ending that corresponds to
Skt. -tt, Gr. -, Lat. -t(d) is not found. The 2.pl. uses the secondary end-
ing.
d. The perfect has its own set of endings only in the indicative. These
endings, in large part, agree with the Skt. forms: sg. -a, -tha, -a, 1.pl. -ma. The
3.du. -atar comes out of *-atr , while Skt. -atur reflects *-atr , with the same
ending *-r that Skt. has in the 3.pl., -ur. Av. has the 3.pl. in -ar
< *-r (the *-r

appears only in the plupf. cikitr).
e. The participles of the present active are formed with the suffix *-nt-
(inflection in 18.3.); its feminine is marked with the suffix -- (19.3.): ht-,
f. hait-; iis, f. iieit.The perf.act.part. has its own suffix, -uuh-/-u- (inflec-
tion 18.8): vduuh-/vdu- (vid); dauuh-/dadu- (d); its feminine takes the
suffix -- (19.3.), e.g., jam- (gam; Skt. jagm-).
30.2 Middle Endings
The earlier remarks made in the paragraph above, 30.1, apply equally to
Table 17.
table 17 The verbal endings of the middle
Primary Secondary Imperative Perfect
1st sg. -e
, -i -i
; -a
-e
2nd -he
, - he
, -e
*-ha, -ha
, -a
-huua
, -vha
, -uua
w.a.
3rd -te
; -e
-ta
; -i
-tm; -m -e
, -oi
1st du. w.a. -uuaid w.a.
2nd w.a. w.a. w.a. w.a.
3rd -ate
, -ie; -ite
-atm; -tm w.a. -ite
1st pl. -maide
, maie -maidi
w.a.
2nd -duii, -e, -e -dm, -m -dm, -m w.a.
3rd -te
, -aite
; -re, -ire -ta
, -ata
; -rm -tm -re

N.B.: a. Primary Endings: The endings correspond, in large part, to the series
of endings in Skt., sg. -e (< *-ai ), -se (< *-soi ), -te (< *-toi ), 3.du. -ete (the-
matic) and -te, pl. -mahe (*-madhai ), -dhve, -nte/ate.The 3.sg. has a fur-
ther ending -e
, Skt. -e, which is used for verbs with stative meaning; it is the
same form as the ending of the 3.sg.perf.mid.The co-occurrence in the 3.du.
of -ate
and -ie (thematic endings) is explained by the fact that the end-
ing of the 2.du.mid. *-ai ai (Skt. -ethe) was adopted by the 3.du. (30.1.b).
The endings of the 3.pl., -re and -ire (< *-re; cf. Skt. -re) were used in PIE
for the stative. They originate in the PIE ending *-ro of the 3.pl. of the mid-
dle.
b. Secondary Endings: The endings correspond to the series of endings in
Skt., 1sg. -i/-a, 3.sg. -ta, 1.du. -vahi (< *-u adhi), 3.du. -etm (thematic)/-tm,
pl. -mahi, -dhvam, -nta/-ata and -(i)ram/-(i)ranFor the 3.du. -atm, it must
again be presumed that the final short vowel was introduced from the 2.du.
*-ai tam (Skt. has innovated with -ethm).In 3.sg., we find an ending -i
(Skt.
-i), which is utilized for the isolated passive aorist (25.2).The endings of the
3.pl. with -r- are endings of the stative, which continues the old PIE middle
endings in *-ro with a secondary *-nt, thus *-ront.
c. The subjunctive has a unique ending in the 1.sg. -ne, which coexists with
the old -i (*a-ai ), cf. Skt. -ai versus OP -naiy. The remaining persons of the
middle have primary endings.The imperative shows forms that correspond
to endings of Skt., 2.sg. -sva, 3.sg. -tm, 2.pl. -dhvam (= SE), 3.pl. -ntm. It is
possible that there may have been a passive form (parallel to the 3.sg. of
the passive aorist, 25.2): 3.sg. Av. -m (Skt. -m), but the evidence is not
conclusive.
31 the augment 87
table 18 The use of the augment
Unaugmented Augmented
Present inj. impf.
Aorist inj. ind.
Perfect ind. plupf. (absent in Av.)
d. In the perfect, there are very few forms attested, but they agree with the
corresponding forms of Skt.: 1.sg. -e, 3.sg. -e, 3.du. -te, 3.pl. -re.
e. The middle participle exhibits a suffix -na- (also -na- and -ana-; Skt.
-na-) in athematic forms and -mna- (OP -mna-, cf. Skt. -mna- versus Pli
-mina) inthematic forms. This distributionof allomorphs is due to the develop-
ment of the PIE suffix *-mh
1
no- > IIr. *C-m Hna- > *-Cna- (after a consonant)
versus *V-mHna- > *-Vmna- (after a vowel). The perf.part.mid. makes use only
of the suffix -na- (Skt. -na-).
31 The Augment
An accented prefix *- (*h
1
e-?), through which indicatives with preterital value
were made, existed in Proto-Indo-European. When added to an injunctive
present form (the basic, unmarked form), the augment transforms an injunc-
tive into an imperfect; if an augment is added to a perfect stem, it would forma
pluperfect. The preterital value of the aorist stem was also formally expressed
through aorist injunctive forms with the augment.
The augment is systematically employed in Classical Greek (pres. : impf.
), inSanskrit (bhrmi: bharam), and inOld Persian(barmiy: abaram);
its use in Avestan, however, seems to be much more limited, since the majority
of preterital forms donot showthe augment, the forms of the injunctive serving
to express the past tense. This use could be due, in many cases, to the particular
content of the texts. Inany case, the (new) functionof the injunctive inAvestan
is not very well understood.
In Old Av., there is no clear evidence for any imperfect forms, while in Young
Av., the few imperfect forms seem to have disparate functions (in some cases,
it can be interpreted as an anterior to the injunctive).
Descriptively, forms with secondary endings can bear or not bear the aug-
ment. In the case of forms without the augment, they are termed injunctives,
following tradition. In the case of forms that bear the augment, they are called
imperfects (in the present stem) or indicative (in the aorist stem). For its part,
the perfect, with its own set of endings, has both forms without the augment,
indicatives, and forms with the augment, pluperfects.
Furthermore, distinguishing betweenthe augment andthe preverbis often
difficult, on account of the partial phonetic confusion between a and in Av.,
depending upon phonetic environment (e.g., 7.3). Thus, a form paitiimraot
he addressed canreflect *paiti mraot (withthe preverb), *paiti amraot (with
impf.), or *paiti amraot .
32 Paradigms
In the following section, we present verbal paradigms consisting of attested
forms, with one form to exemplify each person, as well as the distinct phonetic
variants that an ending may exhibit (cf. e.g., N.B. under 30.1). The forms given
in the tables belong solely to YAv., except where expressly indicated to the
contrary using italics. Forms of the subjunctive with secondary endings appear
in bolditalic type; historically, these forms are more archaic, but they are in the
process of being replaced by primary endings in Avestan.
32.1 Athematic Root Present and Root Aorist
Given that the inflection of both of these formations is the same, they are pre-
sented together in Table 19. To provide examples of the paradigm, we offer
forms from the following present stems: ah-/h-/x -/s-/z-/- be, aii-/i-/y- go,
aoj- speak, h- be seated, is- control, rule, jan-/ ja-/n- kill, tac- run, t-/ta-
shape, create, p- protect, mrauu-/mr- speak, vas-/us- wish, v- blow, saii-
lie, sh- teach, stauu-/stu- praise, aii-/i- dwell. Likewise, we offer forms from
the following aorist stems: car-/kr- make, do, jam-/gm-/ga- come, caii-
/s
ii- perceive, ci-/ci- show, ci-/ci- gather, dar-/dr- hold, dars-/drs-

look, d-/d- give; put, bauu-/b- become, man- think, mi-/mi- throw,
yaog-/yuj- yoke, van- win, 1var- cover, 2var- want, varz- work, rd- make,
srauu-/sru- hear, hac-/sc- follow.
32 paradigms 89
table 19 The root present and root aorist: indicative and injunctive
a. Pres. Ind.act. Ind.mid. Inj./impf.act. Inj.mid.
1st sg. ahmi ni-ne, mruiie, aoji mraom aoj
2nd ahi, vai taxe, pa
h mrao pairii-ao
3rd asti, jaiti, mraoiti mrite, ite, aoxte; mraot , tt, s aogd, aoxta
mruiie, is
1st du. usuuah w.a. w.a. w.a.
3rd st w.a. impf. auuitm w.a.
(auua-i-)
1st pl. mahi mrmaide w.a. w.a.
2nd st w.a. mraot w.a.
3rd hti, yeiti, va
ti aojaite; a
hire, sare impf. auuin w.a.

b. Aor. Inj.act. Inj.mid.
1st sg. darsm w.a.
2nd ci, da
, var-c da
h, m gh
3rd j n, dt , mist mat
1st du. w.a. duuaid
3rd w.a. a-sruutm
1st pl. dm varmaid
2nd dt a-srdm
3rd gmn, dn, yj n v-s
iit
c. Passive Aorist
Within the aorist system, there is an isolated 3rd pers. indicative/injunctive
form that adds an ending -i
(Skt. -i) directly to the aorist stem, in which the

root seems to show LG (by Brugmann, 7.4.6?). The roots that exhibit this
formation are ci gather, vac say, and sru hear (and perhaps 2mr mistreat).
Forms of the 3rd.sg.ind./inj. of the passive aor. are: OAv. cuu, vc anda-uuc
(Skt. vci), sruu (Skt. ra
vi). Later, this ending is extended to other secondary

formations, such as the present stems rnu- (ar set in motion), jan kill, or the
perfect stem - say: YAv. rnuui, jaini, ii.
table 20 The root present and root aorist: other moods
d. Subj. Pres.act. Pres.mid. Aor.act. Aor.mid.
1st sg. fra-mrauua, fra-mra- mrauune; yaoj; varn m ni, varn
uuni aoji
2nd ah w.a. dh da
h
3rd ahaiti, ahat w.a. dit; dt yao-dit,
ciait
1st du. w.a w.a. w.a. jamat
3rd w.a. w.a. w.a. w.a.
1st pl. a
hm, aii-aiiama imaid darsma w.a.

2nd w.a. w.a. v-caiia daduii
3rd ahn, apa.tacin w.a. dait, rdt; jimn da
t, yaojat
e. Opt. Pres.act. Pres.mid. Aor.act. Aor.mid
1st sg. x ii m w.a. diim dii
2nd mruiia
mruu jamiia
d
3rd mruiit ; sht aojta jamiit ; vaint drt
1st pl. x iim w.a. jamiim, buiiama; mainimaid
sruum
2nd x iit w.a. diiata w.a.
3rd hiir w.a. jamiir; jamiin w.a.
f. Impv. Pres.act. Pres.mid. Aor.act. Aor.mid.
2nd sg. mrii, zd, id w.a. did kruu,
dhuu
3rd mraotu, ast w.a. dt w.a.
2nd pl. staota w.a. dt w.a.
3rd yatu, h t w.a. scat w.a.
g. Part. Pres.act. Pres.mid. Aor.act. Aor.mid.
ht-, f. hait-; mruuat- mruuna-, dat- sr(a)uuana-
nom.sg.m. iis, aojana-
f. iieit-
32 paradigms 91
32.2 Athematic Reduplicated Present
To provide examples of the paradigm, we offer forms of the following present
stems (with partial reduplication): kuxnu- (xnu greet), cikaii- (ci perceive),
dad-/dad- (d give; put), didaii-/ddi- (d watch), mim- (m measure), zaz-
/zaz- (z leave), zaz- (z rise), hi.hk-/hic- (hac follow). The rare forms with
total reduplication (26.2.b.) have the same endings.
table 21 The reduplicated present
Ind.act. Ind.mid. Subj.act. Subj.mid
1st sg. dami daie, dad dani yao-dane
2nd dahi w.a. zazhi; yao-da w.a.
3rd daiti daste, dazd yao-daiti; daat daaite
3rd du. w.a. zazite cikaiiat w.a.
1st pl. dadmah dadmaid dama hicamaid
2nd w.a. w.a. fra-mma w.a.
3rd dadait w.a. zazti, dan yao-date
Inj.act. Inj.mid. Opt.act. Opt.mid.
1st sg. dam, dam w.a. w.a. daie
2nd dada
w.a. dai; daiiia
daia
3rd dat , dids dasta, dazd dait ; daiiit daita, daidt
3rd du. w.a. w.a. daitm w.a.
3rd pl. -dadat w.a. daiiin; daiiir w.a.
Impv.act. Impv.mid. Part.act. Part.mid.
2nd sg. dazdi dasuu daat-, daat-; dana-, kuxnuuna-
yao-dat-
3rd dadt w.a.
2nd pl. dasta mz-dazdm
32.3 Presents in -n, -nu, and -n
In what follows, we present a synopsis of all types of presents with a nasal infix.
Membership in each specific type of formation will be expressly indicated (in
the text and in the tables) using the following numbering system: 1 presents in
-n; 2 presents in -nu; 3 presents in -n.
To provide examples of the paradigm, we offer forms of the following stems:
1. cina- to ci show, cinah-/c (< *ciN-; cf. 7.15) to ci gather, mrc-
to marc destroy, minas- to miias mix, vinad-/vid- to vid find; 2. rnauu-
/rnu- to ar set in motion, krnauu-/krnu- to kar make, do, tanauu-/tanu-
to tan stretch, dbnauu- to dab deceive, vrnauu-/vrnu- to var cover,
spanauu-/spanu- to spas watch, srinauu-/srinu- to sri lean on, surunauu-
/surunu- to sru hear, hunauu-/hunu- to hu press out; 3. grn- to grab
seize, drn- to dar tear, prn-/prn- to par fight, frn-/ frn- to fr please,
vrn- to var want, strn-/strn- to star deploy, scatter, hun-/hun- to h
impel.
table 22 Nasal presents
a. Ind.act. Ind.mid. b. Subj.act. Subj.mid.
1st sg. 1. cinahm 2. krnauuni; 2. krnauune
2. krnaomi tanauua 3. prnne;
3. frnmi 3. prne 3. frnni frni
2nd 3.

hunhi w.a. 2.

ni-srinauuhi w.a.
3rd 1. cinasti 1.mrte 2. hunauuat 2. vrnauuait
2. krnaoiti (*mrte) 3. frnt 3. prnite
3. grniti 2. vrnite
3. vrt
3rd du. 2. hunut
1st pl. 1. cmah 1. cmaide w.a. 1. cinamaide
3. friinmah <
*frnmahi
2nd 2.
+
spanu 1. mrgduii
3rd 1. mrciti 1. mrcait 2. krnaon 2. rnauuate
2. krnuuaiti 3. grnn
3. frnti
32 paradigms 93
c. Inj.act. Inj.mid. d. Opt.act. Opt.mid.
1st sg. w.a. w.a. w.a. 2. tanuiia
2nd 1. mina w.a. 2. surunuiia
w.a.
3.
x
strniia
?
3rd 1. cinas 2. hunta 1. ciit 1. mrcta
2. krnaot 3. vrta 2. krnuiit
3rd du. w.a w.a. w.a.
2nd pl. 2. dbnaot w.a.
3rd 1. vidn 3. vrnt 1. cii w.a.
2.
+
hunuiir
e. Impv.act. Impv.mid. Part.act. Part.mid.
2nd sg. 1. cd w.a. 2. krnuuat- 1. vidna-
2. krnii 3. drnat-, 2. hunuuana-
3. prn f. uruunait-
2nd pl. 2. srinaota w.a.
3rd 3. frntu w.a.
32.4 Thematic Presents and Aorists
To provide examples of the paradigm, we offer forms of the following stems:
pres. xraosa- shout, xaiia- rule, control, ga- hear, cara- go, jaiiia- pray,
jasa- come, juua- live, taca- run, tauruuaiia- conquer, daoiia- be deceived,
draiia- hold, ddraa- want to hold, riia- protect, prsa- ask, bauua-
become, bara- bear, nasiia- disappear, maniia- think, yasa- take, yaza- wor-
ship, sacrifice, yiiia- fight, vrca- pull apart, vrziia- work, vsa- serve,
sruuaiia- proclaim, srariia- lean, zaiia- be born, haca- follow, hita- stand;
aor. vaoca- say.
table 23 The thematic present and aorist
1st sg. barmi, jaiiiemi baire, -iiese barni, sruu- jasi
prs aiieni; xaii
2nd barahi, jaiiiehi hitahe, xaiiehe barhi; srraiia
prsa
he
3rd baraiti, nasiieti baraite jasiti; jast yazite
1st du. w.a. w.a. juuuua w.a.
3rd barat; yiiia vsate; fra-carie jast w.a.
1st pl. barmahi barmaide juumah; ba- w.a.
rma
2nd xaiia fra-carae; ddra- w.a. w.a.
.duii
3rd barti, bauuaiti, vste, riiete, bara
ti; barn zaiia
te
taciti, jaiiieiti hacite
c. Inj./impf.act. Inj./impf.mid. d. Opt.act. Opt.mid.
1st sg. barm, abaom baire w.a. isii
v-ram
2nd jas zaiiaha v-raiii yazaa
3rd jasat zaiiata jasit maniiata
3rd du. jasatm jasatm w.a. a-prsaiiatm
1st pl. aor. -uuaocm w.a. jasama vanimaid,
biiiimaie11
2nd tauruuaiiata w.a. fr-rsata rmim
3rd jasn, baon mainiiata, fraorci- baraiin, yazaiiata
ta, vsta kraiin
e. Impv.act. Impv.mid. f. Part.act. Part.mid.
1st sg. jasa, nase baravha, gahuu jasat-, barmna-,
vrziiat-/ hacimna-,
2nd baratu vrziitm vrzit- a-aoiiamna-
2nd pl. barata draiiam, gdm
3rd bartu xraostm
11 With primary ending!
32 paradigms 95
32.5 Sigmatic Aorist
To provide examples of the paradigm, we offer forms of the following aorist
stems: uruuax- (uruuaj walk), xnao- (xnu greet), ar- (ars cut,
shape), ra
h- (r protect), dar- (dar hold), da
h- (d put), di-/di-
(dis point), fra- ( fras ask), bax- (baj divide), na- (nas bring), na- (n
lead, guide), m gh-/m h- (man think), vah-/v gh- (van win), var- (varz
work), va-/va- (vaz go, lead), rax- (iric abandon, remove), rh-/ra
h- (r
bestow), ss- (sad appear), sta
h- (st stand), zh- (z leave).

table 24 The sigmatic aorist
a. Ind./inj.act. Ind./inj.mid. b. Subj.act. Subj.mid.
1st sg. w.a. fra; m gh var xnaoi, mghi;
varsn
2nd di w.a. w.a. ra
hahi
3rd drt, ss, vs xnaota, sta
haiti, varait varait; vaata

msta baxait; v ghat , naat
1st pl. w.a. a-m hmaid nm w.a.
2nd w.a. ardm w.a. maz-da
h.dm
3rd uruuxat w.a. vart; xnaon, vaa
te
v ghn
c. Opt.act. Opt.mid. d. Impv.act. Impv.mid.
1st sg. w.a. w.a.
2nd w.a. raxa di frauu
3rd fra-zaht w.a.
1st pl. nma w.a.
2nd w.a. w.a. sst rzdm
3rd aiin w.a.
e. Part.act. Part.mid.
vaht-? mahna-
32.6 Perfect
To provide examples of the paradigm, we offer forms of the following perfect
stems: d- (to ad say), h- (ah be), iiei- (i go), irri- (iri- die), urraod-
(urud obstruct), cikit-/cicit- (cit understand), jam- (gamcome), jar- (gar
wake), jaauruu- (grab seize), jiga- ( j live), dad-/d-/da-, pfr-/pafr-
(par fill), mamn- (man think), yaiiat-/yat-/yit- ( yat occupy a position),
vauuan-/vaon- (van win), vauuac-/vaoc- (vac say), vauuaz-/vaoz- (vaz go,
lead), vad-/vid- (vid find), vaorz- (uruuz be glad), rar- (r bestow),
susru- (sru hear).
table 25 The perfect
1st sg. jaauruua, vad susruiie w.a. w.a.
2nd dada, vauuaxa, vist w.a. w.a. w.a.
3rd jaauruua, jara, vaa, pafre, daie vaat
+
pafrite
a
ha; urraost
3rd du. yatatar mamnit w.a. w.a.
1st pl. ssrma, yama w.a. w.a. w.a.
2nd w.a. w.a. vaorza w.a.
3rd irriar, a
har; cikitr
+
raire vati, iieiin w.a.
c. Opt.act. Opt.mid. d. Part.act. Part.mid.
1st sg. jamiim w.a. irriu-, irrina-
2nd w.a. jiaa vuuh-/vidu-,
3rd vaoniit , a
ht w.a. f. jam-
3rd du. a
ht .tm w.a.
2nd pl. w.a. w.a.
3rd
+
dain; bariin vaozirm
33 Non-Finite and Nominal Forms
Some formations have a double nature, in that they simultaneously showprop-
erties of both verb and noun. Formally, they have nominal character in virtue
of being declined or bearing specifically nominal suffixes, while they belong to
the verbal systeminsofar as they are able to express tense, voice, and/or aspect.
33 non-finite and nominal forms 97
These formations include infinitives, participles (which have already been dis-
cussedabove; cf. 30.1.e, 30.2.e), as well as various verbal nouns andadjectives.
The most important formations will be treated in the following paragraphs.
1. The infinitive: The majority of formations categorized as infinitives show
an ending *-ai (OAv. -i, YAv. -e), which matches the ending *-ai of the dat.sg. in
nominal inflection (18.1.7); hence, one can infer that the infinitive originally
marked the goal of a verbal action. The Avestan infinitive occupies an interme-
diate positionbetweenthe original use as a verbal noun(taking part innominal
inflection) and use as an impersonal verbal form (being incorporated into the
verbal system).
The infinitive in -diii (Skt. -dhyai) is added to the root in the ZG (e.g.,
sridiii to sru hear), or to the present stem (e.g., riiidiii built on riia-
protect). This suffix provides a middle voice infinitive. Other suffixes that are
added either to the root or to the verbal stem are: -ti/-te: mrite to say, OAv.
sti to be; -uuani/-uua
ne
: OAv. vduuani to vid (cf. the Gr. inf. suffix -,

, Cypriot ); -uui/-uue: OAv. duui to give; the infinitive in -he:
OAv. vaoca h say is formed from the present stem.
2. The verbal action noun, formed with the accented PIE suffix *-t- and ZG
of the root, is also found in Avestan with the same function: krti- act, deed
to kar, maiti- thought to man. In Av., forms with FG of the root also occur, e.g.,
xvarti- food to xvar, which prove that the suffix remained productive until a
late period. The majority of nomina agentis are attestedinthe dat.sg. *-aiai (- e,
-aiiaca), and, given that the finality of the verbal action is thereby expressed,
this formation comes very close to usage as an infinitive.
3. The perfect passiveparticiple is formedwitha suffix -ta- addedtothe root
in the ZG: krta- done to kar, druxta- tricked to druj, basta- bound to bad,
gata- come to gam. This formation continues the old PIE verbal adjective with
ZGof the root andaccentedsuffix *-t-: cf. Skt. gat-, Gr. , Lat. uentus, from
PIE *g
u
m -t- come. However, the adjective -ta- was productive into younger
stages of the language. Hence, forms with FG occur, to se roots (i.e., in -), and
to some roots whose ZG formally obscured its paradigmatic connection to the
rest of the finite forms: dta- put to d (cf. Skt. hit- with ZG < *dhh
1
-t-), rta-
bestowed to r, tata- constructed to ta, yata- honored to yaz (cf. Skt. i-
with FG) and baxta- distributed to baj (Skt. bhakt-).
chapter 4
Syntax
34 Syntax
The functions of the various nominal and verbal formations are, in large part,
described in the chapter on morphology. In this chapter, details that concern
the functioning of the constituent elements in the phrase are the focus of
discussion.
35 Number
An important feature is agreement of the nom.pl.neut. with a verb in the
singular, which is constistent in OAv., but only sporadic in YAv., which prefers
the plural: Y 29.4 saxvr y z vuurzi (3.sg.perf.mid.ind. to varz) insults
that have been made. Avestan has preserved here an important archaism, in
which Sanskrit, Greek, and also Hittite share.
Dual forms are usually found in three different situations: 1. in conjunction
with the number duua- two: Yt 5.131 duua auruuata ysmi I ask for two
coursers; 2. in relation to a natural pair: Yt 10.91 frasntaibiia zastaibiia with
both hands washed, OAv. Y 32.10 vana h aibii in order to see with his eyes;
3. in a dvandva, i.e., a compound in which both members are coordinated with
one another. Each member of the compound takes the dual: Y 9.4 pasu vra
cattle and men, miraahura Mira and Ahura (cf. Skt. mitra
vru Mitra and

Varuna).
Old Av. always employs verbal forms in the dual to refer to two entities
together; in Young Av., the dual is disappearing, gradually being substituted by
the plural. The dual is, though, still regularly usedto refer to natural pairs andin
dvandvas: Y12.5 aprsatmmazda
sczarautrasc M. and Z. discussed. The

number two usually takes plural forms or, in late texts (cf. 40.2, 41.1), even the
singular: Vd 2.41 duua nara us.zaiieiti two men are born; Vd 7.52 duua mainiiu
r na auua.sta
hat the two spirits will begin battle.

100 chapter 4 syntax
36 Case Syntax
1. The nominative is the case of the subject, as well as the case of any predi-
cation that refers to the subject. The nominative can also assume the function
of the vocative, which has its own endings only in the singular. The vocative
is used for appellative functions, and serves to directly address a referent. If
several vocatives occur in a clause, they are usually coordinated asyndetically:
mazd as Mazd and As a!. The accusative functions as the direct object of
transitive verbs. Also used is the accusative of direction, in order to indicate the
end or the direction of a verbal action: Y 46.1 km zm aiien to which land will
I go?. The accusative of (temporal or spatial) extension is also a common func-
tion: Y 65.5 hminmca zaiianmca in summer and in winter, Vd 3.17 rigim
haca three steps from. Finally, the accusative of relation occurs: cuuat drj
howlong?, Yt 5.98 tmsuuita mighty in heritage. Meanwhile, constructions
with a double accusative (some of doubtless antiquitiy) are not foreign to Av.:
Y 11.5 y mm tat draon zint v trfiit v who leaves me this share or robs
me of it (cf. Skt. indro maruta sahasram ajint Indra left a thousand to the
Maruts; OP xaamim adam adinam I left the kingship to him).
2. The genitive is the case of nominal dependency, andthus serves to express
an enormous range of functions (appurtenance, material, partitive, temporal,
etc.), which are also present in the other IE languages: Y 4.23 tar pura
ahurahe mazda
fire, son of Ahura Mazd, Yt 10.98 y sit yazatanm who

is the quickest among the Yazatas, Vd 9.2 upa.arti uruuaranm he cuts (a
certainquantity) of plants. Onthe other hand, the genitive alsofunctions as the
obligatory complement of many verbs (is provide, x order, baj apportion,
etc.): Yt 17.21 upa m sraiiavha vs ahe lean on my cart, Y 58.4 f carkmah
we celebrate the cattle-herder.
The ablative case has its own forms only in part (14.15, 15.3, 16.6). The
ablative refers to a point of departure (origin): Yt 17.13 utra
h
+
uzaiiat
zmat camels that rise from the earth. The ablative functions as the obligatory
complement of many verbs (nas disappear, rah depart, mi deprive, etc.):
Y 9.28 yast m xat miat who deprives him of sovereignty. The ablative
is used especially with verbs of salvation, purification, and protection: Aog 57
yahmat haca naci bujaiit from which no one may be saved. An important
further use is as the proper case for the second element of a comparison: Y 45.11
nit mi vst xmat anii I have no other shepherd than you [pl.]
3. The dative governs a referent to whom the verbal action is directed. It
is used to indicate the person who is advantaged or disadvantaged by what
occurs: Y 29.10 aibii aog dt give them strength, Vd 15.12 ptrbii striieiti
he sins against his parents; it also indicates that the person has a stake in the
37 the use of verbal moods 101
outcome of the action (ethical dative): Y 45.5 yi mi ahmi sraom dn they
who gave him obedience for me. A temporal usage is also in evidence: Vd 4.45
maiiii asnmca xafnmca at midday and at midnight.
The instrumental expresses a referent that participates in the verbal action
withthe subject or the object; this refererent canbe instrumental proper, comi-
tative, causative, or agentive: Y 57.10 snaiia xvarm jaiti with the weapon,
he struck [him] a wound, Y 29.2 y drguu.db amm vdiiit who gives
chase to anger along with the liars, Y 47.2 ii cist for (the sake of) this idea.
In quasi-adverbial use, the instrumental indicates a feature or characteristic: Yt
5.61 mrahe khrpa with the body of a bird.
The locative governs a referent in which or even towards which the verbal
action takes place. For example, Y 11.6 ahmi nmne zaiia
te they will be born in

this house, Y 50.4 pai dmn on the way to the house. A temporal usage is
also in evidence: Vd 21.3 uahuua biaziit he should heal [it] in the morning.
37 The Use of Verbal Moods
1. The present indicative expresses a real and current state of affairs, whether it
be specific or general, whether a single or repeated action: Yt 17.17 k ahi y mm
zbaiiehi who are you who calls upon me?, Y 65.3 vspa
ima
p ya
zm paiti
frataciti all these waters which run across the earth. The perfect indicative
refers to the result of an action in the past, or an action that contributed to the
formation of a present state: Y 1.1 niuuaaiiemi dau y n daa I worship
on account of the creator who has created us, Y 9.1 k nar ahi yim azm
vspahe ah u astuuat sratm ddarsa who, o man, are you, the most
beautiful whom I have seen in all my mortal life?
2. The present and aorist injunctive refer to a real past event, without any
indication as to the perfectivity or imperfectivity of the verbal action. More-
over, there seems to have been no functional difference between the few aorist
indicative forms and the aorist injunctive. The injunctive is the mood of narra-
tionpar excellence; it is thus functionally equivalent tothe imperfect of Sanskrit
or Old Persian. It is used profusely in the great mythological and cosmogonic
narratives, and in doctrinal and didactic texts: Y 19.1 prsat zarautr ahurm
mazdmZ. askedAM. Independent of the moodof the mainclause, the injunc-
tive is always used in any subordinate clause that refers to an real past event.
The injunctive is the mood employed in prohibitive statements (39.2).
As was already mentioned in 31, it is difficult to find unmistakable forms
of the imperfect (such as, for example, forms belonging to ah be or i go).
Moreover, its use does not seem to differ fundamentally from the use of the
injunctive, although some rare passages may be noted in which the imperfect
expresses an action prior to the action expressed by the injunctive: Yt 5.68 tm
yazata jmsp yat spm pairi.auuanat Jmspa sacrificed to him once he
had caught sight of the army.
3. The present and aorist subjunctive indicate an action that has not yet
been initiated, and which will be actualized only if certain conditions hold.
From this point of view, the subjunctive takes on a voluntative value (very
close to a future) for the 1st pers. in main or independent clauses: Yt 19.47 atat
xvarn hagrfne I want to seize that xvarnah; in the 2nd pers., a strongly
prescriptive (almost imperatival) value is observable: Vd 18.51 imm t narm
nisrinaomi imm m narm nisrraiia
upa srm fra.krtm I entrust this

man to you; return this man to me upon [the day of] the mighty fra.krti;
in the 3rd pers. (sometimes also in the 1st), a consecutive (sequential) value,
depending on a previous process, can be felt: Yt 1.9 yazaa mm zarautra
jasni te auua haca rafna haca azmy ahur mazda
make sacrifices to me,

o Zarathustra I, who amAhura Mazd, will come to you in aid and assistance
[of you]; Vd 2.23 riat ca g u apa.jast and a third of the cows will perish.
The subjunctive is used to ask about an action which has not yet been
initiated: Vd 8.12 cii at ate masma ahn yaibii frasnaiia
te varsa
sca
tanmca which are the (kinds of) urine with which hair and body should be
washed? The perfect subjunctive is a rare YAv. formation, which expresses
prospective anteriority: Yt 8.39 titrm yazamaide y pairika
titraiieiti ya
uza
hat ar mainiiu we worship Titria who passes [i.e., kills] the Pairikas,
which Angra Manyu would have sent up.
4. The optative instantiates a number of disparate functions. a. Above all,
the optative is used in a main or independent clause with direct speech, and
its usage implies that the speaker wants for the verbal action to be realized by
the interlocutor (cupitative): Yt 10.23 apa n haca zahat mira bari bring us
away from fear, o Mira. A related usage is the prescriptive optative, which is
very common in the Vidvdd: Vd 21.5
x
uziiara uzra huuarxata arise, get up,
o solar orb. b. The optative also serves in the expression of comparisons and
similies involving an imaginary action: Yt 8.55 titrii pairikm darzaiieiti
mnaiin ahe yaa hazarm narm im narm darzaiiit Titria holds
the Pairik like a thousand men would hold a single man. c. Likewise, the
present optative (almost always with the augment) functions as a past tense
iterative. This usage has left traces in many Old, Middle, and Modern Iranian
languages; this functioncould, therefore, go back to Proto-Iranian. Vd19.27 kuua
t dra paiti hajasti mas ii astuuaiti ahuu hauui urune para.daiiit
where do the donations go, which the mortal has been offering for his own
spirit throughout his bony life?
38 clausal syntax 103
d. The aorist optative occurs only indirect speech. This usage, termedpreca-
tive, seems to have been the product of a Proto-Indo-Iranian specialization:
Y 70.4 yaa a vcim nma because we would bear the word with the liba-
tion.
e. The perfect optative has, as its primary function, the expressionof the past
irrealis: Yt 8.24 yeii z m yazaiiata, auui mm auui.bariim dasanm
aspanm aoj if they would have made a sacrifice to me I would have
acquired the strength of ten horses.
5. The imperative expresses orders and commands of an immediate char-
acter: Vd 2.4 at m gaa
fraiia at m gaa
varaiia make my creatures

prosper, make my creatures grow.
38 Clausal Syntax
1. In the Avestan manuscripts, no sign that serves as a marker of interrogation
or exclamation appears. Nevertheless, wh-questions are usually introduced by
an interrogative pronoun (cf. 22.5) or adverb: OAv. kad, YAv. ka when?;
kaa
how?; OAv. k, kud where?; YAv. kudat from where?; kua how?;
kura
where?, k(a)uua where?. Yes/no-questions are usually accompanied by

specific particles, e.g., kat (Skt. kt): Vd 6.42 kat t haoma
+
yaodaiin ahn
should these haomas be purified?
2. Coordination. Twoor more constituents or clauses canbe conjoinedusing
the enclitic particle -ca
(Gr. -, Lat. -que) or the word uta: Y 57.3 ahe raiia

xvarnahaca through his wealth and abundance. The particle v introduces
disjunctive coordination: Y 46.10 n gn v man or woman. The OAv. particle
at , at ct , YAv. at , introduces a proposition that contrasts with what was pre-
viously said, while aa
indicates a consequence of a previous statement. Av.

z (Skt. h ) expresses causality, Vd 5.36 juu z sptahe mainii u dmanm
hm.raaiieiti because (while it is) alive, it mingles with the creatures of the
beneficient spirit.
3. Subordination. Subordination is introduced by relative pronouns (22.3)
or by conjunctions.
a. In explanatory clauses, OAv. hiiat , YAv. yat (Skt. yt) or yaa
(Skt. yth)
that; because, since is normally used: Yt 5.34 auuat iiaptm dazdi m yat
bauuni aii.vaniia
am dahk m do me this favor that I may be victor

over Ai Dahka.Final or consecutive clauses also begin with hiiat /yat or
yaa in order that; that, so that: Vd 3.32 yat yauu pouru bauut aa mrm
taa mairiit in order that the grain be plentiful, this mantra should now be
recited.
b. Concessive clauses are served by yat cit although (Skt. yc cid): Yt 10.85
ye he vx v hapta karuun jasaiti yat cit nmaha vcim baraiti whose
voice spreads throughout the seven karvars, although he raises his voice
with reverence.Comparative clauses also begin with Av. yaa
like, as (with
correlatives aa
, auuaa, etc.): Vd 2.32 at yim auuaa krnaot yaa dim iat

ahur mazda
then Yima did just as AM. commanded him.

c. The formation of conditional clauses depends on the type of statement.
The usual conjunction is yezi, but hiiat /yat , yaa
also appears. When dealing

with a necessary relation (reality), the protasis usually occurs in the indicative
and the apodosis in any mood other than the optative: Vd 18.76 yezi azite tm
ahm paiiiite yim as aonm if he receives [the penance], then he will arrive
to the existence of the followers of As a; yeii z huuar nit uzuxiieiti if the
sun does not rise. An irrealis relationship is expressed using the conjunction
ye(i)i (z) if (Skt. ydi) with the perfect optative (37.4.e): Yt 8.52 yeii z
azm nit daiiim aom strm pairika ahuum auuahisiiit if I had not
created that star, the Pairika would have entered into existence. A contingent
relation is established through use of the optative in both the protasis and
apodosis.
d. For local subordinate clauses, yara
(Skt. ytra) where, and again yaa,

are used: Y10.4 staomi zmyaaraoahe hubaoii I praise the earthwhere you
grow fragrant.Temporal clauses are introduced with hiiat /yat or OAv. yad
(Skt. yada
). Other temporal conjunctions with more or less specific meanings

also exist: yezi
as soon as: Yt 19.43 yezi bauuni prniiu zm caxrm kr-

nauune as soon as I become older, I will make the earth my wheel; and also
constructions such as para ahmt yat before, vspm ahmt yat until and
while, yauuata
until, while, at yat or yauuat (Skt. ya
vat since), pasca yat

after: Vd 16.8 yezi nirika vohun aii.vant yat h rii xafna saca
te h
nihiata vspm ahmat yat h car xafna saca
te if a woman notices
blood when three nights have passed she should lie down until four nights
pass.
39 Negation
1. The basic negative particle is nit not; neither, nor: Yt 15.16 yimahe xare
nit aotm a
ha nit garmm during the reign of Yima, there was neither

cold nor heat. A repeated negation can be carried out with the OAv. particle
nad, YAv. naa: Y11.6 nit ahmi nmne znaite rauuanaeraata
naa
vstrii fuiis in this house will be born neither a priest nor a charioteer nor a
cattle-herding farmer.
40 the system changes 105
The forms nit and naa are built from the IE negative particle *nei , which
is attested in OP naiy, Lat. n, OCS ni-. In Avestan, this particle is preserved in
the first syllable of the indefinite pronoun naci no one, nothing: Y 43.6 y m
naci dbaiieit whomno one deceives, Yt 1.6 yat mm naci tauruuaiit that
no one defeat me.
2. The particle m (= Skt. ma
, Gr. ) negates a command (with the imper-

ative or the injunctive) or a desire (with the optative; equivalent to nit ). In
Av., an inhibitive sense predominates:1 Yt 17.60 m auui asmanm fras
usa do
not move towards heaven, H 2.17 m dim prs yim prsahi stop questioning
himwhomyou question, Vd. 6.2 mca tmzmkraiinm p hrzaiin and
do not let them sow this land, let them not run water (through it); Y 31.17 m
uuduua
aip.d buuaiiat that the ignorant one ceases to lie.

3. To negate a noun or an adjective, the privative prefix a/an/ana- is fre-
quently used; its functions are similar to the prefixes un- (inherited Germanic),
in- (borrowed fromLatin and Romance languages), and a-/an- (borrowed from
Greek) in English. The origin of all three lies in the IE privative *n-, which is
attested in Gothic un-, Latin in-, Gr. -/-, etc. Just as in Greek, IIr. shows two
variants, *aC- (preconsonantal) and *anV- (prevocalic): Av. ams a-, Skt. amr ta-
immortal from *n -mr -ta-, Av. an-as a- unmilled from *n-rta-. This prefix was
still productive in Young Avestan. Any participle could be negated using this
prefix: anauuavhabdmna- who does not not remain asleep (pres.part.mid.
of auua-xvabda-), afratat kuuah- which does not flow (perf.part. fra-ta-tk-).
40 The System Changes
Not all Young Av. texts represent the same phase of the language. Indeed,
they were composed in a living language, subject to the same processes of
change as any other language. Some examples that illustrate how the system
has undergone simplification have already been pointed out above (cf., e.g.,
the diffusion of the ablative -t under 19.1.6). Other innovations include the
following:
1. For purely formal reasons, feminine -stems come to follow the inflec-
tion of neuters in -ah, on account of the coincidence of -a
as the ending of
the nom.acc.pl. in both inflectional types: YAv. han- army is twice found
1 Recall that prohibition is done in Skt. with the particle ma
and the aorist injunctive: ma
no
rria do not let us suffer harm. Cf. further Gr. do not do that! versus
refrain from doing that!.
inflected as a neuter, cf. Y 57.25 pairi druuat bii han bii, where the form dru-
uat bii leaves no doubt as to the neuter gender of han bii.
The neut. noun va- herd (from *van win) attests an inflection in - in
late texts. This inflection is probably due to the fact that va- is primarily
used as a collective (the herds the livestock), whose nom.acc.pl. va is
formally identical to the nom.sg. of an -stem.
The acc.pl. paa
to pat-/pa- path, instead of the original pa (cf. ch. 3, fn.

2), supplies another example. The substitution in this case probably proceeds
from the acc.sg. patm, which was interpreted as the acc.sg. of a fem. in -. In
turn, the ending of the acc.pl.f. -a
was, by extension, applied to the weak stem

pa-, from which paa
arose.
2. A further semantic motivation has promoted the development of the
m. stem raat- driver, charioteer (Skt. ratheha
-) into a nomen agentis

raatar-. Since the nom.sg. of the root noun and of the nomina agentis (e.g.
dt giver) had an identical ending, the necessary formal condition was avail-
able for a change of inflectional type, and was sustained by semantic affin-
ity: acc. raatrm, nom.pl. raatr; later, with thematicization: gen.sg.
raatrahe, acc.pl. raatr sca.
A semantic reason for morphological change is also evident from the ten-
dency, found in late texts, not to decline numerals: Vd 14.91 mat riss aii.ari
with thirty metal spikes, Vd 4.11 tir sata upzananm three hundred blows.
These cases follow the model of the majority of the numerals (21.1), which
were always indeclinable.
chapter 5
Texts
41 Introduction
The Avestan texts, which have been transmitted only in manuscript form, are
traditionally divided into several books, which constitute a single canon for the
believers who preserved the recited text. Each book is usually subdivided into
chapters. The most extensive book is the Yasna (Liturgical Prayer), which is
in turn divided into 72 chapters or hitis; the book principally contains texts
for the yasna liturgy, i.e., prayers and mantras that were recited during the
celebration of the religious ceremony. In the middle of the Yasna lie the gs
(Gathas, songs) of Zarathustra and the Yasna haptahiti (the Yasna of Seven
Chapters), the only texts in Old Avestan.
Other purely ritual books include collections of lesser scope, such as the
Vsperad (All of the Ratus, 24 chapters), the Nyyin (Elegy, 5 chapters), the
Gh (Parts of the Day, 5 chapters), the Srza (The 30 Days, 2 chapters), and
the frngn (Thanks, 4 chapters). The Yat (Hymns, 21 chapters) are each
dedicated to anAvestandeity and are, inlarge part, writteninpoetic form, with
verses of eight syllables. Their content is mythological, at times encomiastic,
andthus allows us to see into the backgroundinwhichMazdayasnianismcame
into being.
The Vdvdd (The Anti-Demon Law, 22 chapters) narrates the origin of the
world and of humanity. Later, the book turns to the enumeration of religious
rules and laws that a believer must observe inorder to be a goodMazdayasnian.
The book is composed in prose with some small poetic remnants.
Some relatively minor fragments, such as the Nrangestn (Cultic Prescrip-
tions), the Pursinh (Questions), the Hdxt Nask (Book of Scriptures), and
the Frahang m (an Avestan-Pahlavi dictionary), also exist, but they have
passed through a transmission even more deficient than the rest of the texts.
41.1 Young Avestan Texts
A. Yat 8 relates the battle between the star Titria and the demon Apaoa.
Titria is the star Sirius, which announced the arrival of the rainy season.
108 chapter 5 texts
Yt 8.13 paoiriia
dasa xapan | spitama

zarautra | titrii rauua
xvarnavha
| khrpm raaiieiti |
raoxnuuua vazmn | nar
khrpa paca.dasah | xatahe
spiti.dirahe | brzat auui.amahe
| amauuat hunairiia
c
The first ten nights, o Spitama
Zarathustra, splendid, beautiful
Titria mixes his body flying among
the lights in the form of a fifteen
year-old man, radiant, with white
eyes, tall, very strong, vigorous,
dextrous.
8.14 taa aiio yaa paoirm | vrm
auui.ya
bauuaiti | taa aiiao yaa

paoirm | vrm auui.am aiti | taa
aiiao yaa paoirm | vrm rzum
adaste
He is then of the age when the girdle
first comes onto a young man. He is
then of the age when strength first
comes into a young man. He is then
of the age when a young man is given
to puberty.
8.15 h ira viixmaniiete | h ira
prsaniieiti | k mm nrm
friiazite | gaomauuaitibii
haomauuaitibii zaorbii |
kahmi azm dam | vriim tm
vriim vm | hauuaheca urun
yaodrm | nrm ahmi yesniiasca
| vahmiiasca avhe astuuaite | as t
haca yat vahitt
He now proclaims, he now asks:
Who will now honor me with
libations of milk and haoma? To
whom will I give wealth in men, a
flock of men, purification of his own
spriti? Now I am to be worshipped
and to be praised for the bony life
according to the best truth.
8.16 bitiia

xvarnavha
| khrpm raaiieiti |
raoxnuuua vazmn | g u khrpa
zaranii.sruuahe
The second ten nights, o Spitama
the lights in the form of a bull having
golden horns.
haca yat vahitt
sprit? Now I am to be worshipped
41 introduction 109
8.18 ritiia

xvarnavha
| khrpm raaiieiti
| raoxnuuua vazmn | aspahe
khrpa auruahe | srrahe
zairi.gaoahe | zaranii.aiinahe
The third ten nights, o Spitama
the lights in the form of a white
horse, lovely, having yellow ears,
having a golden bridle.
haca yat vahitt
sprit? Now I am to be worshipped
8.20 at paiti auuiti | spitama zarautra
| titrii rauua
xvarnavha
| auui
zraii vourukas m | aspahe khrpa
auruahe | srrahe zairi.gaoahe |
zaranii.aiinahe
Then he descends, o Spitama
Titria, towards the lake Vourukas a,
in the form of a white horse, lovely,
having yellow ears, having a golden
bridle.
8.21 dim paiti.y niduuaraiti | dauu
y apao | aspahe khrpa smahe |
kauruuahe kauruu.gaoahe
| kauruuahe kauruu.barahe |
kauruuahe kauruu.dmahe | daahe
aiit.tarti
The demon Apaoa runs against him
in pursuit, in the form of a black
horse, hairless with hairless ears,
hairless with a hairless back,
hairless with a hairless tail, reddish,
frighteningly harnessed.
8.22 hm tcit bzu barat | spitama
zarautra | titriiasca rauua
xvarnavha
| dauuasca y apao |
ta
yiiia spitama zarautra

| ri.aiiarm ri.xaparm |
dim bauuaiti aii.aoja
| dim
bauuaiti aii.vaniia
| dauu y
apao | titrm rauuatm
xvarnavhtm
The two intermingle their legs, o
Spitama Zarathustra, splendid,
beautiful Titria and Apaoa the
demon. They fight, o Spitama
Zarathustra, for three days and three
nights. He becomes overwhelming
in might, he becomes victorious,
Apaoa the demon over splendid,
beautiful Titria.
110 chapter 5 texts
8.23 apa dim at viieiti | zraiiahat haca
vourukas t | hr.masahm
aanm | sdrm uruuitrmca
nimrite | titrii rauua
xvarnavha
| sdrm m ahura mazda |

uruuitrm p uruuara
sca | baxtm
dane mzdaiiesne | nit mm
nrm mas iika aoxt.nmana yasna
yazte | yaa aniie yazata
h |
aoxt.nmana yasna yaziti
He chases him away from there, from
the lake Vourukas a, for the distance
of a hra in length. Defeat and
retreat! utters splendid, beautiful
Titria. Defeat to me, o Ahura
Mazda, retreat, o waters and plants,
disgrace, o Mazdayasnian religion!
Men do not now worship me with an
invocatory sacrifice, as the other gods
are worshiped with an invocatory
sacrifice.
8.24 yeii z m mas iika | aoxt.nmana
yasna yazaiiata | yaa aniie
yazata
h | aoxt.nmana yasna
yazinti | auui mm auui bariim |
dasanm aspanm aoj | dasanm
utranm aoj | dasanm gauum
aoj | dasanm gairinm aoj |
dasanm apm nuuaiianm aoj
For if men had worshiped me with
an invocatory sacrifice as the
other gods are worshiped with an
invocatory sacrifice, then I would
have obtained the might of ten
horses, the might of ten camels,
the might of ten bulls, the might of
ten mountains, the might of ten
abundant waters.
8.25 azm y ahur mazda
| titrm
rauuatm xvarnahatm |
aoxt.nmana yasna yaze | auui dim
auui.barmi | dasanm aspan aoj |
dasanm utranm aoj | dasanm
gauum aoj | dasanm gairinm
aoj | dasanm apm nuuaiianm
aoj
I, Ahura Mazd, worship splendid,
beautiful Titria with an invocatory
sacrifice. To him I bring the might of
ten horses, the might of ten camels,
the might of ten bulls, the might of
ten mountains, the might of ten
abundant waters.
Yt 8.26 and 8.27 = Yt 8.20 and 8.21, respectively.
8.28 hm tcit bzu barat | spitama
zarautra | titriiasca rauua
xvarnavha
| dauuasca y apao |
ta
yiiia spitama zarautra

| rapiinm zruunm |
dim bauuaiti aii.aoja
| dim
bauuaiti aii.vaniia
| titrii rauua
xvarnavha
| dam yim apaom

The two intermingle their legs, o
Spitama Zarathustra, splendid,
beautiful Titria and Apaoa the
demon. They fight, o Spitama
Zarathustra, until midday. To him
great might comes, to him comes
victory, splendid, beautiful Titria
over Apaoa the demon.
41 introduction 111
8.29 apa dim at viieiti | zraiiahat haca
vourukas t | hr.masahm
aanm | utattm nimrauuaite
| titrii rauua
xvarnavha
|
uta m ahura mazda | uta p
uruuara
sca | uta dane mzdaiiesne

| uta bauut da hauu | us
v apm aauu | apaiti.rta
jasa
ti | a.dnunmca yauuanm |
kasu.dnunmca vs tranm |
gaanmca astuuaitinm
He chases him away from there,
from the lake Vourukas a, for the
distance of a hra in length. Good
fortune! utters splendid, beautiful
Titria. Good fortune to me, Ahura
Mazda, good fortune, o waters and
plants, good fortune, o Mazdayasnian
religion! There will be good fortune, o
countries! Your water canals will
run over unimpeded, (those) of the
wheat of plump grain, (those) of the
meadows of fine grain, and (those) of
the bony creatures.
B. One of the most important animals in the Avesta is the dog. Chapter 13 of the
Vdvdd enumerates the prescriptions that believers must observe regarding
the animal.
Vd 13.12 dtar gaanm astuuaitinm as um y spnm jaiti yim
pasu.haurum frazbaoahm snam vkrt.utnm: k h asti cia.
at mraot ahur mazda
: ata sata upzananm upzit aspahe atraiia

ata sata srao.caranaiia
Creator of the bony world, true one, he who strikes a blow on a shep-
herd dog which takes away (its) consciousness, which cuts away (its) life:
with what is he punished? Thus spoke Ahura Mazd: 800 lashes with the
whip, 800 with the scourge, will be imposed on him.
13.13 dtar etc. y spnm jaiti yim vi.haurum frazbaoahm
snam vkrt.utnm: k h asti cia. at mraot ahur mazda
: hapta
sata upzananm upzit aspahe atraiia ata hapta srao.caranaiia
Creator of the bony world, true one, he who strikes a blow on a guard
dog which takes away (its) consciousness, which cuts away (its) life: with
what is he punished? Thus spoke Ahura Mazd: 700 lashes with the whip,
700 with the scourge, will be imposed on him.
13.14 dtar etc. y spnm jaiti yim vohunazgm frazbaoahm
snam vkrt.utnm: k h asti cia. at mraot ahur mazda
: xuua
sata etc.
Creator of the bony world, true one, he who strikes a blowon a hunting
112 chapter 5 texts
13.15 dtar etc. y spnmjaiti yimtaurunmfrazbaoahmsnam
vkrt.utnm: k h asti cia. at mraot ahur mazda
: paca sata etc.

Creator of the bony world, true one, he who strikes a blow on a young
13.16 hu jaao hu vzao h sukurunahe hu
x
urupi tii.dtahe
hu
x
raopi
x
yaon.xvaptahe hu vspanm spt.mainiiauuanm
spaciranm aniia udra uppa
This (is the punishment for the death) of a hedgehog, this () of a vzu,
this () of a porcupine, this () of a weasel having sharp teeth, this () of
a fox asleep in its den, this () of all the canines of the beneficient spirit,
other than the aquatic otter.1
13.17 dtar etc. kuua asti sp pasu.hauruuditii.gtu. at mraot ahur
mazda
: y yjiiastm haca gabii pariti sraomn tii vhrkmca

Creator etc., where is it that the shepherd dog is in its proper place?
Thus spoke Ahura Mazd: (with him) who goes a yjiiasti in length from
the herds, listening for thieves and the wolf.
13.18 dtar etc. kuua asti sp vi.hauruu ditii.gtu. at mraot ahur
mazda
: y hr.masahm aanm haca vibii pariti sraomn

tii vhrkmca
Creator etc., where is it that the guard dog is in its proper place? Thus
spoke Ahura Mazd: (withhim) whogoes the distance of a hrainlength
from the settlement, listening for thieves and the wolf.
13.19 dtar etc. kuua asti sp vohunazg ditii.gtu. at mraot ahur
mazda
: y nacim isaiti hunaranm tanuiie isaite rrm

Creator etc., where is it that the hunting dog is in its proper place?
Thus spoke Ahura Mazd: (with him) who seeks no skill, (but) requires
protection for his body.
1 The punishment for the death of an otter is more severe, and is addressed in Vd 14.
41 introduction 113
13.20 dtar etc. y spnm tar.pim daste yim pasu.haurum, cuuat
atams
iiaonanmstraiieiti. at mraot ahur mazda
: yaa atahmi
ahuuyat astuuaiti fratm.nmnahe nmn.paitmpaiti tar.pim
daiiit, aa striieiti.
Creator etc., he who leaves a shepherd dog without sufficient food,
how much does he sin (on account of) these actions? Thus spoke Ahura
Mazd: as if in this life, which is bony, the lord of a foremost house were
left without sufficient food, he sins that much.
13.21 dtar etc. y spnm tar.pim daste yim vihaurum, cuuat
atams
: yaa atahmi
ahuu yat astuuaiti mam.nmnahe nmn.paitm paiti tar.pi-
m daiiit, aa striieiti.
Creator etc., he who leaves a guard dog without sufficient food, how
much does he sin (on account of) these actions? Thus spoke Ahura
Mazd: as if in this life, which is bony, the lord of a middling house were
left without sufficient food, he sins that much.
13.22 dtar etc. y spnm tar.pim daste yim vohunazgm, cuuat
atam s
iiaonanm straiieiti. at mraot ahur mazda
: narm bit
ia as auuanm jastm ahmiia nmne mat auuabii daxtbii yaa
rauua paiti tar.pim daiiit aa striieiti
Creator etc., he who leaves a hunting dog without sufficient food,
how much does he sin (on account of) these actions? Thus spoke Ahura
Mazd: as if in this life, which is bony, no less than a truthful man who
enters into his house with his appearance like a priest were left without
sufficient food, he sins that much.
13.23 dtar etc. y spnm tar.pim daste yim taurunm, cuuat
atams
: yaa atahmi
ahuu yat astuuaiti aprniikm dahm.krtm s
iiaonuuar-
zm vrziit s
iiaonm paiti tar.pim daiiit aa striieiti

Creator etc., he who leaves a young dog without sufficient food, how
much does he sin (on account of) these actions? Thus spoke Ahura
Mazd: as if in this life, which is bony, a working child, the offspring of
a dahma who does work, were left without sufficient food, he sins that
much.
(Vd 13.2427 lays out the punishment for the offenses commited in Vd 13.2023,
as 200, 90, 70, and 50 lashes, respectively.)
114 chapter 5 texts
13.28 atm z atahmi ahuu yat astuuaiti spitama zarautra, sp-
tahe mainii u dmanm sitm zauruunm upiti yat spn yi
hitte axvar upa xvartm par spasn uuidn, par xuuisca
zitica g u mat baratu xvaranm snahe, auua he ditii.pim
Because in this life, which is bony, o Spitama Zarathustra, the fastest
among the creatures of the beneficient spirit to go into old age, as regards
dogs, are those which stand without eating alongside one who is eating,
lying in wait without receiving anything. Bring milk and fat along with
meat as the food of the dog. That is its proper food.
13.29 dtar etc. yat ahmi nmne yat mzdaiiasni sp auuaca
v bauuat
aitii.xratu, kua t vrziin ate yi mazdaiiasna
Creator etc., if in a house which is of a Mazdayasnian there is a mute
dog [scil. a dog that bites], lacking proper judgement, how shall those
Mazdayasnians act here?
13.30 at mraot ahur mazda
: auua h baraiin ttm duru upa tm

manaorm, stamanm h at niizaiin ati.mas xraoduuahe bi
atauuat varduuahe.
Thus spoke Ahura Mazd: they shall place a cut piece of wood about
its neck, they shall close shut its mouth with this, an ati in length in case
of a hard piece, twice of that in length in case of a soft piece.
13.31 atahmt cit nidarzaiin, fra hmcit nidarzaiin yezi nit sp
auuaca
v aitii.xratu pasm v narm v raaiit

They shall tie it [scil. the wood] onit [scil. the mouth], they shall secure
him with it, so that the mute dog lacking proper judgment does not harm
the livestock or a man.
41.2 The Yasna Haptahiti
Inthe middle of the Yasna (chapters 35 to41) is foundthe only OAv. text inprose
that we possess. Its liturgical style is marked by parallelism and repetition.
Y 37.13 i t yazamaid ahurm mazdm, y gmc as mc dt , apasc
dt uruuara
sc vavh, raoca
sc dt bmmca vspc voh | ahii

xarc maz nc hauuapahica. t mat yasnanmpauruuatt yaza-
maid, yi g u hac s
iieit | t m at hirii nm n mazd.var

+
spt.tm yazamaid. t m ahmki
+
azdbc utnic yaza-
maid. t m as unm frauuas narmc nirinmc yazamaid.
41 introduction 115
So now we worship Ahura Mazda, who created the cow and the truth,
(who) created the waters and the good plants, (who) created the lights
and the earth and all good things | through his control and greatness
and skillfulness. We worship him with the excellence of the sacrifices
that dwell alongside the cow. | We worship him with the Ahurian names,
which Mazd chooses, the most beneficient ones. We worship him with
our bodies and lives. We worship him in the Fravais of the truthful ones,
of the men and of the women.
41.3 The Gathas (Gs)
Chapters 2834, 4351, and 53 contain the Gathas, which are poetic texts in var-
ious meters making a total of 855 verses. As regards their content and language,
the Gathas represent the oldest formof Avestan. Given that Zarathustra speaks
in the first person in some poems, and on account of the special religious con-
tent of the texts, it is thought that the Gathas may have been composed by the
prophet Zarathustra himself. Although such a claim is unverifiable, the value
of the Gathas for Iranian religious history and for comparative linguistics is, on
all points, inestimable. It should be kept in mind that the form of the verses,
and the compact and metaphorical style employed therein, makes the Gathas
almost always difficult to interpret.
In Yasna 43, Zarathustra tells his god Ahura Mazd how he received his
prophetic calling from the Beneficient Spirit. Each strophe contains five verses
of 4+7 syllables.
43.5 sptm at mazd m h ahur |
hiiat ah u zi darsm
paouruum | hiiat da
iiaoan
mdauun yc ux | akm aki
vavhm as m vahauu | hunar
dmi uruuas ap m
I realized that you are beneficient, o
Ahura Mazd, because I see you as
the first one in the generation of
existence, because you establish
deeds and words as having prizes:
bad for the bad, a good reward for the
good, through your skill, at the final
turn of creation.
43.6 yahm spt mainii uruuas
jas | mazd xar ahm voh
manah | yehii s
iiaoani gaa
as frdt | aibii rat s ghait

rmaiti | ahii xrat u y m naci
dbaiieit.
At the turn to which you come
through your beneficient spirit, o
Mazd, by your power, I am with
good thought, through whose deeds
the creatures prosper along with
truth. Right-mindedness explains to
them the judgments of your thinking,
which no one can deceive.
116 chapter 5 texts
hiiat m voh pair.jasat manah |
prsat ca m ci ah kahii ah | ka
aiiar daxr frasaiii d | aib
h gah tanuuc
AM, when he surrounds me with
good thought and asks: Who are
you? Whose are you? Why, o daxr,
would you submit some day to
questioning regarding your creatures
and yourself?
43.8 at hi aoj zarautr paouruum |
haiii duuaa
hiiat isii
drguuit | at as un rafn
x ii m aojhuuat | hiiat
+
+
bt
vasas.xarahii dii | yauuat
mazd staom ufiic
I respond to him then, first (I am)
Zarathustra. And that I being
truthful, would seek hostilities
against the liar. Thus I would be a
mighty support to the truthful, if I
might gain the adornments of the
one who rules according to his own
will, inasmuch as I praise and extol
you, o M.
ahii fras m kahmi vuuduii va
| at ahmi r rtm nmah |
as ahii m yauuat isi mainiii
good thought (and asks me): For
whom do you wish to examine the
consultation of this?For your fire,
a gift of reverence, for the truth, I will
understand as much as I can.
43.10 at t mi di as m hiiat m
zaozaom | rmait hacimn t
rm | prsc na
y ti hm
part | part m z
+
yan tat
mauuatm | hiiat xaiis am
diit mauuatm
Show me then the truth that I
continue to call to myselfin the
company of right-mindedness I have
come for itand ask us what has
been asked to us by you, for what is
asked by your strength is like that of
the strong, if the ruling one may
make you vigorous and strong
43.11 sptm at mazd m h
ahur | hiiat m voh pair.jasat
manah | hiiat xm uxi dda h
paouruum | sdr mi ss mas iia
zarazditi | tat vrziieidiii hiiat
mi mraot vahitm
good thought. When I learn of the
first thing through your [pl.] words,
faith in mortals seems disappointing,
in order to carry out that which you
tell me is best.
41 introduction 117
43.12 hiiat c mi mrao as m jas
frxnn | at t mi nit asrut
pairiiao |
+
uzridiii par hiiat
mi .jimat | srao as mz.raii
hacimn | y v as rnibii sauui
[v]diit
And when you tell me you reach
truth in your foresight, then you
do not impose on me things that
can be disobeyed, so that I arise
before obedience will come to me,
accompanied by reward having great
wealth, who will distribute rewards
on both sides in abundance.
ar vizdiii kmahii t m mi
dt | dargahii yao y m va
naci
drt it | vairiia
sti y ahm
xari vc
good thought, to achieve the
aspirations of (my) desire, you give
me that of a long life, which no one
holds you [pl.] to go (through with),
that of a desirable existence, which is
said to be within your control.
43.14 hiiat n friii
+
vadmn isuu daidt
| maibii mazd tauu rafn
frxnnm | hiiat xar
as t hac frt |
+
uzridiii az
sardana
s ghahii | mat ti vspi

yi ti mra
mart
Because a man who receives it and
has it available would give it to a
friend, (give) me your forethoughtful
support, o Wisdom. Because (this)
is attained with your rule in the
company of truth, in order that I
rise up together with all those who
memorize your mantras, to expel
those who challenge your teaching.
daxat us
iii tunmaitis vahit

| nit n pour drguuat x iit
cixnu | at ti vsp g agr g
as aon dar
good thought. A peaceful mind
teaches one to say what is best; a
man should not wish to satisfy the
many liars. These ones say that all the
wicked ones are truthful.
43.16 at ahur huu mainiim zarautr |
vrt mazd yast cic sp nit |
astuuat as m x iit utn aojhuuat
| xv g darsi xari x iit rmaiti |
as m s
iiaoani voh daidt

manah
O Ahura, this Zarathustra chooses
the spirit, o Mazd, whichever of
yours is the most beneficial. May the
bony truth be strong through its
vitality. May right-mindedness in its
reign be in the sight of the sun. Grant
reward through actions by means of
good thought.
Bibliography
Electronic Editions
http://titus.uni-frankfurt.de/texte/etcs/iran/airan/avesta/avest.htm
Avestan Digital Archive: http://www.avesta-archive.com
Chrestomathy
Reichelt, Hans, Avesta Reader. Texts, Notes, Glossary, and Index. Strasbourg 1911.
Translations
Darmesteter, J. & Mills, L. , The Zend-Avesta, Oxford 18801887 (repr. 1972).
Darmesteter, J., Le Zend-Avesta, Paris 18921893 (repr. Paris 1960).
Wolff, Fritz, Avesta: die heiligen Bcher der Parsen, bersetzt auf der Grundlage von
Chr. Bartholomaes altiranischemWrterbuch, Strasbourg 1910.
Secondary Literature and Annotated Bibliographies
Duchesne-Guillemin, Jacques, L tude de l iranien ancien au vingtime sicle,
Kratylos 7 (1962) 144.
Kellens, Jean L avestique de 1962 1972, Kratylos 16 (19711973), 130; addenda et
corrigenda, Kratylos 18 (1973), 15.
L avestique de 1976 1990, Kratylos 36 (1991), 131.
Schlerath, Bernfried, Awesta-Wrterbuch. VorarbeitenI: index locorumzur Sekundr-
literatur des Awesta. Wiesbaden 1968.
Tremblay, Xavier, Iranian Historical Linguistics in the Twentieth Century, Indo-
European Studies Bulletin 11 (2005), 123.
Iranian Historical Linguistics in the Twentieth CenturyPart Two Indo-
European Studies Bulletin 13 (2008), 151.
Edition
Geldner, Karl, Avesta. The Sacred Books of the Parsis, Stuttgart 18861896 (repr. 1991).
120 bibliography
Partial Editions with Commentary
Gershevitch, Ilya, The Avestan Hymn to Mithra, Cambridge 1959.
Hintze, Almut, Der Zamyd-Yat, Wiesbaden 1994.
AZoroastrian Liturgy: The Worship in Seven Chapters (Yasna 3541). Wiesbaden
2007.
Humbach, Helmut, Die Gathas des Zarathustra. III, Heidelberg 1959.
The Gths of Zarathushtra and the Other Old Avestan Texts. In collaboration
with J. Elfenbein & P.O. Skjrv, Heidelberg 1991.
Humbach, Helmut & Faiss, Klaus, Zarathutra and his Antagonists: A Sociolinguistic
Study with English and German Translation of his Gathas, Wiesbaden 2010.
Humbach, Helmut & Ichaporia, Pallan R., Zamyd Yasht. Yasht 19 of the Younger
Avesta. Wiesbaden 1998.
Insler, Stanley, The Gathas of Zarathustra (= AcIr 8), Tehran / Lige 1975.
Kellens, Jean & Pirart, ric, Les textes vieil-avestiques, Wiesbaden 19881991.
Monna, Maria C., The Gathas of Zarathustra, Amsterdam 1978.
Narten, Johanna, Der Yasna Haptahiti, Wiesbaden 1986.
Panaino, Antonio, Titriya. Part I: The Avestan Hymn to Sirius, Rome 1990.
West, Martin, Old Avestan Syntax and Stylistics. With an edition of the texts, Berlin/
Boston 2011.
Grammars and Lexicon of Avestan
Bartholomae, Christian, 1. Vorgeschichte 2. Awestasprache und Altpersisch, Grun-
dri der Iranischen Philologie I, 1, ed. W. Geiger / E. Kuhn, Strasbourg 18941895.
Altiranisches Wrterbuch, Strasbourg 1904 (various repr.)
Beekes, Robert S.P., A Grammar of Gatha-Avestan, Leiden 1988.
Benveniste, mile, Les infinitifs avestiques, Paris 1935.
Boyce, Mary, Zoroastrians, New York 1979.
Cantera, Alberto (ed.), The Transmission of the Avesta, Wiesbaden 2012.
Duchesne-Guillemin, Jacques, Les composs de l Avesta, Lige / Paris 1936.
Hoffmann, Karl, Zum Zeicheninventar der Avesta-Schrift, in Festgabe deutscher
Iranisten zur 2500 Jahrfeier Irans, 1971, 6473.
Avestan Language, Encyclopedia Iranica III, 1 (1987) 4762.
Hoffmann, Karl &Forssman, Bernhard, Avestische Laut- undFlexionslehre. 2., durch-
gesehene und erweiterte Auflage, Innsbruck 2004.
Hoffmann, Karl & Narten, Johanna, Der sasanidische Archetypus: Untersuchungen
zur Schreibung und Lautgestalt des Avestischen, Wiesbaden 1989.
Kellens, Jean, Les noms-racines de l Avesta, Wiesbaden 1974.
bibliography 121
Le verbe avestique, Wiesbaden 1984.
Avesta in Encyclopedia Iranica III, 1 (1987) 3444.
Avestique, in CompendiumLinguarumIranicum (ed. R. Schmitt), Wiesbaden
1989, 3256.
Liste du verbe avestique. Avec un appendice sur l orthographe des racines aves-
tiques par Eric Pirart, Wiesbaden 1995.
Mayrhofer, Manfred, Iranisches Personennamenbuch. Band I: die altiranischen Na-
men, Vienna 1977.
Etymologisches Wrterbuch des Altindoarischen, Heidelberg 19861996.
Rastorgueva, V.S. & delman, D., Etimologieskij slovar iranskix jazykov (Etymo-
logical Dictionary of the Iranian Languages), Moscow 2000.
Reichelt, Hans, Awestisches Elementarbuch, Heidelberg 1909.
Schmitt, Rdiger, Die altiranischen Sprachen im berblick, in: Compendium Lin-
guarumIranicarum (ed. R. Schmitt), 1989, 2531.
Skjrv, Prods Oktor, Old Iranian and Old Persian Morphology, in: Morphologies of
Asia and Africa (ed. A.S. Kaye) Winona Lake 2007, 853940.
Old Iranian in: The Iranian Languages (ed. G. Windfuhr), London / NewYork
2009, 43195.
de Vaan, Michiel, The Avestan Vowels, Amsterdam 2003.
On the Iranian Languages in General
Cheung, Johnny, Etymological Dictionary of the Iranian Verb, Leiden 2007.
Schmitt, Rdiger (ed.), CompendiumLinguarumIranicarum, Wiesbaden 1989.
Sims-Williams, Nicolas, Eastern Iranian Languages in: Encyclopedia Iranica VII/6,
1996, 649652
Skjrv, Prods Oktor, Iraniv. IranianLanguages and Scripts, in: EncyclopediaIranica
XIII/3, 2006, 344377
Windfuhr, Gernot (ed.), The Iranian Languages, London 2009.
Yarshater, Ehsan (ed.), Encyclopedia Iranica, London 1982.
The Reconstruction of Proto-Indo-European
Beekes, Robert S.P. & de Vaan, Michiel, Comparative Indo-European Linguistics. An
Introduction. Second edition, revised and corrected by Michiel de Vaan, Amsterdam /
Philadelphia 2001.
Cowgill, Warren & Mayrhofer, Manfred, Indogermanische Grammatik. I. Einlei-
tung, Lautlehre, Heidelberg 1986.
122 bibliography
Fortson, Bejamin, Indo-European Language and Culture. An Introduction. 2nd ed.,
Oxford 2010.
Meier-Brgger, Michael, Indogermanische Sprachwissenschaft. 9th ed., Berlin 2010
[8th ed. translated as Indo-European Linguistics, Berlin 2003].
Reviews of Introduccin al avstico (2001)
Hintze, Almut, in: Kratylos 50 (2005), 200203.
Huyse, Philip, in: Abstracta Iranica 25 (2002). See abstractairanica.revues.org/
document4116.html.
Tremblay, Xavier, in: Bulletin de la Socit de Linguistique de Paris 100 (2005), fasc. 2,
144145.
Yakubovich, Ilya, in: Journal of Indo-European Studies 29 (2001), 476481.
Glossary
In the following pages, we offer a glossary of the words contained in the texts
of 41. The numbers refer to the corresponding nominal or verbal class. Young
Avestan forms are unmarked. In verbs, active voice is unmarked.
The alphabetical order followed is: a a
e o i u k x x xv g ()
c j t d t p f b n v n m m () y v r s z s
s h.
a-/i- dem.pron. this 22.2.1: gen.sg.m. ahii
(OAv.), dat.pl.m.n. aibii, loc.sg.m.n. ahmi,
ahmiia
auua- one 21.1; this 20.2
ata- that there 22.2.2: nom.sg.n. atm,
abl.sg.m.n. atahmt , loc.sg.m.n. atahmi,
nom.du.n. ate, gen.pl.m.n. atam
atauuat- as, so, so much 18.3: gen.sg.n.
atauuat
aa- vigorous 19.1/2: acc.sg.m. am
aoxt.nman(a)- invocatory (lit.
speaking the name) 18.4/19.1: inst.sg.m.
aoxt.nmana
aoj- speak 32.1: 1.sg.pres.mid.inj. aoj (OAv.)
aojah- n. might 18.7: acc.sg. aoj
aojhuuat- (OAv.) mighty, strong 18.3:
nom.sg.n. aojhuuat
aiiao- iiu-
aiiar- n. day 18.6: acc.sg. aiiar (OAv.)
aib prep. about, regarding + loc. 23.3
aii.aojah- overwhelming in might 18.7:
nom.sg.m aii.aoja
aiit.tarti- frighteningly harnessed 19.5:

gen.sg.m. aiit.tarti
aii.vaniiah- victorious 18.7: nom.sg.m
aii.vaniia
auua adv. towards

auua- dem.pron. that there 22.2.5: nom.sg.m.
huu (OAv.), nom.sg.f. hu, inst.pl.f.
auuabii
auuacah- mute 18.7: nom.sg.m. auuaca
auui- descend: 3.sg.pres.ind. auuiti

auui prep. + acc. towards 23.3
auui.ama- very strong 19.1: gen.sg.m.
auui.amahe
auui.bar- bring: 1.sg.pres.ind. auui.barmi
32.4, 1.sg.perf.opt. auui.bariim 32.6
aurua- white 19.1/2: gen.sg.m. auruahe
aka- bad, wicked 19.1/2: acc.sg.n. akm,
dat.sg.n. aki
axvar(a)- without eating: nom.pl.m. axvar
aa so, thus 37.2
adaste d-
adt from there
aitii.xratu- without proper judgment
19.5: nom.sg.m. aitii.xratu
au- f. canal 19.5: nom.pl. aauu
aan- m. distance 18.4: acc.sg. aanm
at (OAv.): so, then
apa adv. away, dis-
apaoa- m. Apaoa 19.1: nom.sg. apao,
acc.sg. apaom
apaiti.rta- unimpeded 19.1/2: nom.pl.f.
apaiti.rta
apasc p-
apm p-
aprniika- child, youth 19.1/2: acc.sg.m.
aprniiukm
ap ma- final 19.1/2: loc.sg.m. ap m (OAv.)
ahuu ahu-
ah u ahu-
avhe ahu-
aniia- other 19.1/2, 20.2.: nom.pl.m. aniia,
nom.pl.m. aniie
agra- (OAv.) evil, wicked 19.1/2: acc.pl.m.
agr g
ama- m. strength 19.1: nom.sg. am
amauuat- vigorous 18.3.: gen.sg.m.
amauuat
ar- come 32.1: 1.sg.aor.inj. rm (OAv.)
124 glossary
ara- n. intent, aspiration 19.1: acc.pl. ar
(OAv.)
ast- n. bone 18.1: inst.pl. azdb (OAv.)
asti ah-
astuuat- bony 18.3: nom.sg.n. astuuat ,
dat.sg.m. astuuaite, loc.sg.m. astuuaiti,
gen.pl.f. astuuaitinm
aspa- horse 19.1: gen.sg. aspahe, gen.pl.
aspanm
aspiia- of a horse, pertaining to a horse 19.1/2:
acc.sg.f. aspiim
asruta- disobeyed 19.1/2: acc.pl.n. asrut
(OAv.)
aza- drive, impel 32.4: az 32.1
azm pers.pron. of the 1st pers.sg. 22.1: acc.sg.
mm, acc.sg.encl. m, gen.dat.sg.encl. m,
mi (OAv.), dat.sg. maibii (OAv.)
azdb ast-
ata eight 21.1
ati.masah- of an ati in length 18.7: acc.sg.n.
ati.mas
atr- f. whip 19.2: inst.sg. atraiia
a.dnu- having plump grain 19.5: gen.pl.m.
a.dnunm
as a- n. truth 19.1: acc.sg. as m, gen.sg. as ahii
(OAv.), abl.sg. as t , inst.sg. as
as auuan- truthful 18.4.: voc.sg. aum,
acc.sg.m. as auuanm, dat.sg.m. as un
(OAv.), acc.pl.m. as aon, gen.pl.m.
as unm (OAv.)
as i- reward 19.5: acc.sg. as m, inst.sg. as ,
acc.pl. as
ah- be 32.1: 1.sg.pres.ind. ahmi, ahm (OAv.),
2.sg. ah (OAv.), 3.sg. asti, 1.sg.pres.opt.
x ii m (OAv.), 3.sg. x iit (OAv.)
ahii a-/i-
ahu- life, existence 19.5: gen.sg. ah u
(OAv.), dat.sg. avhe, loc.sg. ahuu
ahura- m. Ahura 19.1: nom.sg. ahur, voc.sg.
ahura, ahur (OAv.), acc.sg. ahurm
ahmka- our 19.1: inst.pl.m. ahmki
ahmi ah-; a-/i-
ahmiia a-/i-
ahm ah-
adv. against, until, towards, for + acc. 23.9
at so, then, but
iiu- n. age, life 19.5: gen.sg. yao (OAv.),
aiiao
tar- n. fire 18.5: dat.sg. r (OAv.)
rauuan- m. priest 18.4: nom.sg. rauua
d- say 32.6: 3.pl.perf.ind. dar (OAv.)
t (OAv.) then, so, for 37.2
p- f. water 18.1: voc.pl. p, acc.pl. apas-c,
gen.pl. apm
rm ar-
rmaiti- f. right-mindedness 19.5: nom.sg.
rmaiti, inst.sg. rmait (OAv.)
sita- swiftest 19.1/2: nom.sg.n. sitm
-straiia- make oneself sinful 32.4:
3.sg.pres.ind. straiiaeiti
striia- sin 32.4: 3.sg.pres.ind. striieiti
ziti f. fat 19.5: nom.sg. ziti-ca
hiriia- ahurian, pertaining to Ahura 19.1/2:
acc.pl.n. hirii (OAv.)
uuidan- without receiving anything 18.4:
nom.pl.m. uuidn
rzu- f. puberty 19.2: acc.sg. rzum
mauuat- (OAv.) strong 18.3: acc.sg.m.
mauuatm, gen.pl.m. mauuatm
hm vam
i-/aii- go 32.1: 3.sg.pres.ind. aiti, inf. it (OAv.)
it i-/aii-
i (OAv.) thus, so
ira here, now
ia here
isa- seek 32.4: 3.sg.pres.ind. isaiti,
3.sg.pres.mid.ind. isaite, 1.sg.pres.mid.opt.
isii (OAv.)
is(a)- be able 32.1/4: 1.sg.pres.mid.subj. isi
(OAv.)
isuuan- (be/have) available 18.4: nom.sg.m.
isuu
t i-/h-
ti- f. wealth 19.5: acc.sg. tm
i-/h- encl.pers.pron. of the 3rd pers. 22.1.6:
acc.sg.m.f. dim, acc.sg.f. hm, acc.sg.n. t ,
gen.dat.sg. h, hi (OAv.)
uxa- n. word 19.1: acc.pl. ux (OAv.), inst.pl.
uxi
udra- m. otter 19.1: nom.pl. udra
glossary 125
upa prep. around, (up) to + acc. 23.10
upi- = upa + i-/aii- approach, draw near to
(+ acc.): 3.sg.pres.ind. upiti
uppa- aquatic 19.1: nom.pl. uppa
upza- (= upa + aza-) impose 32.4:
3.sg.pres.opt. upzit
upzana- n. lash 19.1: gen.pl. upzananm
ufiia- extol 32.4.: 1.sg.pres.ind. ufii-c
(OAv.)
uruuasa- m. turn 19.1: loc.sg. uruuas
(OAv.)
uruuan- m. soul, spirit 18.4: gen.sg. urun
uruuar- f. plant 19.2: voc.pl. uruuara
s-ca,
acc.pl. uruuara
s-c (OAv.)
uruuitra- n. retreat 19.1: acc.sg. uruuitrm
urupi- m. weasel 19.5: gen.sg. urupi
urun uruuan-
us adv. outside, above
uz-ar- raise, rise: mid.inf. 32.1 uzridiii
(OAv.)
uta- wish 19.1: nom.pl. uta
utatt- f. good fortune 18.2: acc.sg. utattm
utna- m./n. life, vitality 19.1: inst.sg. utn
(OAv.), inst.pl. utni
utra- m. camel 19.1: gen.pl. utranm
us
iii vac-
ka- who? 22.5: nom.sg.m. k, nom.sg.f. k,
gen.sg. kahii (OAv.), dat.sg.m. kahmi
kauruua- bald, hairless 19.1/2: gen.sg.m.
kauruuahe
kauruu.gaoa- having hairless ears 19.1/2:
gen.sg.m. kauruu.gaoahe
kauruu.dma- having a hairless tail 19.1/2:
gen.sg.m. kauruu.dmahe
kauruu.bara- having a hairless back
19.1/2: gen.sg.m. kauruu.barahe
ka (OAv.) how, why?
kasu.dnu- having fine grains 19.5: gen.pl.n.
kasu.dnunm
kma- m. desire 19.1: gen.sg. kmahii (OAv.)
khrp- f. body, form 18.1: inst.sg. khrpa,
acc.sg. khrpm
kuua where?
kua how
xraoduua- hard 19.1: gen.sg.n. xraoduuahe
xratu- m. thought 19.5: gen.sg. xrat u (OAv.)
xara- n. rule, reign, control 19.1: inst.sg.
xar (OAv.), loc.sg. xari (OAv.)
xapan- f. night 18.4: acc.pl. xapan
xaiia- rule 32.4: nom.sg.m. of pres.part.
xaiis
xuua six
xuuid- m. milk 18.1: nom.sg. xuuis-ca
xm ym
x iit ah-
x ii m ah-
xvara- eat 32.4: acc.sg.m. of pres.part.
xvartm
xvara- n. food 19.1: gen.pl. xvaranm
xv g- huuar-
gairi- m. mountain 19.5: gen.pl. gairinm
ga- f. creature 19.2: nom.pl. gaa
, gen.pl.
gaanm, abl.pl. gabii, loc.pl.
gah (OAv.)
gaoiia- of a cow, pertaining to a cow 19.1/2:
acc.sg.f. gaoiim
gaomauuat- with milk 18.3: dat.abl.pl.f.
gaomauuaitibii
gauu- cow, bull; meat: acc.sg. gm, gen.abl.sg.
g u, gen.pl. gauum
gam go, come 32.1: 3.sg.aor.subj. jimat
-ca, -c (OAv.) encl. and
cixnua- wish to satisfy 25.17, 32.4:
nom.sg.m. of pres.part. cixnu
ci- f. punishment, penance 19.2: nom.sg.
cia
-cit , -ct (OAv.) affirmative or emphatic
pariticle
ci-/ca- who?, someone 22.5.1: nom.sg.m. ci,
ci-c (OAv.)
cuuat how much?
jan-/n- beat, strike 32.1: 3.sg.pres.ind. jaiti
jasa- go, come 25.16, 32.4: 2.sg.pres.inj.
jas, 3.pl.pres.subj. jasa
ti, acc.sg.m. of
pres.part. jastm
jau- m. hedgehog 19.5: gen.sg. jaao
jimat gam-
tauu tm
tauruna- young 19.1/2: acc.sg.m. taurunm
taa so, then
126 glossary
tat ha-/ta-
tan- m.f. body 19.4: dat.sg. tanuiie, loc.pl.
tanuu-c (OAv.)
tar.pia- lacking in food 19.1/2: acc.sg.m.
tar.pim
tcit ha-/ta-
tiiu- m. thief 19.5: acc.pl. tii
ti ha-/ta-
tta- shaped, formed 19.1/2: acc.sg.n. ttm
ta
ha-/ta-
tm ha-/ta-
t m ha-/ta-
t tm; ha-/ta-
ti tm; ha-/ta-
titriia- m. Titria 19.1: nom.sg. titriias-ca,
titrii, acc.sg. titrm
tii.dta- having sharp teeth 19.1/2: gen.sg.m
tii.dtahe
tuna- peaceful 19.1/2: nom.sg.f. tun (OAv.)
t tm
tm pers.pron. of the 2nd pers.sg. you 22.1:
nom.encl. t, acc.encl. , gen. tauu
(OAv.), gen.dat.encl. t, ti (OAv.)
a- possessive pron. your 22.4: gen.sg.m.
ahii (OAv.), inst.sg.m.n. (OAv.),
loc.sg.m. ahm (OAv.), loc.pl.f. h
(OAv.)
tm
rra- n. protection 19.1: acc.sg. rrm
ri.aiiara- lasting three days 19.1/2: acc.sg.n.
ri.aiiarm
ri.xapara- lasting three nights 19.1/2:
acc.sg.n. ri.xaparm
ritiia- third 19.1/2, 21.2: nom.pl.f ritiia
dauua- m. demon 19.1: nom.sg. dauuas-ca,

dauu, acc.sg. dam
dan- f. religion 19.2: voc.sg. dane
daiiit d-
daidt d-
daxa- show 32.4: 3.sg.pres.inj. daxsat
daxra- m. daxra (meaning unknown)
19.1: voc.sg. daxr (OAv.)
daxta- n. characteristic 19.1: dat.abl.pl.
daxtbii (19.2)
daa- reddish 19.1: gen.sg.m daahe
dam d-
da hu- f. country 19.5: voc.pl. da hauu
dar- hold, support 32.5: 3.sg.aor.inj. drt
(OAv.)
darga- (OAv.) long 19.1/2: gen.sg.n.
dargahii
darsa- m. appearance; sight 19.1: loc.sg.
darsi (OAv.)
dars- see, look 32.1: 1.sg.aor.inj. darsm
dasa(-) ten 21.1: indecl. dasa, gen.pl.m.f.
dasanm
daste d-
dahm.krta- made by a dahma 19.1/2:
acc.sg.m. dahm.krtm
d give; put; make 32.1/2: 1.sg.pres.inj.
dam, 3.sg.pres.mid.inj. daste, adaste,
3.sg.pres.opt. daiiit , daidt (OAv.),
2.sg.aor.inj. da
, 3.sg.aor.inj. dt (OAv.),
2.pl.aor.inj. dt (OAv.), 3.sg.aor.opt. diit ,
diit (OAv.), 1.sg.aor.mid.opt. dii (OAv.),
2.sg.aor.mid.opt. d
ditii.gtu- in the proper place 19.5:
nom.sg.m. ditii.gtu
ditii.pia- n. proper food: nom.sg.
ditii.pim
di dis-
duru- n. wood 19.5: acc.sg. duru
dtar- m. creator 18.5: voc.sg. dtar
dbaiia- deceive 32.4: 3.sg.ind.pres. dbaiieit
(OAv.)
dman- n. creature 18.4: gen.pl. dmanm
dmi- f. creation 19.5: gen.sg. dmi
drt dar-
dh- teach 32.2: 1.sg.pres.mid.ind. dda h
(OAv.)
dii d-
diit d-
dim i-/h-
dis point, show 32.5: 2.sg.aor.inj. di (OAv.)
dda h dh-
d d-
duuaah- n. (OAv.) hostility 18.7: acc.pl.
duuaa
drguuat- m. (OAv.) liar 18.3: dat.sg.

drguuit, acc.pl. drguuat
glossary 127
paoiriia- first 19.1/2, 21.2: acc.sg.n. paoirm,
acc.pl.f. paoiriia
paouruiia- (OAv.) first 19.1/2: acc.sg.m.

paouruum
paiti adv. to(wards) 23.14
paitiia
c- against, opposed to 18.2: nom.sg.m.

paiti.y
pairiiaoj- (pairi + aoj-) impose 32.1:
2.sg.pres.mid.inj. pairiiao (OAv.)
pair.jasa- surround 32.4: 3.sg.pres.inj.
pair.jasat
pauruuatt- excellence 18.2: inst.sg.
pauruuatt (OAv.)
paca five 21.1
paca.dasah- fifteen 18.7: gen.sg.m.
paca.dasah
par adv. before 23.16
pari = para + i-/aii- go foward 32.1:
3.sg.pres.ind. pariti
par adv. beside, further 23.17
parta- asked 19.1: nom.sg.n. part m (OAv.),
nom.pl.n. part (OAv.)
pasu- m. cattle, livestock 19.5: acc.sg.
pasm
pasu.hauruua- shepherding 19.1/2:
nom.sg.m. pasu.hauruu, acc.sg.m.
pasu.haurum
prsa- ask 32.4: 3.sg.pres.inj. prsat ,
2.sg.pres.impv. prs-c
prsaniia- ask 32.4: 3.sg.pres.ind.
prsaniieiti
frasa- m./n. (OAv.) consultation 19.1: acc.sg.
fras m
fras- (OAv.) f. question 19.2: dat.sg.
frasaiii
frauuas i- f. Fravai 19.5: acc.pl. frauuas
fratmn.nmna- n. foremost house 19.1:
gen.sg. fratmn.nmnahe
frazbaoah- taking away the consciousness
18.7: acc.sg.m. frazbaoahm
fr adv. forth, forward 23.19
friiaza- ( fr + yaza-) honor 32.4:
3.sg.pres.mid.subj. friiazite
frxnna- (OAv.) m./n. foresight 19.1: loc.sg.
fraxnan
frxnna- (OAv.) foresightful one 19.1:
acc.sg.n. frxnnm
frda- prosper 32.4: 3.pl.pres.mid.ind.
frdt (OAv.)
frs- attain ( fra- + s nas-) 32.1:
3.sg.aor.mid.inj.
friia- friend 19.1/2: dat.sg.m. friii
bauua- be(come) 32.4: 3.sg.pres.ind. bauuaiti,
3.sg.pres.inj. bauuat , 3.sg.pres.subj. bauut
baxta- n. destiny; disgrace 19.1: acc.sg. baxtm
bara bear, carry 32.4: 3.sg.pres.impv. baratu,
3.du.pres.ind. barat, 3.pl.pres.opt.
baraiin
bzu- m. arm, leg 19.5: acc.pl. bzu
brzat- high 18.3: gen.sg.m. brzat
bit nothing less than
bitiia- second 19.1/2, 21.1: acc.pl.f. bitiia
bi twice
bmi- f. earth 19.5: acc.sg. bmm
bti- f. adornment 19.5: acc.pl. bt
naci-/naca- no one 22.5.1: nom.sg.m.
naci, acc.sg.m. nacim
nar- m. man 18.5: nom.sg. n, acc.sg. narm,
gen.sg. nar, gen.pl. narm
nas- reach
nir- f. woman 19.3: gen.pl. nirinm
nuuaiia- rich in waters 19.1/2: gen.pl.
nuuaiianm
nman- n. name 18.4: acc.pl. nm n (OAv.)
na
vam
nmah- n. reverance, homage 18.7: gen.sg.
nmah
nit not
niiza- (ni + aza-) close, tighten 32.4:
3.pl.pres.opt. niizaiin
ni-darza- tie, bind; hold down 32.4:
3.pl.pres.opt nidarzaiin
ni-mr- utter 32.1: 3.sg.pres.mid.ind.
nimrite
ni-mrauua- utter 32.4: 3.sg.pres.mid.ind.
nimrauuaite
ni-duuara- run/chase (out): 3.sg.pres.ind.
niduuaraiti
nrm now
nmna- n. house 19.1: loc.sg. nmne
128 glossary
nmn.paiti- m. master of the house 19.5:
acc.sg. nmn.paitm
maiti- f. mind 19.5: nom.sg. maiti
maibii azm
mam.nmna- n. house of
middling importance 19.1: gen.sg.
mam.nmnahe
mat prep./posp. + gen. with
man- think, realize: 1.sg.aor.mid.inj. m gh
manaor- f. neck 19.3: acc.sg. manaorm
manah- n. thought 18.7: inst.sg. manah
(OAv.)
maiia- be aware, think about 32.4:
1.sg.pres.mid.subj. mainiii
maiiu- spirit 19.5: acc.sg. mainiim, gen.sg.
mainii u, inst.sg. mainii (OAv.)
mara- memorize 32.4: 3.pl.pres.ind. mart
(OAv.)
mazan- n. greatness 18.4: inst.sg. maz n-c
(OAv.)
mazd- m. Mazd 18.1: nom.sg. mazda
,
voc.sg. mazda, mazd (OAv.), acc.sg.
mazdm
mazdaiiasna- Mazdayasnian 19.1/2: nom.pl.
mazdaiiasna
mazd.vara- which Mazd chooses 19.1/2:
acc.pl.n. mazd.var (OAv.)
mas iia- m. mortal, man 19.1: loc.pl. mas iias
mas iika- m. man 19.1: nom.pl. mas iika
m azm
mzdaiiesni- Mazdayasnian 19.5: voc.sg.f.
mzdaiiesne
mra- m. mantra 19.1.: acc.pl. (sic!) mra
mm azm
m gh man-
mdauuat- having prizes 18.3: acc.pl.n.
mdauun (OAv.)
mrauu-/mr- speak 32.1: 2.sg.pres.inj. mrao,
3.sg.pres.inj. mraot , 2.pl.pres.inj. mraot
(OAv.)
ya- relative pron. which, who 22.3: nom.sg.m
y (OAv.), yas-t, y, acc.sg.m. y m (OAv.),
yim, gen.sg.n. yehii, inst.sg.m. y, loc.sg.m.
yahm (OAv.), nom.pl.m. yi, nom.acc.pl.n.
y(-c) (OAv.), nom.sg.f. y
yaon.xvapta- sleeping in its den 19.1/2:
gen.sg.m. yaon.xvaptahe
yao iiu-
yaodra- n. purification 19.1: acc.sg.
yaodrm
yauua- wheat 19.1: gen.pl. yauuanm
yauuat as much as
yaa as, when 37.3
yana- m. strength, force 19.1: inst.sg.
yan
yasna- m. sacrifice, act of worship 19.1: inst.sg.
yasna, gen.pl. yasnanm
yaza- worship, honor 32.4: 1.sg.pres.mid.ind.
yaze, 1.pl.pres.mid.ind. yazamaid
(OAv.), 3.pl.pres.ind.mid. yazte,
3.pl.pres.mid.opt. yazaiiata
yaziia- be worshiped 32.4: 3.pl.pres.ind.
yaziti
yh- n. girdle 18.1: nom.sg. ya
yeii if 37.3
yesniia- deserving of worship 19.1/2:
nom.sg.m. yesniias-ca
yezi if 37.3
yiiia- fight 32.4: 3.du.pres.ind. yiiia
yjiiasti- f. yjiiasti- (a measure of distance):
acc.sg. yjiiastm
ym pers.pron. of the 2nd pers.pl. you, you
all 22.1: acc.encl. va
, acc.gen.dat.encl. v,
inst. xm (OAv.)
vada- find 32.4: nom.sg.m. of pres.mid.part.
vadmn (OAv.)
vam pers.pron of the 1st pers.pl. we 22.1: acc.
hm (OAv.), acc.encl. na
(OAv.)
vairiia- to be chosen, desirable 19.1/2:
gen.sg.f. vairiia
vac speak, say 32.1: 3.sg.aor.pass.inj. vc

(OAv.), inf. us
iii (OAv.)
vahu- good 19.3,5: acc.sg.f. vavm, dat.sg.n.
vahauu (OAv.), inst.sg.n. voh (OAv.),
acc.pl.f. vavh, acc.pl.n. voh (OAv.)
var choose, want 32.3: 3.sg.pres.mid.ind.
vrt (OAv.)
varduua- soft 19.1: gen.sg.n. varduuahe
vas-/us- wish, want 32.1: 2.sg.pres.ind. va
(OAv.)
glossary 129
vasas.xara- m. (OAv.) who rules
according to his own will 19.1: gen.sg.
vasas.xarahii
vaza- lead, drive 32.4: nom.sg.m. of
pres.mid.part. vazmn
vahita- best 19.1/2: acc.sg.n. vahitm
vahmiia- deserving of praise 19.1/2:
nom.sg.m. vahmiias-ca
v or 37.2
vc vac-
vstra- n. pasture, meadow 19.1: gen.pl.
vstranm
va
ym
v- f. herd: acc.sg. vm
vrt var-
vrziia- carry out, execute, do 32.4:
3.sg.pres.subj. vrziit , 3.pl.pres.subj.
vrziin, mid.inf. vrziieidiii (OAv.)
vhrka- m. wolf 19.1: acc.sg. vhrkm
vourukas a- Vourukas a (lit.: having a wide
bay) 19.1/2: acc.sg.n. vourukas m, abl.sg.
vourukas t
voh vahu-
vohunazga- for hunting (of dogs) 19.1/2:
nom.sg.m. vohunazg, acc.sg.m.
vohunazgm
v ym
vizdiii vid-
viia- pursue, chase 32.4: 3.sg.pres.ind.
viieiti
viixmaiia- proclaim, speak aloud 32.4:
3.sg.pres.mid.ind. viixmaniiete
vid- find; know 32.6: act.inf. v-uuduii,
mid.inf. vizdiii
vi.hauruua- guardian (lit: protecting the
village) 19.1/2: nom.sg.m. vi.hauruu,
acc.sg.m. vi.haurum
v adv. apart
vuuduii vid-
vkrt .utna- which cuts away life 19.1/2:
acc.sg.m. vkrt .utanm
vra- m. man 19.1.: acc.sg. vrm
vriia- pertaining to men, manly 19.1/2:
acc.sg.f. vriim
vs- f. tribe, village 18.1: abl.pl. vibii
vspa- all 19.1/2: acc.pl.m. vsp g (OAv.),
acc.pl.n. vsp-c (OAv.), gen.pl.m.
vspanm, inst.pl.m. vspi
vzu- m. vzu (an animal) 19.5: gen.sg. vzao
vibii vs-
rauuat- brilliant, splendid 18.3: nom.sg.m.
rauua
, acc.sg.m. rauuatm
raaiia- mix 32.4: 3.sg.pres.ind. raaiieiti
raaiia- damage 32.4: 3.sg.pres.subj. raaiit
raoxnu- n.? light: loc.pl. raoxnuuua
raocah- n. light 18.7: acc.pl. raoca
s-c (OAv.)
raopi- m. fox 19.5: gen.sg. raopi
ratu- m. judgment 18,5: acc.pl. rat
rapiina- pertaining to midday 19.1/2:
acc.sg.m. rapiinm
rafnah- n. support 18.7: nom.acc.sg. rafn
rt- f. gift 19.2: acc.sg. rtm
rna- m. side, part 19.1: dat.pl. rnibii
(OAv.)
sauua- n. abundance: loc.sg. sauui (OAv.)
sata- one hundred 21.1: acc.pl.n. sata
sad- seem 32.5: 3.sg.aor.inj. ss (OAv.)
sardan- f. challenge; challenger 19.2: acc.pl.
sardana
sdra- n. defeat 19.1: acc.sg. sdrm

sdra- deceptive 19.1/2: nom.sg.f. sdr (OAv.)
sma- black 19.1/2: gen.sg.m. smahe
ss sad-
sraoa- m. (OAv.) obedience: nom.sg.
srao
s gha- (OAv.) explain: 3.sg.pres.ind.
s ghait
s gha- (OAv.) m. explanation: gen.sg.
s ghahii
sukuruna- m. porcupine 19.1: gen.sg.
sukurunahe
snahe span/sn-
stauu-/stu- praise 32.1: 1.sg.pres.ind. staom
(OAv.)
staman- m. mouth 18.4.: stamm
sti- f. existence 19.5: gen.sg. sti
spacira- canine 19.1/2: gen.pl.m.
spaciranm
span-/sn- m. dog 18.1/4: nom.sg. sp, acc.sg.
spnm, gen.sg. snahe, nom.pl. spn
130 glossary
spasan- spying 18.4: nom.pl.m. spasn
spta- beneficient 19.1/2: acc.sg.m. sptm,
gen.sg.m. sptahe, inst.sg.m. spt (OAv.)
spt.tma- most beneficient 19.1/2:
acc.pl.n. spt.tm (OAv.)
spt.mainiiauua- pertaining to the
beneficient spirit 19.1/2: gen.pl.m.
spt.mainiiauuanm
sp nita- (OAv.) most beneficient 19.1/2:
nom.sg.m. sp nit
spitama- m. Spitama 19.1: voc.sg. spitama
spiti.dira- having white eyes 19.1/2:
gen.sg.m. spiti.dirahe
snaa- m. blow, hit 19.1: acc.sg. snam
sraoa- pay attention: nom.sg.m. of pres.part.
sraomn
srao.caran- scourge 19.2: inst.sg.
srao.caranaiia
srra- beautiful 19.1/2: gen.sg.m. srrahe
zaor- f. libation 19.2: dat.abl.pl. zaorbii
zaozao-/zaoz- call 32.1 (present intensive
25.2b): 1.sg.pres.ind. zaozaom (OAv.)
zairi.gaoa- having yellow ears 19.1/2:
gen.sg.m. ziri.gaoahe
zauruuan- m. old age 18.4: acc.sg.
zauruunm
zarautra- m. Zarathustra 19.1: nom.sg.
zarautr, voc.sg. zarautra
zaranii.aiina- having a golden bridle
19.1/2: gen.sg.m. zaranii.aiinahe
zaranii.sruua- having golden horns 19.1/2:
gen.sg.m. zaranii.sruuahe
zarazditi- f. faith 19.5: nom.sg. zarazditi
za- n. creation, generation 19.1: loc.sg. zi
(OAv.)
z because
zraiiah- n. lake 18.7: acc.sg. zraii, abl.sg.
zraiiahat
zruuan- m. time 18.4: acc.sg. zruunm
s
iia-/s
ii- dwell 32.1: 3.pl.pres.ind. s
iieit
(OAv.)
s
iiaoana- n. deed, act 19.1: acc.pl. s
iiaoan,
inst.pl. s
iiaoani
s
iiaona- n. deed, act 19.1: acc.sg. s
iiaonm,
gen.pl. s
iiaonanm
s
iiaonuuarz- doing deeds 18.1: acc.sg.m.

s
iiaonuuarzm
ha-/ta- dem.pron. this 22.2: nom.sg.m. h,
nom.sg.n. tat , acc.sg.m. t m (OAv.), acc.sg.f.
tm, nom.du.m. t-cit (ta
), nom.pl.m. t, ti
(OAv.), inst.pl.m. ti
haomauuat- with haoma 18.3: dat.abl.pl.f.
haomauuaitibii
haiiia- truthful 19.1/2: nom.sg.m. haiii
hauua- own 22.4: gen.sg.m. hauuahe-ca
hauuapaha- (OAv.) skillfulness 19.1: inst.pl.
hauuapahi
haca
(OAv.): posp. + gen.abl. (outside) of,

from; according to; together with 23.24
haca- follow 32.4: nom.sg.m. of pres.part.
hacimn
hapta seven 21.1
hu auua-
hr.masah- a hra in length 18.7:
acc.sg.m. hr.masahm
hm adv. together 23.27
h i-/h-
h ha-/ta-
hi i-/h-
hiiat (OAv.) that, because, since, when 37.3
hita- stand: 3.pl.pres.mid.ind. hitte
hm i-/h-
huuar- n. sun 18.6: gen.sg. xv g (OAv.)
huu auua-
hunairiia
k- dextrous 18.2: gen.sg.m.

hunairiia
c
hunara- m. skill 19.1: gen.pl. hunaranm,
inst.sg. hunar
Word Index
Avestan (131), Old Church Slavic (149), Old English (149), Gothic (149), Greek (149), Hittite (150),
Old Irish (150), Latin (150), Lithuanian (151), Oscan (151), Pli (151), Old Persian (151), Prkrit (151),
Proto-Indo-European (151), Proto-Indo-Iranian (153), Sanskrit (153), Umbrian (158). Words are
indicated by section numbers, footnotes are preceded by the number of the chapter in which
they occur. For instance, fn. 2.1 = footnote 1 in chapter 2, etc.
Avestan
a/an/ana- 39.3
a-/i- 22.1.6
a-/i-/ima- 22.2.1, 22.5.2
aibii 8.1, 22.1.1, 22.2.1, 36.3
aibi 22.2.1
auua- 20.2, 21, 21.1, 22.2.1
auua ha
20.2
auuadasa- 21.2
auuahmi 20.2
ata- 7.5.2, 22.2.1, 22.2.3, 22.3
ata 22.2.3
ataibii 22.2.3
ataca 22.2.3
atam 22.2.3
atau 22.2.3
atauua 22.2.3
ataiia 22.2.3
ataiia
22.2.3
ataiia
scit 22.2.3
atat 22.2.3
atahm 22.2.3
ata ha
7.5.2, 22.2.3
atahmi 22.2.3
atahmt 22.2.3
atahmi 22.2.3
atahmiia 22.2.3
atcit 22.2.3
atbii 22.2.3
ata
22.2.3
ata
s.t 22.2.3
at 22.2.3
atm 22.2.3
at 22.2.3
ate 7.11.3, 22.2.3, 37.3
arapaiti 19.5.9
am 10.1.2, 22.2.1
asma- 10.1.1
asm 7.9.3
aa- 10.1.1
aa 22.2.3
a 22.2.3
am 22.2.1
ama- 10.1.1
a 22.2.3
aiin 32.5
au 22.2.1
auua 22.2.1
aoi 10.3.2, 11.1.4, 23
aoui 11.1.4
aoxta 11.11.1, 32.1
aogd 11.11.1, 32.1
aog 11.11.1, 11.20.5
aog 11.12.2, 36.3
aoj 11.11.1, 26.1, 32.1
aojaite 32.1
aojah- 10.3.1, 11.12.2
aojah 11.12.2
aoji 32.1
aoji 32.1
aojiiah- 20.1.1
aojita- 20.1.1
aoj 32.1
aojta 32.1
aom 22.2.5
aoah- 10.3.1
aii-/i-/y- 32.1
aiia 22.2.1
aiiar 18.6
aiir 18.6
aiia
22.1.1
aii 18.6
aiin 18.6
aii m 10.1.2, 22.2.1
aiien 36.1
aiti 23
aidiiu- 7.17.6, 19.5
aidii 7.17.6
aipi
23
aib 10.3.2, 11.1.4, 23
aii 10.3.2, 11.1.4, 23
aiiiasca 22.2.1
aiii 18.1, 18.1.18
aiiaiiama 32.1
aini
m 7.1.4: fn. 2.6

ainm 7.1.4: fn. 2.6
airiiaman- 8.1
auua- 22.2.5
auua 22.1: fn. 3.7, 22.2.5
auuam 22.2.5
auuauuat 11.1.4
auuaxvabda- 39.3
auuaa 38.3
auuat 22.2.5, 38.3
auuabii 22.2.5
auuahi 22.2.5
auuaha
22.2.5
auuah 19.5.37
auua ht 22.2.5
auua ha
22.2.5
auua he 22.2.5
auuah- 18.7
auuaheca 22.2.5
auu 22.2.5
auuitm 32.1
auuin 32.1
auui 22.2.5
auuc 32.1
132 word index
auua
22.2.5
auum 22.2.5
auue 22.2.5
auruuanta 35
auruuantm 7.8.3
aka- 19.1
axtiii 19.5, 19.5.6
axti 7.15.2, 19.5
axtm 7.15.2
ax iii 11.29, 22.2.1
ax iica 22.2.1
aruu 19.4.8
ar- 19.4
aa
38.2, 38.3
ad 32.6
ads 18.1.5
adrujiiat- 11.25.6
aaoiiamna- 32.4
aairi 23
at 38.2
at ca he 22.1.6
at ct 38.2
apa 18.1, 18.1.8, 23
apacit 18.1.7
apam 10.1: fn. 2.7
apa.tacin 32.1
apca 18.2
apc- 18.2
apm 18.1, 18.1.17
ap 18.2
aprsaiiatm 32.4
ape 18.1, 18.1.7
ap 18.1, 18.1.5, 18.1.15
apura- 20
apur- 20
afnavhat- 11.18
afratat kuuah- 39.3
afmn 18.4
abaom 10.3.3, 32.4
abauuat 11.1.4
ara- 11.26.3
aha 11.27
ahaiti 32.1
ahauua 19.5.26
ahat 11.27, 32.1
ahn 32.1, 37.3, 38.1
ah u 11.29: fn. 2.14, 19.5.22
ah 32.1
a hat 25.2
a hi 11.29, 22.2.1
a ht 22.2.1
a ha
22.2.1
a ha
sca 22.2.1
a he 22.2.1
a himnaiia
25.2
ahu- 19.5
ahuu 19.5, 19.5.31
ahu 11.29: fn. 2.14
avhe 19.5.24
aiia- 11.1.3
aiiam 20.2
aiiahmi 20.2
aiia
s-cit 19.1.16
aiie 20.2
ana- 22.2.1, 22.2.4, 22.5.2
ana 22.2.1, 22.2.4, 23
anaocah- 18.7
anauuavhabdmna- 39.3
anafmm 18.4
anas a- 39.3
anahe 22.2.4
anahmt 22.2.4
an 22.2.1, 22.2.4
ani 22.2.4
aniia- 7.9.1, 7.14: fn. 2.6, 19.1.3
anii m 7.14: fn. 2.6, 19.1.3
anii.t kaa- 11.10.2
anu 23.7
aguta- 11.12
atar 11.3.2, 23.8
amrtatt- 18.2
ams a- 7.9.3, 11.2.3, 19.1, 39.3
ams aibii 19.1, 19.1.19
ams 7.9.3
ams g 7.9.3
ams s-ca 19.1.16
am hmaid 32.5
ar 32.1, 32.3
arduu- 11.13
ardu- 18.8
ara- 11.13, 21.4
arti- 11.20.5
a
s 11.10.1
asaiia- 7.3.2
as-ca 18.1, 18.1.4
asct 22.1, 22.1.1
ast- 9.2.5, 18.1
astaca 18.1, 18.1.11
astm 18.1.17
astm 18.1.4
ast 18.1.5
ast 25.1.1
asti
11.20.1, 18.1, 18.1.16, 32.1

astuuat- 11.17.4
astuua
18.3
ast 32.1
aspa- 11.10.3, 11.20.3
aspaibiia 19.1.14
aspaca 19.1.9
aspm 7.8.1
asna- 11.13, 11.24.2
asman- 11.20.3
asmanm 18.4, 39.4
asruutm 32.1
asrdm 32.1
az
m 7.9.2, 10.1.2, 22.1, 37.1,

37.3, 38.3
az m 22.1
azd 11.24.3
azdb 9.2.4
azdib 9.2.4, 18.1.19
az 19.3
a.aojastara- 20.1.2
as.aojah- 20.1.2
aibii 35
ata 11.21, 21
atahuua- 21.4
atiti 21, 21.1
atma- 21, 21.1
as a- 19.1
as aon 18.4
as aoniii 19.3
as aoniia
19.3
as aonibii 19.3
as aoninm 19.3
word index 133
as aon- 19.3
as aonm 19.3
as aon 19.3
as aoni
19.3
as aii 19.5
as auuaoii 11.1.3
as auuaobii 11.1.3
as auuan- 10.6.2, 11.1.3, 18.4, 19.3
as auuan 18.4
as x ii-c 19.1.5
as a-cinah- 5.4: fn. 2.1
as ahii 19.1.5
as 19.1.2, 36.1
as at c 19.16
as un 18.4
as um 10.6.2, 11.3.2, 18.4
as t 19.1, 19.1.6
as t ct 19.1.6
as .yec 7.10: fn. 2.5
as m 19.1
ai- 19.5
ai 19.5, 19.5.10
as iuua
19.5
aibiia 19.5, 19.5.12
as i 11.20.5, 19.5
as 19.5
as m 19.5
as i 19.5
ai- 11.25.6
ah 7.8, 11.10.1, 11.20.1, 11.24.2,
11.26.1, 11.26.2, 11.27, 11.28,
26.1, 32.1, 32.6, 37.2
2ah 11.28
ahe 22.2.1
ahi 11.26.1, 32.1
ahii 11.28, 22.2.1
ahu- 11.1.4, 11.26.1, 19.5.31
ahuii 11.1.4, 19.5.24
ahuua
19.5.29
ahubiia
19.5
ahubii 19.5.30
ahura- 7.8.1, 7.16, 19.1
ahura
19.1
ahuraibiia 19.1
ahurahe 19.1
ahurahii 19.1
ahur 19.1.2
ahuri 19.1
ahuri. 19.1.7
ahura
h 19.1, 19.1.15
ahurm 7.8.1, 19.1, 37.2
ahur 19.1, 37.3
ah 19.5.25
ahm 11.29: fn. 2.14, 19.5.20
ahma 7.9.2, 11.26.2, 22.1
ahmaibii 22.1, 22.1.3
ahmat 7.3, 7.9.2, 22.1
ahmi 22.2.1
ahmka- 22.4
ahmki 22.4
ahmkm 22.1, 22.4
ahmk g 22.4
ahmt 22.2.1
ahmt cit 19.1.6
ahmi 22.2.1, 32.1, 36.3, 39.1
ahmiia 22.2.1
ahm 22.2.1
23.9
+hac 11.20.1
at 37.3, 38.2, 38.3
at yat 38.3
iiapta- 11.10.3
iit 11.1.3
iiese 32.4
iiu- 19.5, 19.5.38, 19.5.39
ii 32.1
i 22.2.1
uuaocm 32.4
uui 7.15.1
uuiia- 7.15.1
kriti- 8.1, 9.1
xtirm 21.2
tar- 11.2, 18.5
tar 18.5
tar ahurahe mazda
36.2
trm 11.2.1, 18.5
d- 32.6
-dadat 32.2
d- 18.1
-d- 25.1.2
- 32.1
pa 18.1, 18.1.10
pat 18.1.6
pm 18.1
p 18.1, 18.1.14, 18.1.15, 37.1,
39.2
f 18.1.1
brt- 18.1
brta 18.1.10
brtm 11.14.3, 18.1.2
brte 18.1.7
br 18.1.1
biia 22.2.1
biiasc 22.2.1
bii 22.2.1
b 22.2.1
rmaiti- 7.4.2
s 32.1
sit 36.2
skiti- 11.20.1
h- 18.1, 32.1, 32.6
huua 22.2.1
h 22.2.1
a
ha 18.1.8, 32.6
a
har 11.27, 32.6

a
hire 32.1
a
hm 32.1
a
hm 2.2.1
a
h 18.1.5
a
ht 32.6
a
ht .tm 32.6
a
sc 22.2.1
xnah- 7.6.1
sa- 19.1.12
saiia
19.1.12
zah- 7.6.1
uuisti- 7.15.1
uuduua
7.8.2, 39.2
rduua- 11.13
ra- 11.13
rnauu-/rnu- 32.3
rnuui 32.1
rnu- 32.1
rzao 19.5.22
rzu- 9.1, 11.2.1
134 word index
rzu 19.5
rzuu 19.5.19, 19.5.22
r 5.4
r.vaca
5.4
r-jiii 18.1.7
r-j 18.1.1
uu 22.1
n 23.7
a
h 18.1.8
hm 7.9.2, 11.26.2, 22.1
ii 22.2.1
iim 21.1
im 21.1, 37.4
m 21.1
i 26.1, 32.6, 37.2
iiei- 32.6
iieiin 32.6
iiiejah- 7.10.1, 8.1
id 32.1
it 22.1.6
ima- 10.2, 22.1.6, 22.2.1
imat 22.2.1
im 22.2.1
ima
22.2.1, 37.1
im 7.9.3, 22.2.1
imm 22.2.1
imm 22.2.1, 37.3
ime 22.2.1
irixta- 8.1
iric 32.5
iri 32.5
irri 32.6
irrar 32.6
irrina 32.6
irriu- 32.6
is- 32.1
is 36.2
ismaid 32.1
is 32.1
isii 32.4
iziia- 26.12
i 10.5.1
ia- 26.7
ite 32.1
22.1.6
t 22.1.6
m 22.1.6, 22.2.1
- 18.1
22.1.6
m 18.1.2
18.1.14, 18.1.15
tm suuita 36.1
uiie 10.2.3, 11.1.4, 19.1.11, 21.1
uua- 21.1
uua 21.1
uuaibiia 21.1
uuaiia
21.1
uuaii 19.1.13, 21.1
uxa- 11.17.2, 19.1
ux 19.1
uxibii 19.1.19
uxiia- 25.1
ugra- 11.12
ura- 11.12, 20.1.1
upa
23.10, 36.2
upairi
23.11
upa.arti 36.2
upaskabm 11.3.2, 11.14
upmaca 19.1.8
uba- 10.2.3, 21.1
ub 10.2.3, 11.1.4, 19.2, 21.1
ubii 19.1.13, 21.1
ubibii 19.1.14, 21.2
ubda 11.11.1
ubdana- 11.11.1
uruua 18.4
uruuaire 19.2, 19.2.3
uruuax- 32.5
uruuaj 32.5
uruuata- 8.2
uruuan- 7.6.3
uruuaraiit 19.2.5
uruuaraiia
19.2
uruuaranm 36.2
uruuar- 19.2
uruuarbiiasca 19.2.13
uruuara
sca
19.2.11
uruu 18.4
uruuxat 32.5
uruudah- 8.2
uruun 7.6.3
uruuz 32.6
uruun 7.6.3
uruunait- 32.3
urud 27.3, 32.6
urdiia- 26.14
urpaiia- 26.14
urraod- 32.6
urraost 11.17.4, 32.6
us 23.12
usuuah 32.1
ustna- 11.20.2
ustna-zasta- 5.4: fn. 2.1
us-d- 25.1.2
uz- 23.12
uzaiiat 36.2
uz-uuaat 11.25.5
uzuxiinca 11.3.2
uzuxiieiti 38.3
uah- 18.7
ua
18.7
ua
hm 18.7
uibiia 19.5.12
utra- 11.17.4
utra
h 36.2
ka- 7.13.1, 11.20.1, 11.26.1, 22.2.1,
22.5
kaoiim 19.5.16
kaine 18.4
kainii 7.8.3
kainibii 7.8.3
kainn- 7.8.3, 18.4
kainn 8.1
k(a)uua 38.1
kauuam 19.5.2
kauu 7.8.2
kaa
38.1
kad 38.1
kat 11.9, 38.1
ka he 22.5.2
kana 22.5.2
kamnnar- 5.4: fn. 2.1
kar 7.13.3, 11.2.1, 11.3.2, 11.9,
11.10.2, 26.5, 32.3, 33.2, 33.3
kar- 18.1
word index 135
kar 18.6
karuuar 18.6
karuun 18.6
kasiiah- 20.1.1
kasita- 20.1.1
kasu- 20.1.1
kas-ci
t 19.1.1
kas-n 11.20
kahe 22.5.2
kahii 11.26.1, 22.5.2
k 22.5.2
kuuaiias-ca 7.4.2
kma- 11.9
kraiin 32.4, 39.2
km 36.1
kuuna- 7.8.2
krta- 11.2.1, 33.3
krti- 33.2
krnaoiti 26.5, 32.3
krnaot 32.3
krnaon 32.3
krnaomi 32.3
krnauu-/krnu- 26.4, 32.3
krnauune 32.3, 38.3
krnauuni 32.3
krnuiit 32.3
krnuuaiti 32.3
krnuuat- 32.3
krnii 32.3
krta- 26.9
krf 18.1.1
kruu 32.1
khrp- 11.2.2, 18.1
khrpa 18.1.8
khrpas-ca 18.1.15
khrpm 18.1.17
khrpm 18.1.2
khrp m 18.1.2
khrpiia 18.1.9
k 7.13.1, 11.20.1
k 7.13.1
kuua 11.1.3
kuxnu- 32.2
kuxnuuna- 32.2
kua 38.1
kura
38.1
kudat 38.1
kud 38.1
k 38.1
x- 18.1
xa
18.1.1, 18.1.4
xm 18.1.17
xraosa- 32.4
xraostm 32.4
xrataot 19.5, 19.5.23
xratauu 19.5
xratu 10.6.1, 19.5.26
xrata
19.5.26
xratu- 7.17.6, 10.6.1, 11.16, 19.5
xratuma
18.3
xrat 19.5, 19.5.25
xratm 19.5
xrat 7.17.6, 19.5, 19.5.33
xra 19.5.25
xrae 19.5.24
xra 19.5.22
xaiia- 32.4
xaiia 32.4
xaiiamn 7.8.3
xaii 32.4
xaiieite 7.10
xaiiehe 32.4
xaiieh 7.10.1
xara- 19.1
xara 19.1
xart 11.10.2
xari 19.1.9
xari. 19.1.9
xariu 19.3
x 7.8.3, 7.10.1, 36.2
x- 18.1
xuua 11.16, 21, 21.1, 21.3
xuuati- 21
xudra- 11.13
xtuua- 11.17.4, 21, 21.2
xnao- 32.5
xnaoi 32.5
xnaon 32.5
xnaota 32.5
xn 11.16
xnu 10.6.1
xnu 32.2, 32.5
xma- 22.1.4
xmaibii 22.1, 22.1.4
xmat 22.1, 36.2
xm 22.1
xmuuiia
xmka- 22.4
xmkahii 22.4
xmk 22.4
xmki 22.4
xmkm 22.4
xmkm 22.1, 22.4
x iit 32.1
x iit 11.28
x iim 32.1
x ii m 11.28, 32.1
x iia
11.28
xva- 11.28, 11.30, 22.4
xvana- 11.13
xvafna- 9.1, 11.30
xvafna- 11.30
xvaha 18.5
xvahar- 18.5
xvan- 7.17.7, 9.1, 11.30
xvanuuat- 7.4
xvar 33.2
xvarti- 33.2
xvahe 22.1.7
xvahmi 22.5
xvi 22.1.7
xvra- 11.30
xvr 19.1.9
xvrii 19.1.9
xvpaiiia- 8.1
xvstra- 11.30
xv nuuat- 7.9.1
xv nuut 7.4.3
xv g 11.30, 18.6
xvta- 11.30
gaanm 19.2
ga- 19.2
gauuii 19.2, 19.2.13
gabii 19.2
gab 19.2
136 word index
gahuua 19.2, 19.2.15
gah 19.2, 19.2.15
gaa
19.2, 37.5
gam 10.1.2, 19.1.3
gaoaibiia 19.1.14
gaoae 19.1.14
gaiia- 10.1.2, 19.1.3
gaiias-c 19.1.1
gair 18.1.7
gairi- 7.15.3, 19.5
gairibii 19.5
gairinm 19.5, 19.5.16
gair 7.15.3, 19.5
gauu- 11.12, 19.5, 19.5.18
gauustriia- 11.1.3, 11.17.4
gata- 33.3
gam 11.3.2, 11.12.2, 11.20.4, 26.16,
27.1, 28, 30.1, 32.6, 33.3
gar 32.6
gar- 9.2.4, 18.1
gara 19.5
garma- 11.12
gar 18.1.5, 18.1.15
garit 19.5
garib 18.1.19
garb 9.2.4
garz 11.25.3
giia- 10.5.2
gim 10.5.2
gu 10.6.1, 11.2, 19.5.18
gtu- 7.17.5, 19.5
gtuu 7.13.4, 19.5.26
gtm 7.17.5
guruuaiia- 11.1.4
-gt 11.10.2
gn 38.2
gmn 32.1
grfa- 26.17
grn- 26.5, 32.3
grniti 11.1.4, 32.3
grnn 32.3
grd 11.25.3
g u 10.4, 37.3
guzaiia- 26.14
gu 11.31
ga- 32.4
gahuu 11.131, 32.4
gdm 32.4
graom 19.1.3
grauua- 19.1.3
grab 11.1.4, 32.6
gr hma- 11.26.2
gruuaiia 19.2
gruu-
nnm(ca) 19.2.12
nat 18.1.6
nm 18.1.17
ne 18.1.7
n 18.1.5
ar- 11.25.5: fn. 2.13
ca
11.3.1, 11.9, 19.1.6

caiiasc 22.5.1
caii 22.5.1
caii-/s
ii- 32.1
caxra- 19.1
catar- 21, 21.1
catar 21.1
catura 21.1
caturm 21.1
catur 21.1
caar- 21
caar-/catur- 21.1
caarasca 7.3.1, 21.1
caarsat- 21
car 7.3, 11.17.1, 21.1, 38.3
carudasa- 21.2
caru 21.3
caruuua- 21.4
car-/kr- 32.1
cara- 32.4
carkr- 26.2.b
caman- 7.6.2, 11.3.3, 18.4
camanat 18.4
camm 7.6.2, 11.3.3
cam g 18.4
cahii 22.5.1
cahmi 22.5.1
cuu 32.1
ci-/ci- 32.1
ciait 32.1
ciat 10.1: fn. 2.7
ci 32.1
ci-/ci- 32.1
cim 10.1: fn. 2.7
cit 11.17.4
crt 7.13.3, 11.10.2
ci 32.2
ci-/ca- 22.5, 22.5.1
cikaii- 32.2
cikaiiat 32.2
cikit-/cicit- 32.6
cikitr 30.1.d, 32.6
ci-ca 22.5.1
ciciu- 19.3
ciciah- 19.3
cit 32.6, 10.1: fn. 2.7
ci 32.3
cira- 7.9.4
cir 7.9.4
cir s-ca 19.1.16
cit 22.5.1
cit 7.2,
ci
t 19.1.6
cina- 32.3
cinamaide 32.3
cinas 32.3
cinasti 32.3
cinah-/c- 32.3
cinahm 32.3
cinmni 18.4
ci- 32.3
ci 7.15.3, 10.1: fn. 2.7, 11.17.4,
22.5.1, 32.1
ci-ca
22.5.1
c-c 22.5.1
cm 22.5.1
ciit 32.3
cii 32.3
cmaide 7.15.3, 32.3
cmah 7.15.3, 32.3
cd 32.3
ci
m 22.5.1
cuuat 18.3, 36.1
cuuat drj 36.1
cuus 18.3
word index 137
ja 18.1.1
jaiiia- 11.25.6, 26.10, 32.4
jaiiiat- 18.3
jaiiieiti 32.4
jaiiiemi 32.4
jaiiiehi 32.4
jaini 32.1
jaiti 11.12.2, 32.1, 36.3
jaauruu- 32.6
jar- 32.6
jara 32.6
jana- 26.8
jam- 28, 32.6
jamiim 32.6
jam- 30.1.e, 32.6
jan 11.12.2, 32.1
jan 18.1
jan-/ja-/n- 32.1
janm 18.1.2
jan 18.1.14, 18.1.15
janiii 19.5.4
jat 11.3.2
jam 11.12.2
jam-/gm-/ga- 27.1, 32.1
jamat 32.1
jamiit 32.1
jamiim 32.1
jamiir 32.1
jamiin 32.1
jamiia
32.1
jasa- 11.12.2, 26.16, 32.4
jasa 32.4
jasatm 32.4
jasama 32.4
jasaiti 11.20.4, 38.3
jasatm 32.4
jasat- 18.3, 32.4
jasi 32.4
jasiti 32.4
jast 32.4
jast 32.4
jasn 32.4
jas 30.1, 32.4
jasit 32.4
jmaspa- 9.2.1
ja
18.1.1
j n 32.1
j ghati-c 8.1, 11.3.2
jiia 11.1.3, 11.12.2
jiga- 32.6
jiaa 32.6
jimn 32.1
j 32.6
j- 18.1
juua- 32.4
juuuua 32.4
juumah 32.4
ta- 7.11.1, 8.1, 11.28, 22.16, 22.1.2,
22.1.6, 22.2.1, 22.2.2, 22.5.3,
22.3, 22.5.2
tact 22.2.3
taibii 22.1
taibii 22.1
tauua 22.1
tauu 22.1
tauruuaiia- 32.4
tauruuaiiata 32.4
taxe 32.1
tac- 26.1, 32.1
taca- 32.4
taciti 32.4
tat 22.2.3, 36.1
tafsa- 26.16
tan- 32.3
tanauu-/tanu- 32.3
tanauua 32.3
tanuiia 32.3
tanuii 10.2.3, 19.4.5
tanuua 19.4.6
tanuuat 19.4.4
tanuu m 7.17.5, 19.4.2
tanuu 19.4.3
tanuu 19.4.8
tanuui 19.4.7
tanubii 19.4.10
tanu 11.20.5, 19.4.1
tanuu 19.4.12
tanu-c 19.4.12
tan- 7.17.5, 10.2.3, 19.4
tanm 7.17.5, 19.4.2
tarasca 23.13
tar 23.13
tar 7.9.4
tar 7.9.4, 23.13
ta 11.17.4, 33.3
taa- 27.2
taan- 11.21.4, 18.4
tanm 18.4
tata- 33.3
tan 18.4
t 22.2.3, 37.4, 38.1
ti 22.2.3
tca 22.2.3
tpaiia- 26.15
t-/ta- 32.1
tt 11.17.4, 32.1
ta
22.2.3
ta
sc 22.2.3
ta
scit 22.2.3
t 22.2.3
tm 22.2.3, 37.2, 39.2
tuui- 18.8
tuui-c 7.15.1
tuu- 7.8.2
tm 22.2.3
t 22.2.3
t g 7.9, 22.2.3
t 7.11.3, 22.1, 22.1.5, 22.2.3
ti 22.1, 22.1.5, 22.2.3
tuu m 7.13.1, 7.17.5, 22.1
t 22.1, 22.1.2
tiriia- 8.2, 21, 21.2
tm 7.17.5, 22.1
tir- 21, 21.1
tirm 21.1
tir 21.1
anuuar 18.6
a- 10.2.2, 22.4
ax iia
22.4
at 22.1
ardm 9.2.4, 32.5
ars 9.2.4, 32.5
ar- 32.5
ahii 11.28, 22.4
ahmi 22.4
138 word index
ahmt 22.4
ahm 22.4
22.1, 22.1.5, 22.4
h 22.4
m 7.6.2, 11.17.1, 22.1
rsa- 26.7
22.4
i 22.4
rtar- 7.13.3, 9.1, 11.14.4
raot 11.26.3
raiias-ca 11.17.1, 21.1
raiim 21.1
r 9.2.4, 11.24.2, 27.3, 32.5
riia- 32.4, 33.1
riiete 32.4
rii 21.1, 38.3.d
riiidiii 33.1
rzdm 9.2.4, 11.24.2, 32.5
ra
h- 27.3, 32.5
ri- 21, 21.1
riim 21.1
rigm haca 36.1
rita- 21.2
ritiia- 21, 21.2
ridasa- 21.2
ribii 21.1
risat- 21, 21.1
risatanm 21.1
risatm 21.1
risaat- 21.1, 21.3
riss 21, 21.1, 40
ri 21.3
riuua- 7.17.5, 21.4
riu
m 7.17.5, 19.1.3
riat- 21.4
riuuat 21.3
r 19.5, 21.1
r 21.1
dauu- 19.3
dadi 26.2.b
dana
19.2, 19.2.7
danaiia 19.2, 19.2.7
danaiii 19.2
danaiit 19.2
danaiia
19.2
dan- 7.5.2, 19.2
dana
7.5.2
danm 19.2
dane 19.2
daoiia- 32.4
daie 32.2
daiiir 32.2
daiiia
32.2
daiiin 32.2
daita 32.2
daia 32.2
daidt 32.2
daie 32.2, 32.6
daiiit 32.2
daitm 32.2
dait 32.2
dain 32.2
dai 32.4
daibit 8.1, 9.2
daibitiia- 11.14.1, 21, 21.2
daibi 11.14.1
daibiaiiat- 8.1
daibit 11.14.1
da(i) haot 19.5.23
dai hauua 19.5.26
da(i) hauue 19.5.24
dai h 19.5.26
dai hu 19.5.27
dai huu 19.5.26
dait 32.1
daxma- 19.1
daxmauua 19.1, 19.1.21
dax ii u 10.4, 11.29, 19.5.22
dax iiu- 7.17.5, 19.5
dax iiunm 11.29: fn. 2.14,
19.5.35
dax iim 7.17.5, 11.29: fn. 2.14,
19.5.20
daaite 32.2
daat 32.2
daat- 32.2
dana- 32.2
dani 32.2
dama 32.2
dan 32.2
dadait 32.2
dad-/dad- 26.2.a 32.2
dad-/d-/da- 28, 32.6
dadt 32.2
dada 11.13, 32.6
dada
32.2
dadmi 7.6.3
dadn 11.10.1
dadmaid 32.2
dadmah 32.2
dad 32.2
daduii 32.1
daat- 32.2
daiti 32.2
dat 32.2
dahi 32.2
dam 32.2
dami 32.2
dauuh-/dadu- 30.1
dab 11.25.2, 32.3
dara- 11.26.3
da haom 11.29: fn. 2.14, 19.5.20
da huu 19.5.32, 19.5.33
da h u 10.4, 11.29
da hu- 10.3.4, 11.29: fn. 2.14,
19.5
da huu 10.3.4
da hubii 19.5, 19.5.36
dat- 32.1
dam- 18.1
dar 7.13.3, 10.1.2, 11.13, 11.17.4,
11.20.5, 32.3, 32.5
dar-/dr- 32.1
darga- 11.2.4, 11.12
dargmc 11.3.2
dara- 11.12
darra- 9.1
dars-/drs- 32.1
darsma 32.1
darsm 9.1 32.1
dar 32.5
dars 7.4.2, 9.1, 11.21.1
darz 11.25.5
dasa
11.13, 11.20.3, 21
dasanm 21.1
word index 139
dasma- 21, 21.2
dasuu 32.2
dasta 32.2
daste
11.11.2, 32.2
dazd 32.2
dazde
11.11.2
dazdi 32.2
dazdiii 11.24.3
daa- 11.25.6
dahk m 38.3
d 7.5.1, 7.6.3, 11.10.1, 11.11.2,
11.13, 11.24.3, 11.31, 25.1, 27.1,
28, 30.1, 32.2, 32.5, 33.3
d-/d- 27.1 32.1
diiata 32.1
did 32.1
di 10.5.1, 32.5
di-/di- 27.3, 32.5
duui 33.1
duru- 8.2, 10.4, 10.6.3, 19.5,
19.5.38
dta- 33.3
dtar- 7.3.1, 19.3
dtaras-ca 7.3.1
dt 32.1, 36.3 40.2
dtrm 7.4.4, 18.5
dtr 7.3.1
dtibiias-c 19.1.19
dt 32.1
da- 19.1.3
daibii 19.1.19
dm-ca 19.1.3
d m 19.13
dr- 19.3
ddarsa 37.1
dt 32.1
dman- 11.31
dmahuua 11.31
dm 32.1
dmn 18.4
dmm 18.4
draiia- 11.1.3, 26.15, 32.4
draiiam 32.4
draiiat 11.13
drt 7.13.3, 32.5
dh 32.1
dhuu 11.31, 32.1
dn 32.1
dm 7.6.2, 18.1.9
dmi- 7.15.2
dmi 18.1.9
dmm 7.15.2
dh 7.6.1, 11.10.1
dhita- 7.6.1
da
32.1
da
h- 32.5
da
h 32.1
da
h 7.5.1, 32.1
da
t 7.5.1, 32.1
dbnaot 32.3
dbnauu- 32.3
dmna- 9.1, 11.13
drn- 32.3
drnat- 32.3
drz- 18.1
drz 18.1.8
dr-c 18.1.1
drta- 11.21.1
d jmspa- 9.2
d g 18.1.5
diraiia
19.1.12
dirbiia 19.1.14
di 10.1: fn. 2.7
di 32.5
drt 7.13.3, 11.17.4, 11.20.5
diiao 10.4
dii 32.1
diim 32.1
didaii-/ddi- 32.2
dids 7.6.1, 11.10.1, 32.2
diaii-/di- 28
dit 22.1.6
dia- 11.25.2, 26.17
diaidiii 11.25.2
dim 22.1.6, 38.3, 39.2
dis 10.1: fn. 2.7, 10.5.1, 25.1.1,
27.3, 32.5
d 22.1.6, 28, 32.2
ddr 11.25.5
ddraa- 32.4
ddra.duii 32.4
dam 32.2
d 22.1.6
d 32.1
duiie 21.1
duua- 21, 21.1 35
duua 21.1
duuaibiia 21.1
duuaca 21.1
duuaah- 11.14.1
duuaiia
21.1
duuaid 32.1
duuadasa- 21, 21.2
duuadasa 21, 21.1
duuar 7.4.3
du 11.25.1, fn. 2.12
duuuacah- 11.25.1
dudh- 11.25.1
du-manah- 11.25.1, fn. 2.12
dumanah- 11.25.1, fn. 2.12
drt 19.1.6
/dbitiia-/ 21.2
drao 10.4, 19.5.40
drafa- 11.21.4
*drguua 5.4
drguuata-c 8.1
drguuat- 5.4, 11.12, 18.3
drguuite 8.1, 10.5.3
drguu.dbii 5.4
drguu.db 5.4, 7.13.2
driao 11.12
driu- 11.12
dri 11.12
drt 32.1
druua- 11.1.3
druuat bii 40.1
druuat- 11.12, 18.3
druu 18.3
druxta- 33.3
drux 18.1, 18.1.1
druj- 7.14.3, 18.1
druj 5.4, 11.25.6, 33.3
drujat 18.1
drujm 18.1.2
druj m 7.14.3, 18.1.2
140 word index
drujiia- 26.11.a
drujim 7.14.3, 18.1.2
drua- 26.11.a
druaiti 11.25.6
drj 18.1, 18.1.5
drm 19.1.3
a- 11.25.6
t kaa- 11.10.2
t bi 11.14.1
t biaiiat- 8.1
paoiriia- 8.2, 10.3.2, 21, 21.2
paoiriii 21.2
paoirm 19.1.3, 21.2
pa(o)uruiia- 8.2
paouruum 19.1.3
pauruua- 21.2
paite 19.5
paiti- 7.15.3, 19.5
paiti 19.5, 22.2.1, 23.14
paiti iia zm 22.2.1
paitiimraot 31
paiti-d 25.1.2
paitm 7.15.3
paie 19.5.6
pai 18.1.9, 36.3
pairi 22.1.6
pairii-ao 32.1
pairi.uruuata- 10.2: fn. 2.7
pairi
23.15
pac 11.9
pacata 11.9
pa- 18.1, 18.1.1, 18.1.19, 40.1
paa 18.1.8
paana- 20
paan- 20
paa
40.1
pa 11.17.3, 18.1.1, 18.1.5, 18.1.15,
40.1
pad- 18.1
padb 18.1.19
pa 18.1.15
pafrite 32.6
pafre 32.6
patahuua- 21.4
patavhum
paca
11.3.2, 21, 21.2

pacadasa 21.1
pacanm 21.1
pac-ca 21.1
pacca vsaiti- 21.1
pacsat- 21
pacsat b-ca 21.1
pat-/pa- 40.1
pat- 18.1.1
patnm 18.1.2
patn 18.1.14
pata
11.17.3, 18.1.1
patm 18.1.2, 40.1
par 7.10, 32.3, 32.6
para
23.16, 38.3
para ahmt tat 38.3
parao 19.5.22
para.kauuistma 5.4
para-d- 25.1.2
parna- 18.4
parnin- 18.4
pardi- 11.3.2
par 7.9.4
par 7.9.4, 23.17
par.kauum 5.4, 7.13
partm 19.1.3
part m 19.1.3
pasu- 7.12.1, 11.21.4
pasu vra 35
pasu-/fu-26.13
pasuua
19.5.29
pasuum 19.5.35
pasuu 19.5.22, 19.5.33
pasu.huua 19.5.38
pasca 23.18, 38.3
pasca yat 38.3
p- 32.1
pa 18.1.10
pm 18.1.2
pfr-/pafr- 32.6
pna
- 11.20.4
pnaibiia 19.1.14
pnbiia 19.1.4
pa
h 32.1
psnu- 7.6.1
prtu- 19.5
prtu 19.5.18
prt 19.5.33
pru- 11.2.1, 11.17.3, 20
pre 19.5.24
pr 19.5.33
pr- 20
prn 32.3
prn-/prn- 32.3
prnite 32.3.a
prnne 32.3.a
prne 32.3.a
prsa- 26.16, 32.4
prsaite
11.20.4
prsat 11.10.2, 37.2
prsauha 11.31
prsavha 11.31
prs 32.4.a
prsa
he 32.4.a
ps auu 19.5.33
ps an- 11.2.3
pouru- 7.12.1, 8.2
pouru-ca 19.5.28
pour 8.2
pita 7.14.4
pitar- 18.5
pitarm 7.4, 18.5
pire 10.2.2
pitra- 11.17.4, 11.20.5
puxa- 21, 21.2
pur 7.9.3
ptar- 9.1, 10.2.2, 11.10, 11.10.3,
11.17.2
fri 7.14.4, 9.1, 10.2.2, 11.10.3,
11.17.2
fra- 23.19
fraaotra- 11.17.4
frauu 32.5.c
fra
11.18, 21.2, 23.19

fra+nas 11.21.1
fra+nu- 11.21.1
fraorcita 32.4
fraiiaire 7.10
fra-uuart 9.2, 11.2.4
frauuas e 19.5.6
word index 141
fra-uuizdm 11.24.3
fra-carae 32.4
fra-carie 32.4
fra-tatk- 39.3
fratma- 21, 21.2
fra-d- 25.1.2
frapinaoiti 25.2
frapinuuata 25.2
fra-mma 32.2
fra-mrauua 32.1.d
fra-mrauuni 32.1.d
fras 11.10.2, 11.20.4, 11.21.1, 19.1.3,
32.5.a
fras- 11.21.2
frascibana- 11.3.2, 11.14
frasntaibiia zastaibiia 35
fra-zaht 32.5.c
fra- 32.5
fra 32.5.a
frana- 11.21.2
franu- 11.21.2
fr-rsata 32.4.d
frnaoiti 11.21.2
fra
c- 18.2
fr 7.6, 18.2
friia- 7.14: fn. 2.6, 11.1.2
friinmah 32.3.a
fr- 32.3
fr- 18.1
frn-/frn- 26.5. 32.3
frni 32.3.b
frnt 32.3
frnni 32.3.b
frnmi 32.3.a
frnti 32.3.a
frntu 32.3.e
fri
m 7.14: fn. 2.6

fu 11.21.4
fuiia- 26.13
fumat- 18.3
f carkmah 36.2
bauuani 18.6
bauuar 21.1
baon 32.4.c
baire 32.4.a
bauu-/b- 32.1
bauuaiti 11.1.3, 32.4.a
bauuait- 19.3
bauuat- 19.3
bauuatm-ca 7.8.3
bauuar-/bar- 28
baxta- 11.10.2: fn. 2.11, 11.16, 33.3
bax- 27.3, 32.5
baxait 32.5.b
baga- 11.12
baa- 11.12
baj 11.10.2: fn. 2.11, 11.16, 11.25.6,
27.3, 32.5, 33.3, 36.2
bariin 25.1.1, 32.6
ban 7.8.3
bad 11.3.2, 33.3
badaiieiti 11.3.2
bar 7.8.1, 11.10.1, 11.14, 25.1, 28
bara- 25.1.1, 26.6, 32.4
baraiin 32.4.d
baraiti 32.4.a
baraiti
8.1
baraite 32.4.a
barata 32.4.e
barat 32.4.a
baratu 32.4.e
baravha 32.4.e
barahi 32.4.a
barni 32.4.b
barma 32.4.b
barmaide 32.4.a
barmahi 32.4.a
barmi 32.4.a
barmi 30.1.a
barhi 32.4.b
bara
ti 32.4.b
barn 32.4.b
barriit 19.3
barriuua 19.3
barn 7.8, 11.10.1, 11.14
barti 32.4.a
bartu 32.4.e
barm 32.4.c
barmna- 32.4.f
barsmn 18.4
barzan- 11.25.4
barj 10.1
bar 18.1
basta- 33.3
baat 11.25.6
bbuuar 11.14.3
braiia- 25.1.1
bzauua 19.5.27
bzu 19.5.18
bzubii 19.5.37
bzue 19.5.30
bzu 19.5.18
bnaiin 7.8.3
brjam 10.1
brjaiin 10.1.1: fn. 2.8
brdubii 9.2
brz- 18.1
brzat- 18.3
brzm 18.1.2
brz 18.1.5
biiah- 18.7
bitiia- 11.14.1, 21, 21.2
bitiii 21.2
bitm 21.2
bi 21.2
biaziia- 26.13
biuuat 21.3
buiiama 32.1.e
buiir 30.1.b
buiin 30.1.b
buuat 11.10.2
bud 10.1.1: fn. 2.7
bura- 11.13
bu
na- 11.13
bujaiit 36.2
b 10.3.3, 11.1.4, 11.10.2, 32.1
biiiata 10.1.1: fn. 2.7
biiiimaie 10.1.1: fn. 2.7,
32.4.d, fn. 3.10
biri- 19.5, 19.5.3
brtar- 18.5
brtu
iriia- 8.2
bruuat .biim 15.3
na- 22.1.5
na-cit 22.5.1
142 word index
na-ci 22.5.1, 36.2, 39.1
na-ci
m 22.5.1
nad 39.1
naa 39.1
nania- 11.25.6
nama- 21.4
nastar- 10.2: fn. 2.9
na- 32.5
naat 32.5.b
naoma- 10.3, 21, 21.2
naire 10.2, 18.1.7
nauua 21
nauuaiti- 21
nauuaca nauuaiti- 21.1
nauuanm 21.1
nauuza- 7.3.3
napt- 11.10.3, 11.21.4, 18.2
naptm 18.2
napa
18.2
naptar-/[napt-] 11.10.3
napt 18.2
naptiia- 11.10.3
nafr 11.10.3
nafu 18.2
nafu-c 11.21.4
nab-nazdita- 11.18
nar- 9.1, 10.2, 11.1.4, 11.20.5, 18.1,
18.5, 19.3
nara
18.1, 18.1.12
narm 18.1.17
nar 18.1, 18.1.3
narm 18.1.2, 37.1.3, 37.1.4
nar 18.1.14
nari 10.2, 18.1.7
nar 11.20.5, 18.1.5
nas 11.21, 27.2b, 32.5, 36.2
nasuu 19.5.32
nasum 10.6.2, 19.1.3, 19.5.20
nase 32.4
nasiia- 32.4
nasiieiti 32.4.a
nasu- 10.6.2, 19.1.3
nas 19.5.33
nazdii 11.13
nazdita- 11.18, 11.24.2
na- 32.5
nata- 11.21
n 11.20, 18.1.1, 22.1, 38.2
nir- 19.3
nist 10.5.1, 11.17.4
nism 10.5.1
nfa- 11.18
nman- 7.6.2, 11.3.1, 11.3.3, 18.4
nmanm 7.6, 11.3.1
nmm 7.6.2, 11.3.2
nm ni 18.4
ns- 18.1
nm 32.5.b
nma 32.5.c
n 18.1
na
22.1, 22.1.5
nma 7.6.2, 7.7, 18.4
nsa- 27.2b
nmax iimah 11.28
nmaxvait 11.30
nmah 11.27
nmah- 7.8.3, 11.3, 11.27, 25.1
nm 18.7
nm 18.7
nrt 18.5
nrbiia 18.1, 18.1.13
nrbiias-c 18.1.18
nrbii 11.1.4, 18.1.18
nr 11.20.5, 18.1.5, 18.1.15
nr 9.1
nr 18.5
nruii 11.1.4, 18.1.18
n 22.1, 22.1.5
n 22.1, 22.1.5
nit 38.3.c, 39.1, 39.2
ni 23.20, 25.2
nii- 23.20
niia
c- 18.2
niia
c 18.2
niia
cim 18.2
ni-ne 32.1.a
nid 10.2: fn. 2.9, 10.5.1, 11.17.4
ni-d- 25.1.2
niptaiiaca 19.5.6
ni-sirinaoiti 9.2
ni-srinauuhi 32.3
nihida- 26.8
ni- 23.21
n 32.5
ni
23.21
nmna- 11.13
nmnau 19.1.21
nmnaiia-ca 19.1.9
nmnat haca 19.1.6
ma- 7.9.2, 22.1.1
maaibiia 19.1.14
masma- 10.2: fn. 2.9
maza- 10.2: fn. 2.9
maii 22.4
maiti- 33.2
maiiii 10.2
maibii 22.1, 22.1.3
maibii 22.1, 22.1.1
mainiia-/maiia- 26.10
mainiiauue 7.10
mainiiata 32.4.c
mainimaid 32.1.e
max iia
22.4
mada- 11.28
madahii 11.28
mau- 7.12
mat 11.26.2, 22.1, 23.22
mahna- 32.5.e
man 7.9.3, 13.1, 27.3, 32.1, 32.5,
32.6, 33.2
mana 22.1
manaha 7.9
manah 7.9, 11.27
manah- 7.8.3, 7.9.3, 11.25: fn.
2.12, 11.26.1, 11.27, 18.7
manahi-c 11.26.1
man.vista- 5.4: fn. 2.1
maniiata 32.4.d
maiia- 13.1, 26.10
maii u 10.4
maii 19.5
maii 19.5.19
maiiu- 7.10, 19.5
maiiu 7.8.3
maii 19.5
word index 143
maiuua
19.5
mat 32.1.b
mamn- 32.6
mamnit 32.6.a
marka- 11.2.2
marka-c 11.2.2
martaibii 19.1.19
martn 18.4
marc- 32.3
masiia- 11.7
maz- 9.2, 18.1
mazat- 7.5, 20.1.1
maza
tm 7.5
maz 18.1.5
mazi 18.1.7
maziiah- 20.1.1
mazib 9.2, 18.1, 18.1.19
mazita- 20.1.1
mazit am 7.9
mazit am 7.9
mazga- 11.12, 11.24.2
mazdaiiasna- 7.3.2
mazd- 11.24.2, 18.1
mazd 18.1.3, 36.1
mazda
18.1.1, 18.1.5, 36.2

mazda
h.dm 32.5
mazdm 18.1.2
mz-dazdm 32.2
mas iia- 11.2.3, 19.1
mas iia
19.1.15
mas iia 19.1.21
mas iii-ca
19.1.20
mas iinm 19.1.18
mas iia
h 19.1.15
mas iis-ca
19.1.16
mas ii g 19.1
mas m 19.1, 19.1.3
mahii 22.4
mahi
11.26.2, 32.1
mahmi 7.9, 22.2.5
mahm 22.4
mahrka- 11.2.2
m 22.1, 22.1.5, 22.4, 32.2, 39.2
muuiia 22.1, 22.1.1, 22.1.7
mca 39.2
mtar- 18.5
mh- 7.5, 11.28, 18.1
mhiia- 11.28
ma
18.1.1
ma
hm 7.5, 18.1.2
ma
h 18.1.5, 18.1.14, 18.1.15

mra- 7.6, 19.1
mri 19.1.20
mrn- 18.4
mrn 18.4
mrs-ca 19.1.16
mm 7.6, 22.1, 22.1.1
msta 32.5.a
mriiiao 10.4, 19.5.22
mriiiu- 8.1, 9.1, 11.17.1
mrte 32.3.a
mrgduii 32.3.a
mrc- 32.3
mrcait 32.3.a
mrciti 32.3.a
mrcta 32.3.d
m 7.9.4, 22.4
m.n 22.1
m ni 32.1.d
m gh-/m h- 27.3, 32.5
m gh 32.1.b
m ghi 7.9, 32.5.b
m gh 32.5.a
m 22.1, 22.1.5, 36.2
mourum 11.12
mou 7.12
mou-c 11.21.4
mi 22.1, 22.1.5, 36.3
mi-/mi- 32.1
miat 10.1.1: fn. 2.7, 36.2
mist 32.1.b
mist 11.17.4
miias- 32.3
mi 10.1.1: fn. 2.7, 10.2: fn. 2.9,
11.17.4, 36.2
mira- 19.1
mira ahura 35
miribii 19.1.19
minas- 32.3
mina 32.3.c
mim- 32.2
miriia- 26.11b
mda- 7.15, 11.20.5
mraoiti 11.3, 32.1.a
mraot 32.1.a
mraotu 10.3.1, 32.1.f
mraot 10.3
mraot 32.1.a
mraom 10.3, 32.1.a
mraomi 25.1.1
mrao 30.1, 32.1.a
mrauu-/mr- 32.1
mrauune 32.1.d
mruiit 32.1.e
mruiia
32.1.e
mruiie 32.1.a
mruii 7.11, 10.2, 25.1.1, 32.1.a
mruuat- 32.1.g
mruuna- 32.1.g
mruu 32.1.e
mr 10.2, 10.3, 11.3.1, 25.1.1, 26.1
mrite 32.1.a, 33.1
mrii 32.1.f
mrmaide 32.1.a
ya- 7.5, 7.10, 7.14, 11.31, 22.2.1,
22.3, 22.5.2
yaibii 22.1.1
yatatar 32.6.a
yama 32.6.a
yaiia- 26.12
yaog-/yuj- 11.10.2, 32.1
yaogt 11.10.2
yaojat 32.1.d
yaoj 32.1
yao 19.5.40
yao-daiti 32.2
yao-dane 32.2
yao-dat- 32.2
yao-date 32.2
yao-da 32.2
yao-dit 32.1.d
yaiiat-/yat-/yit- 32.6
yauuajii 18.1.15
yauuajibii 18.1.18
yauua-jim 18.1.2
144 word index
yauuatt- 18.2
yauua-sum 18.1.2
yauuata
38.3.d
yauuat 38.3.d
yauu 19.5.42
yauukm 22.1, 22.1.3
yauu 19.5.41
yauui 19.5.41
yat 32.6
yaa
37, 38.3
yara 38.3.d
yad 38.3.d
yat 22.3, 38.3
yat cit 38.3
yatu 32.1.f
yasa- 32.4
yas-ca
19.1.1
yasna- 7.10, 11.21.2, 19.1
yasnanm 19.1.18
yasnas-ca 19.1.1
yasns-ca
19.1.16
yasnibii 19.1
yaz 11.21.2, 33.3
yaza- 26.6, 32.4
yazaa 32.4d, 37.3
yazaiiata 32.4.d, 37.4.e
yazatanm 36.2
yazamaide 19.1.8, 37.3
yazite 32.4.b
yata- 33.3
yahmat 36.2
yahm 7.10
yakar 18.6
yta- 19.1.7
ytii 19.1.7
ytu 19.5.37
yt 11.10.2
yr 18.6
ysanuha 11.31
ysaha 11.31
ysavha 11.31
ysmi 35
yhuua 11.31
ya
hm 7.5
ys 22.3
y g 22.3
y gs-t 22.3
y m 7.14
y ma- 7.14
ye(i)di (z) 37.4.c, 38.3.c
yeiti 32.1.a
yex iic 22.3
ye h-a 22.3
ye ht 22.3
ye ha
22.3
ye he 22.3
yesne
7.10, 19.1
yesniia- 7.10
yesnii-ca
19.1.17
yesnm 19.1.3
yezi
38.3.c, d
yehii 7.10, 22.3
y 22.3
y 36.2
yi 10.2, 22.3, 36.3
yim 7.14
yima- 7.14
yuj 11.10.2
yiiia- 32.4
yiiia 32.4.a
yj n 32.1.b
y 22.1, 22.1.4
yma- 22.1.4
ymaibii 22.1.4
ymat 11.10.2, 22.1
ymka- 22.4
ymkahii 22.4
ymki 22.4
ymkm 22.1, 22.4
ym 22.1, 22.1.2, 22.1.4
y m 22.1, 22.1.4
va- 22.1.5
va 21.1 fn. 3.7
vaa- 10.2
vaat 32.6.b
vati 32.6.b
vad-/vid-/vd- 28, 32.6
vada
7.11, 10.1
vad 32.6.a
vaa 32.6.a
vana h 35
vanimaid 32.4.d
vam 10.1, 22.1, 22.1.2, 22.1.3
vaoca- 27.2b, 32.4
vaocim-ca 10.1: fn. 2.7
vaoniit 32.6.c
vaorz- 32.6
vaorza 32.6
vaozirm 32.6.b
vaiiu- 7.3, 10.3.4
vaint 32.1.e
vauuaxda 11.17.2, 32.6.a
vauuac-/vaoc- 28, 32.6
vauuan-/vaon- 32.6
vauuanuuh-/vaonu- 18.8
vauuaz-/vauz- 32.6
vaxra- 11.17.2
vax- 25.1.1
vax 11.20.5
vaxaiia- 25.1, 26.15
vaxiia- 26.18
vaxii 11.20.5
vaxt 11.20.4
vax iia
18.7
vaibii-ca 18.1.13
vaibi 18.1, 18.1.19
vac- 13.1, 18.1
vac 11.10.2: fn. 2.11, 11.20.5, 13.1,
26.18, 27.2b, 28, 32.1.c, 32.6
vaca 18.1.8
vacah- 5.4, 7.9.4, 7.13, 18.7
vac 7.9.4, 7.13, 18.7
vac b 7.9.4
vac 7.13, 18.1.5, 18.1.14, 18.1.15,
18.7
vadar
9.1
vare 11.13
vaf 11.11.1
vaaka- 11.25.2
vanuh 11.31
vah-/v gh- 32.5
vahauue
19.5
vahauu 11.27
vahu 10.6.1, 19.5, 19.5.26
vaht- 32.5
word index 145
vah u 11.27, 11.29: fn. 2.14
vah 11.31
vahu- 20.1.1
vahuum 19.5.35
vahu 11.27, 11.29: fn. 2.14
vahuu 19.5
vah 11.27, 19.5.33
va hahm 18.7
va hah- 20.1.1
va ha
18.7
va h 11.29
vavhi 19.5, 19.5.28
vavhiia
19.3
vavh- 19.3
vavhbii 19.3.12
vavhnm 19.3.11
van- 32.1
van 7.6, 7.9, 11.10.1, 11.17.4, 18.8,
32.5, 32.6, 40.1
var- 32.3
1var- 32.1
2var- 32.1
var 9.2, 11.2.4
varat 9.2
varn 32.1.d
varn 32.1.d
varc.hc 18.7
varduua- 11.13
varmaid 32.1.b
var- 32.5
varait 32.5.b
varait 32.5.b
var 32.5.b
varn 32.5.b
vart 32.5.b
var-c 32.1.b
varz- 32.1
vard 11.11.2
varz 7.3, 28, 32.5, 35
vas 11.20.5, 11.21, 11.21.2, 11.21.4
vas-/us- 11.21.2, 32.1
vasah- 11.21.1
vasm 11.21
vast 11.20
vasna- 11.21.2
vasn 11.21.2
vaz 7.8, 11.25.5, 32.5, 32.6
vazti 7.8
vaata 32.5.b
vaa
te 32.5
vai
11.20.5, 11.21.4, 32.1.a

vat 11.21.1, 11.21.4
va-/va- 32.5
vadra- 11.25.3
vah 11.20.1
vaheh 11.1.1
vahiiah- 11.1.1, 11.28, 18.7, 20.1.1
vahii 11.29
vahita- 20.1.1
v- 32.1
v 21.1: fn. 3.7, 22.1, 22.1.3, 38.2
vuuarz-/vuurz- 28
vx 11.20.5, 18.1.1
vibii 18.1.18
vcm 18.1.2
vcim 18.1.2
vc 25.2, 32.1
vahe 36.2
vs m 7.4
va
22.1, 22.1.5
va
ti 32.1.a
va 40.1
v- 19.2, 40.1
vbiia 19.2
vs 7.6, 11.10.1, 11.17.4, 32.5.a
vrca- 32.4
vrra-jan- 20.1.2
vrraj 18.1.1
vrrajstma- 20.1.2
vrd- 18.1
vrd 18.1
vrn- 32.3
vrnauu-/vrnu- 32.3
vrnauuait 32.3.b
vrnt 32.3.c
vrnite 32.3.a
vrta 32.3.a
vrt 32.3.a
vrz na- 7.9
vrziia- 26.11a, 32.4
vrziiat-/vrzit- 32.4.f
vrziitm 7.4, 32.4.e
vrzda- 11.11.2, 11.24.3
vhrka- 11.2.2
vhrkat 19.1.6
v 7.9.a, 7.13, 22.1, 22.1.5
v ghait 7.9
v ghat 7.9, 32.5.b
v ghn 32.5.b
vouru 11.2.1
vohu- 7.12, 10.6.1, 11.1.1, 11.28,
11.29, 18.7, 19.3, 19.5
vohu 11.29: fn. 2.14, 19.5, 19.5.21
vohu-ca 19.5.28
vohu-c 19.5.21, 19.5.25
vohunm 11.29: fn. 2.14, 19.5,
19.35
voh 11.27, 19.5, 19.5.25, 19.5.34
vohm 11.29: fn. 2.14
v 7.13, 22.1, 22.1.5
vin- 10.2
vinuii 19.2.13
vist 10.1, 11.20.2, 32.6.a
vi- 23.23
vii- 23.23
viixmainiia- 26.13
viixman- 26.13
vii-darsm 7.4.3
vid 7.15, 10.1, 11.20.2, 11.24.3,
18.8, 27.2, 28, 30.1.e, 32.3,
32.6, 33.1
vinad-/vid- 26.3, 32.3
vid 11.10.1
vidna- 32.3.e
vidn 11.10.1, 32.3.c
vista- 11.20.2
v 7.15, 23.23
v-caiia 32.1.d
vda- 27.2
vdauua- 11.13
v-d- 25.1.2
v-dram 10.1.2
vduuani 33.1
vduuh-/vdu- 18.8, 30.1.e
vduua
18.8
146 word index
v-ram 32.4.c
v-raiin 10.1.2: fn. 2.8
v-raiii 32.4.d
vuuh-/vdu- 32.6
vuua
18.8
vuua
hm 18.8
vra- 19.1
vraiia
19.1, 19.1.12
vs- 18.1
vsa- 32.4
vsa 18.1.8
vsate 32.4.a
vsaiti 21, 21.1
vsaitiuuat- 21.3
vsat 18.1.6
*vsastma- 21.2
vsm 18.1.17
vsstma- 21, 21.2
vsta 32.4.c
vste 32.4.a
vsm 18.1.2
vse 18.1.7
vs 18.1, 18.1.5, 18.1.15
vsi 18.1.9
vsiia 18.1.9
vspa- 7.9.3, 19.1
vspaibiias-ca 19.1.19
vspaca
19.1.9
vspam 20.2
vspau
19.1, 19.1.21
vspanm 20.2
vsp.voh 5.4
vspi 20.2
vspa
h 19.1.15, 20.2
vspm ahmt yat 38.3
vspmi 20.2
vspmca 11.3.2
vsp 7.9
vsp g 7.9
vsp m i 20.2
vsp s-ca 19.1.16
vspe 20.2
vspibii 19.1.19
v 18.1.1
va- 11.20.5
v-s
iit 32.1.b
vibii 18.1, 18.1.18
rax- 32.5
raxa 32.5.c
raa- 10.2: fn. 2.9
rarizaite 11.2.1
ram 10.1.2, 19.5.2
raocaiia- 26.15
raocah- 7.9, 10.3, 11.2.1
raoca
18.7
raoc bii 7.9.4
raoc b 7.9.4
raom 11.12
raos- 27.3
ra
ii- 10.2
raire 32.6
ratu-frii 18.1.14
ratu-friu 18.1.20
ratu-fri
18.1.1
ratu 11.20.5
raa- 11.2.1, 11.17.3
raatar- 40.2
raat- 10.2: fn. 2.9, 40.2
raatrahe 40.2
raatrm 40.2
raatr sca 40.2
raatr 40.2
rait- 10.2: fn. 2.9
ra 19.5.22
rap 11.10.3, 11.17.2
rafra- 11.10.3, 11.17.2
rar- 32.6
ranao 10.4, 19.5.22
rah 7.8, 11.10.1, 36.2
r 32.5, 32.6, 33.3
rii 10.5.2
rta- 33.3
rd- 32.1
rdt 32.1.d
rna- 19.1.12
rnibii 19.1.14
rmim 32.4.d
rriin 11.10.1
rh-/ra
h- 32.5
ra
haiin 7.8.3
ra
hahi 32.5.b
rnaiia
19.1.12
ruu- 7.8, 11.3.2
rjita- 11.3.2
rma- 11.26.3
ric 8.1
riz 11.2.1
rud 11.17.4
sa- 21.3, 22.1.6
sate 25.2
sare 32.1.a
saii- 32.1
saiianm 7.8: fn. 2.3
saite 19.1, 19.1.11
sair 18.1
sauuaiia- 26.15
saxvr 18.6, 35
sata- 21, 21.1
satm 11.7, 11.20.3, 10.20.5
sattma- 21.2
saaiia- 26.14
saha- 7.9
sad 11.17.4, 11.24.3, 32.5
sar- 7.9.4, 18.1
sar 7.9.4, 18.1.5
sar m 18.1.2
sar 18.1.5
sari 18.1, 18.1.7
sazdiii 11.24.3
suuaiia- 26.15
sxv n 18.6
sstar- 11.20.3
ssn- 7.5.: fn. 2.2, 11.20, 19.2
sh- 32.1
sh 11.20, 11.20.3, 27.2
sht 32.1.e
ss 11.17.4, 32.5.a
sst 32.5.d
sh 11.24.3
suuita- 7.8, 7.15
s gha- 7.9.3, 11.3.2
s ghait 8.1
sa- 27.2
sunm 18.1.17
surunaoiti 9.2
word index 147
surunauu-/surunu- 26.4, 32.3
surunuiia
32.3.d
surunuuat-/surunuuat- 18.3
susru- 28, 32.6
susruiie 32.6.a
s- 18.1
s 26.15
sne 18.1.7
sn 18.1.5
sn- 19.3
sra- 7.8
ssrma 32.6.a
skda- 11.20
scat 11.20, 32.1.f
staota 32.1.f
stauu-/stu- 32.1
star- 9.2, 18.1, 18.5
star m-ca 9.2
st 11.20, 32.1.a, 32.5
st- 18.1
stuuita- 7.15
stnauua 19.1.21
sta
h- 32.5
sta
haiti 32.5.b
strn-/strn- 32.3
strniia
32.3.d
st 32.1.a
sti 33.1
stt- 18.1
str mca 9.2, 18.1.17
str 18.5
spainiiahm 7.8.3
spainiiah- 7.8
spad 19.1.10
span- 18.1, 19.3
spaniia
7.8.3
sparz 11.20
spas- 18.1, 32.3
spasm 18.1.2
spas 18.1, 18.1.14
spa 18.1.1
spanauu-/spanu- 32.3
spanu 32.3.a
sp 18.1.1
spna 18.1.10
spnm 18.1.2
spn 18.1.14
spta- 7.8.3, 11.3.2
sptaibii 19.1.19
sptax ii-c 19.1.5
sptat 19.1.6
sptahii 19.1.5
spt.tma- 5.4
snaa- 11.25.6
sna 11.25.6
snat 11.25.6
snait- 11.25.6
srata- 10.2: fn. 2.9, 11.20.3,
20.1.1
sraot 10.3.1
sraiiavha 36.2
sraiiah- 11.1.3, 11.20.3, 20.1.1
srauu-/sru- 27.1, 32.1
sr(a)uuana- 32.1.g
sruuaiia- 32.4
sruuaiieni 32.4.b
sruuahiia- 7.10
sruuahiieit 7.10
sruu 25.2, 32.1.c
srraiia
32.4.b
sruum 32.1.e,
sri 9.2, 32.3
srinaota 32.3.e
srinauu-/srinu- 32.3
srra- 11.1.3, 11.20.3, 20.1.1
sru 9.2, 10.3, 11, 20.3, 27.1, 28,
32.1.c, 32.3, 32.6, 33.1
sridiii 33.1
zaotar- 11.24.1, 18.5
zaoraiit 19.2.5
zaor- 19.2
zaorbiiasca 19.2.13
zaozao- 26.2b
zaoa- 11.24.1
zaiia- 25.1.1, 26.11b, 32.4
zaiiata 32.4.c
zaiiaha 32.4.c
zaiianmca 36.1
zaiia
te 32.4.b, 36.3
zairimiia- 8.1
zauutiia
h 19.1.15
zan 11.24.1, 25.1, 26.18
zanauua 19.5.27
zaga- 11.12
zam- 18.1
zasta- 10.3.4, 19.1
zasta 19.1.10
zastaiia 19.1.9
zastaii 10.3.4, 19.1, 19.1.13
zasti 19.1.20
zast 19.1
zastibii 19.1, 19.1.14
zaz- 32.2
zazti 32.2
zaz-/zaz- 32.2
zazite 32.2
zazmi 11.24.1
zazhi 32.2
zazuua 18.8
z 11.24.1, 32.2, 32.5
zta- 7.4, 11.24.1
znu- 11.25.3, 19.5
znu 19.5.38, 19.5.39
zh- 32.5
zm 18.1.2, 36.1, 39.2
zhiia- 25.1, 26.18
zma
t 18.1.6, 36.2, 37
zmas-ca 18.1.15
zm 18.1.8, 22.2, 37
zma 18.1.6
zmm-ca 18.1.17
zm 18.1.9
zm 18.1.5, 18.1.14
zmi 18.1.9
zrd- 18.1
zrd-c 18.1.8
ziia
11.24.1
ziim 18.1.2
zim- 18.1
zim 18.1.5, 18.1.14
z 35, 37.4.c, 38.1, 38.3.c
zzana- 7.15, 25.1
zuta- 11.20.4
z 11.14.2
zd 11.24.2, 32.1.f
148 word index
zbaiia- 26.14
zbaiieiti 11.14.2
zbar 11.14.2
zbarmna- 11.14.2
zraiiah- 11.24.1, 18.7
zruuan- 7.17, 18.4
zr 7.17, 18.4
at 11.21.6
aii-/i- 32.1
22.1.6
ira- 11.21.6
i 11.21.6
iis 30.1.e, 32.1.g
iieit- 30.1.e, 32.1.g
ti 18.1.7
ta
18.1.1, 18.1.5, 18.1.14

tm 18.1.2
te 18.1.7
s
to 11.23
s
iiaoana- 9.2, 11.23, 19.1

s
iiao(a)nanm 19.1
s
iiaoan 19.1.17
s
iiaoane 19.1.9
s
iiaoani 19.1, 19.1.9, 19.1.11

s
iiaona- 9.2, 11.23, 19.1

s
iiaona 19.1.17
s
iiaoni 19.1
s
ii 11.23
s
iito 11.23
ntar- 11.25.3
nubiias-cit 11.21.2, 11.25.3,
19.5.43
nm 19.5.39
ha- 23.27
hat 7.13.4, 10.3.4
hatu- 10.3.4
han- 40.1
haoma- 19.1
haom 7.9, 19.1
haoms-ca 19.1.16
hait- 30.1.e, 32.1.g
haiiia- 7.14: fn. 2.6, 7.15
haiii m 7.14: fn. 2.6
haim 7.14: fn. 2.6, 7.15
*hau 10.3.4, 22.2.5
hauua- 22.4
hauuanm 22.4
hauue 22.4
hauruua- 7.17, 11.1.3, 11.26.1, 19.1
hauruuatt- 8.2, 18.2
haurm 7.16, 19.1.3
hakrt 21.3
haxa 11.16, 19.5.1
haxai- 19.5.6
haxaiia 19.5.9
haxm 9.1
haxtiia
19.5.11
haxmain 8.1
haxman- 8.1, 9.1, 11.16
hac 7.5, 7.9, 8.1, 11.20, 11.26.1,
32.2
hac- 18.1
hac-/sc- 32.1
haca- 26.6, 32.4
haca 7.3, 19.1.6, 23.24, 36.1
hacaite
8.1
hacaiti 25.2
hacaite 25.2
haca
t 7.5
hac na- 7.9
hac mn 7.9
hacite 7.14, 32.4.a
hacimna- 32.4.f
hara 23.26
had 11.17.4, 28
had 23.25
hadi- 18.8
haa 23.25
hapta 11.10.3, 11.26.1, 21
haptaa- 21, 21.2
haptahuua- 21.4
haptiti- 21, 21.1
han 10.1: fn. 2.7
hana- 27.2
hanamc 10.1: fn. 2.7
ham- 18.1, 23.27
ham+kar 11.3.2
hama 18.1.8
hamastar- 10.2: fn. 2.9
hascit 22.2.2
hastra- 11.17.4
hazara- 21.1
hazare 19.1.11
hazartma- 21.2
hazah- 7.9.4
haz 7.9
hazd- 28
has
a 19.5.7
has
m 19.5.16
has
e 11.16, 19.5.6
hah- 26.1
h 11.24, 22.2.2
hu 22.2.5
hm(-) 23.27
hminmca 36.1
hs 11.20.2
h
(n) 23.27
ht- 30.1
ht-/hat- 18.3
hti 32.1
ht 7.8, 25.1
h-duurt 7.4
h
m 23.27
hrzaiin 39.2
h 22.2.2
h kriti- 11.3.2
h -grabm 11.1.4
h t 32.1
h miisait 7.4
h 22.2.2
h 22.1.6
h
hiiat 22.3, 38.3
hiir 32.1
hiniii 19.5.36
hica- 26.9
hizuu- 19.4
hizuua
19.4.6
hizuu 19.4.3
hizub 19.4.11
hiz- 19.4
hicamaid 32.2
hita- 26.8, 32.4
hitahe 32.4
hi.hak-/hic- 26.2.a, 32.2
word index 149
h 22.1.6
hm 22.1.6
h 22.1.6
hu 32.3
huuar
9.1, 18.6, 38.3

huuar 11.30
huuuuiia 22.1.7
huu 10.3, 22.2.5
huxratu- 10.3, 11.16
huxratuu 10.3
hupur- 20
hunauu-/hunu- 32.3
hunauuat 32.3
hun-/hun- 32.3
hunmi 7.6
hunhi 32.3
hunuiir 32.3
hunuuana- 32.3
hunut 32.3
hunta 32.3
humanah- 18.7
humiia- 10.5.2, 19.1.3
humm 10.5.2, 19.1.3
hu.haxim 19.5.2
h 7.17, 18.6
hxta- 11.10.2: fn. 2.11
hr 18.6
-h yat 22.3
Old Church Slavic
eny 19.2.11
imena 18.4
mene 22.1.1
ni- 39.1
oba 21.1
on 22.2.4
ov 22.2.5
tx 22.2.1
cto, kto 22.5
Old English
hweol 11.9
Gothic
gibos 19.2.11
js 22.1.2
mawi, mauja 19.3.2: fn.
3.6
sa, ata, so 22.2.3
un- 39.3
undar 23.4
waurkjan 26.11.a
weis 22.1.2
Greek
-/- 39.3
7.6.1
11.20.5
11.20.3
11.26.2
23.7
21.1
21.3
23.5
11.12.2
33.3
11.12.2
11.12
19.2.11: fn. 3.5
19.2.11: fn. 3.5
18.4
11.24.1
11.25.3
11.25.4, 19.5.39
7.6.1
8.1
11.14.1
11.13
33.1
11.12
10.6.3, 19.5.39
19.5.40
19.2.3
11.25.1
11.25.1: fn. 2.12
21.1
7.4.4
11.26.1
33.1
18.8
11.26.2
21.3
7.8.1
- 31
22.4
- 33.1
11.2.4
23.2
11.26.1
11.26.1
- 7.4.3
11.20.1
23.1
11.2.1
10.3.1
11.2.11: fn. 3.5
20.1.1
- 20.1.1
- 20.1.1
11.2.2
11.12
- 30.1.c
21.4
- 19.3.1
22.1.6
11.24.2
11.20.1
7.8.3
25.2
11.24.1
11.20.3
11.16
11.21.6
11.9
11.2.1
10.3.1
19.1.10
31
7.12.1
39.2
11.20.5
11.3.1
19.2.3
150 word index
, , 22.2.2
21.1
19.1.9
19.1.9
10.1.1
11.20.2
21.1
10.2.2: fn. 2.9
22.3
11.26.1
7.4.4
11.3.2
23.15
11.9
11.2.1, 11.17.3
19.516
20.1.2
, , 23.14
23.19
20.1.2
20.1.2
11.20.5
, 19.2.11
22.4
- 20.1.2
- 11.9, 38.2
- 20.1.2
11.17.1
, 22.5
11.13
19.2.4
7.11.3
21.1
21.2
- 30.1.c
18.6
33.1.4
23.11
23.10
11.11.1
11.14
18.1.9
, 21.1
11.24.1
11.24.1
19.2.6
- 30.1.a
Hittite
kitta 25.2
uttr 18.6.1
Old Irish
cethoir, toir 21.1
tricho 21.1
Latin
ab 23.5
amb- 23.3
amb 21.1
angustus 7.6.1
augre 10.3.1
c-rus 11.9
centum 11.20.3
clure 11.20.3
coqu 11.9
crre 7.15.3
decem 11.20.3
diem 19.2.11: fn. 3.5
dis- 23.23
d 11.13
fc 11.13
familis 19.2.4
formus 11.12
genus 11.12.2
gntus 7.4
gust 11.20.5, 11.24.1
hiems 11.24.1
iecur 18.6
im 22.2.1
in- 39.3
inter 23.8
intimus 20.1.2
ra 10.1.1
is, ea, id, im, erum, erum
22.2.1
ling
locus 19.1.15
loca 19.1.15
lx 10.3.1, 11.2.1
marium 19.5.17
mlle 21.1
mox 11.21.4
ne 38.1
nemus 11.3.1
n 39.1
nmen 7.6.2, 11.3.1
nmina 18.4
ns, nostrum 22.1.5
- 30.1.a
pater 11.10
pecus 11.21.5
perna 11.20.5
pns 11.17.4
pistus 11.20.5
ps 18.1.2
posc 11.20.4
precor 11.20.4
pro 11.18
-que 11.9, 38.2
quitus 11.23
quisque 22.5
quid 11.9
rota 11.17.3
saluus 11.26.1
sequor 11.26.1
socius 11.17
st 11.20.1
sub 23.10
sunt 7.8.1
super 23.11
-t(d) 30.1.c
trs 11.17.1
tuus 22.4
ueh 11.25.3
uentus 33.3
ui
uis 19.2.4, 19.2.11
uiae 19.2.6
uituls 19.2.11
ullus, ull 20.2
us, uostrum 22.1.5
r 10.3.1
ucem 18.1.2
word index 151
Lithuanian
ans 22.2.4
deimts 11.20.3
mergs 19.2.11
im tas 11.20.3
Oscan
deva 19.2.6
Pli
-mina 30.2
Old Persian
adam 22.1.1
a-draiya 11.1.3
an 22.24
aniya- 11.1.3
-naiy 30.2.c
antara 11.3.2
rtv 18.4
ava, avam, avahy, avaiy,
avai, avm 22.2.5
*atauva- 21.4
azd 11.24.3
-nm 7.3.1
-y 19.2.4
-ha 16.1.1, 19.1.15
iyam, ima, an 22.2.1
ub 21.1
-nm 19.5.35
aiva- 21.1
kas-ciy 22.5
krta- 11.2.1
*auuva- 21.4
*iuva- 21.4
xn 11.16
-tam 30.1.b
tuva-
tuvam 22.1.2
-taiy 22.1.5
dahyva 19.5.32
-dim, d 22.1.6
naiy 39.1
*panauva- 21.4
paruv 7.12.1
paruviya- 10.3.2
bar
barmiy 31
abaram 31
abaran 7.8.1, 11.14
bauvatiy 11.1.3
ma-
mm 7.6.2, 22.1.1
man 22.1.1
-maiy 22.1.5
Margum 11.12
-mna- 30.2.e
rautah- 11.26.3
-im, , aiy 22.1.6
van 11.21.2
hauv 10.3.4, 22.2.5
haruva- 11.1.3
Prkrit
jhara 11.25.5: fn. 2.13
Proto-Indo-European
*b 11.14
*bh 11.14, 11.18
*bher 7.8.1, 11.14, 25.1.1
*d 11.13
*-dd- 11.11.2
*deh
3
7.5.1, 11.13
*dem 11.13, 11.20.3
*di --s 10.5.1
*deru -os 19.5.40
*dm tm 11.20.3, 21.1
*dl h
1
gh-
*dreu -s 19.5.40
*du ei s 8.1
*du - 11.14
*du i- 11.14
*du i-dm ti 21.1
*dh 11.13
*dheh
1
7.5.1, 11.13
*dh
1
-t- 33.3
*d
(h)
h 11.25.5: fn 2.13
*dhg
u h
er 11.25.5: fn. 2.13
*-dhi 30.1.c
DhT, DhS 11.11.1
*dhreu gh 5.4
*e 7, 7.3, 11.7, 11.9, 13.1
*- 31
*-eh
2
19.1.15, 19.1.17, 19.2.1
*-eh
2
-ei 19.2.6
*-eh
2
-es 16.1.2, 19.2.4, 19.2.11
*-eh
2
-ih
1
19.2.8
*-eh
2
19.2.11, 19.2.11: fn. 3.5
*-ei 14.6, 18.1.7
*-enh
2
18.6.3
*-es 16.1, 14.5, 18.1.14
*-eu -es 19.5.32
*-eu m *-m 19.2.11: fn. 3.5
* 7, 7.4, 11.7
*-i 19.5.8
*-u -s 19.5.18
* 11.7, 11.24.1
*enh
1
25.1.1
*eu s 11.24.1
*n 11.16
*neh
3
11.25.4
*n h
1
-t- 7.4
*rh
2
-u en- 18.4
*h 11.7, 11.24.1
*heslii o- 21.1
*heu 11.24.1
*heu H 11.14.2
*
(h)
u 11.14.2
*hu er 11.14.2
*g 11.7, 11.12
*gh 11.7, 11.12
*g
u
11.7, 11.12
*g
u
em 11.12.2
*g
u
m -se- 11.12.2
*g
u
iHeh
2
- 11.12.2
*g
u
eh
3
us 11.12
*g
u h
11.7, 11.12
*g
u h
en 11.12.2
*g
u h
or-m- 11.12
*h
13
7.14.4
*-h
1
14.7, 15.1, 19.1.8
*h
1
- 31
*h
1
eu o- 11.10.3
*h
1
ei -/*h
1
i-, *h
1
e- 22.2.1
*h
1
ono- 22.2.4
152 word index
*h
1
es-
*h
1
es-ti, h
1
s-nti 25.1.1
*h
1
i-i h
1
-t 11.1.3
*h
1
leng
u h
7.8.2, 11.3.2
*h
1
mo- 22.4
*-h
2
16.1.2, 16.3.3, 19.4
*h
2
mhes- 7.6.1
*h
2
enk 7.6.1
*h
2
ep-nes- 11.18
*h
2
u eg-s 11.20.5
*h
2
u es 25.1.1
*h
2
eu o- 22.2.5
*-h
3
k
u
- 18.2: fn. 3.3
*h
3
mei h 10.2.2: fn. 2.9
*i 7, 11.7
*i 10.11.1
*-i eh
1
/*ih
1
- 29
*-i eh
2
-ei 19.3.6
*-i eh
2
0es 19.3.10
*-i eh
2
-h
1
19.3.7
*-i eh
2
m 19.3.2: fn. 3.6
*-i eh
2
s 19.3.4
*i eu K-t 11.10.2
*-ih
1
15.1, 15.2
*-ih
2
19.2, 19.3.1, 19.5.15, 20
*-ih
2
/-i eh
2
19.3
*-ih
2
-m 19.3.2
*im 22.2.1
*-in-s 19.5.14
*-i-om 19.5.16
* 7
* 11.7, 11.20.3
*leu 26.4
*ou -i e- 26.15
*mt- 21.1
*u 11.10.3
*k 11.17, 11.9, 11.16
*keh
2
11.9
*ki eu 11.23
*k
u
11.7, 11.9, 11.16
*-k
u
e 11.9
*k
u
e, *k
u
id 22.5
*k
u
ei s 7.15.3
*k
u
k
u
lo- 11.9
*k
u
etesr- 21.1
*k
u
tores 11.17.1
*k
u
id, *k
u
o-s 22.5
*k
u
tur
*k
u
od 11.9
*l 11.2.1, 11.9
*leu k- 10.3.1
*-l/n- 18.6: fn. 3.4
*m 11.3.1, 11.9
*m 7.3
*mn i e/o- 13.1
*mleu H 25.1.1
mlu H-ti 11.3.1
*mn s-dheh
1
11.24.2
*-m- 21.2
*mosgh- 11.12
*n 11.3.1, 11.9
*n 7.3, 7.4.4
*n- 39.3
*nei 39.1
*n -g
u
h
2
-h
2
- 19.4
*n H 7.4
*ni 7.14.3
*ns 11.27
*nsdno- 11.13
*-n s 17.2, 18.1.15
*n sme- 22.1.5
*-nt 30.2.b
*-n tos 18.6
*o 7, 7.3, 7.4.4, 11.9, 30.1
*o+ei 19.1.7
*o+es 16.1.2, 19.1.15
*-o-h
1
19.1.10
*-oh
2
30.1.a
*-oi 19.1.9
*-o-ih
1
19.1.11
*-oi-m 19.5.2
* 7, 7.4
*-i 19.5.8
*s 19.1.15
*h
3
eteh
3
11.21.1, 21.1
*-ons 19.1.16
*-osi o 7.10.2, 19.1.5
*-ou -es 19.5.32
*p 11.10, 11.18
*pek
u
-i e/o- 11.9
*pk
u
11.9
*penk
u
to- 11.5
*penk
u
e-dm t- 21.1
*penth
1
- 18.1.1: fn. 3.2
*ponth
1
- 18.1.1: fn. 3.2
*ph
2
tr- 11.10
*pu- 11.21.5
*pl th
2
- 11.2.1, 11.17.3
*prh
2
s 21.2
*pr-se/o- 11.20.4
*reh
1
i- 10.1.3
*r 11.2.1
*rH 11.2.4
*-ro 30.2.a, 30.2.b
*roteh
2
- 11.17.3
*rth
2
o- 11.17.3
*s 11.19, 11.20, 11.20.5, 11.26.1
*seh
2
-ul-/*sh
2
-u en- 18.6: fn.
3.4
*sek
u
11.16
*sek
u
h
2
-oi - 11.16
*sem- 21.3
*s 11.20.4
*-se/o- 26.16
*sm -heslo- 21.1
*smih
2
-heslih
2
21.1
*s, *td, *seh
2
22.2.3
*-soi 30.2.a
*-som 22.2.1
*sr- 11.26.3
*-tero- 20.1.2
*t 11.10, 11.17
*teton- 11.21.6
*tH 11.17.3
*-t- 33.2
*tisres 21.1
*t 11.21.6
*-t- 33.3
*-toi 30.2.a
*tri(h
2
)-domt- 21.1
*trins, tri-h
2
21.1
*-tt- 11.11.2, 11.20.2
*u 6
*u 10.11.2
*u eh 7.8.1, 11.25.3
word index 153
*u -si 11.20.5, 11.21.4
*u ei d 7.8.2
*-u ent-/u n t- 18.3
*-u er/-u en 18.6
*-uh
1
19.5.25
*u l h
1
-t 11.2.4
*u obhso- 11.25.2
*-u s/-u os/-us 18.8
* 7
*z 11.25.1
Proto-Indo-Iranian
*-a- 7, 7.3
*ataH-ti- 21.1
*-ai- 7.11.2, 10, 10.1.1, 10.2
*-ai 19.1.9, 19.1.11, 19.2.3
*ai am 22.2.1
*ai -au 19.1.13
*-aibhi as 20.2
*-ai 19.5.4, 19.5.5
*-ai u 20.2
*aiu am 21.1
*-ans 18.3
*-as 7.9.4, 7.13.1
*-asi a 19.1.5
*-ats 18.3
*-au- 7.12.2, 10.3, 10.4
*-au 7.13.4, 10.3.4, 15.4, 19.1.13,
19.5.19, 19.5.26
*au gdha 11.11.1
*-au i- 11.1.4
*-au 10.4, 19.5.22
*-- 7, 7.4, 7.4.4
*-i - 10.5.1
*-i a- 7.3.2, 10.5.2
*-i am 19.5.2
*-m 19.1.18
*-n- 7.6
*-s 8.5.2, 15.4, 16.1, 19.1.15,
19.2.11
*st 11.10.1
*-u - 10.6.1
*-u a 10.6.2
*-bh- 11.18
*-bhias 19.1.19
*tasras 21.1
*C-m Hna- 30.2.e
*-- 11.20.3, 11.21.1, 11.21.2
*rai H-i as- 11.1.3
*-s- 11.21.4
*-u - 11.10.3
*-DDh- 11.11.1
*du i- 11.14.1
*du itii a- 21.2
*-DZh- 11.11.1
*-dhi 29.1
*-H 19.1.8
*-(H)sa- 25.17
*-i- 7, 7.14.1
*-i as- 20.1.1
*i.Ha- 11.28
**-iHs 19.3.10
*-ita- 20.1.1
*-- 7, 7.14.2, 7.15
*- 19.5.9
*j - 11.21.2, 11.24.1, 11.25.3
*-j +s- 11.25.5
*-j n- 11.16
*-j h- 11.24.1, 11.25.3
*-j h+s- 11.25.5
*ktur()i a 21.2
*-l- 11.2.1
*-m- 21.2
*mas-dhaH- 11.24.2
*matsi a- 11.11.2
*-m-m 18.1.1
*-n 11.10.1
*-ns- 7.9.2, 11.27
*-nt 11.10.1
*pntaHs 18.1.1: fn. 3.2
*pn tHs 18.1.1: fn. 3.2
*prHu - 21.2
*-r- 11.2.1
*rta
u an- 18.4
*-s- 30.1
*snts 11.11.2
*-s- 11.20.4
*-st 11.10.1
*suHar 18.6
*-tama- 20.1.2
*-tara- 20.1.2
*tu ar 7.13.3
*-u- 7, 7.17.2
*-u aH 19.5.25
1*u ai d 7.8.2
2*u ai d 7.8.2
*-u as 19.5.22
*u asu-H 19.5.34
*u nst 7.6, 11.10.1, 11.17.4
*-uH 19.5.27
*-uHas 19.4.8
*uma- 22.1.4
*-- 7, 7.17.2
*-z- 11.25.1
*-hasra- 21.1
Sanskrit
m sayo 19.1.12
m has- 7.6.1
agru
- 19.4
aguh- 11.12
-atur 30.1.d
dyu- 7.16.6
addha
11.24.3
ti 23.1
ntama- 20.1.2
antr 11.3.2, 23.8
ana
, anay 22.2.4
nu 23.7
any- 7.9.1, 11.1.3
anym 7.14.3: fn. 2.6
anysmau, any, anym
20.2
p-
aps 18.1.5, 18.1.15
apa
18.1.8
apa 18.1.10
apas 18.1.14
apa
m 18.1.17
adbhys 18.1.18
pa 23.5
apca 18.2
pi 23.2
pnas- 11.18
pnasvant- 11.18
154 word index
abh 10.3.2, 11.1.4, 11.14, 23.3
-bhi 19.2.14
-bhym 19.2.10
-bhyas 19.2.13
amr ta- 11.2.3, 39.3
ay
iya
t 11.1.3
tu 10.3.1
aym
ayh, asys, asyi, ena
, asya
m,
aya
22.2.1
asmt 7.9.2
ebhys 8.1
-ay 19.2.7
-aye 11.1.1
-yo 19.2.9
armati 7.4.2
ardh- 21.4
aryamn- 8.1
rvant- 7.8.3
vidvm s- 7.8.2
vitti- 7.15.1
av 22.2.5
at- 21.1
man- 11.20.3
va- 11.10.3, 11.20.3
ave 19.2.3
aa
11.21.1
aam- 21.2
1as
mi 11.26.2
si 11.26.1
ti 11.20.1, 25.1.1
snti 7.8.1, 25.1.1
sat 11.27
a
st 11.10.1
sr 11.27
edhi 11.24.2
syt 11.28
2as
sya 11.28
-as 7.9.4
su- 11.1.4, 11.26.1
sura- 7.8.1, 11.26.1, 25.1.1
asu 22.2.5
asthanvnt- 11.17.4
asm-
asma
n 11.26.2
asma
kam, asmbhyam 22.1.3,

22.4
asra- 11.26.3
ahm 22.1.1
hi- 11.25.6
a
23.9
a
krti 8.1
-na- 30.2.e
-nm 7.3.1
-ni 18.4
-m 30.2.c
-yai 19.2.6
-ys 19.2.4
-yo 19.1.12
a
viiya- 17.15.1
s-
ss 18.1.5
sa
18.1.7
-s 7.5.2, 16.1.2, 19.2.11
-sas 16.1.1, 19.1.15
skra- 11.20.1
-i 30.2.b
t 22.1.6
idhm- 10.1.1
i- 33.3
yati 10.1.1
- 19.3.1
-nm 19.5.16
22.1.6
ukth- 11.17.2
ukthebhyas 19.1.19
ukn- 7.16.1
ugr- 11.12
ttara- 20.1.2
uttn- 11.20.2
d 23.12
-nm 19.5.35
pa 23.10
upri 23.11
ubdh- 11.11.1
ubh- 21.1
ubh 19.1.11, 19.2.8, 21.1
ubha
21.1
ubhyo 19.1.13, 21.1
ubha
bhym 19.1.14, 21.1

-ur 18.5, 30.1.b, 30.1.d
ur- 11.2.1
urvi
19.5.28
ra- 11.17.4
-e 19.2.3, 19.2.8
ka- 21.1
kasmin, kasys 20.2
-ethm 30.2.b
ethe 30.2.a
ev(m) 21.1
e
ea- 10.1.1
ait 10.5.1
e, ea
, eta
s, etsyas 22.2.3
-ai 30.2.c
jas- 10.3.1
oati 10.3.1
-o 7.5.3, 15.4
-au 19.1.10
rj- 9.1, 11.2.1
rta
van- 7.3.3, 10.6.2, 18.4

kt 11.9, 38.1
katam- 20.1.2
katar- 20.1.2
kanya
7.8.3, 8.1, 18.4

kamnnar- 5.4: fn. 2.1
kar 11.9
krnti 26.5
-kar 7.13.3
krt- 11.2.1
krman-
krm 18.4
krm 18.4
krm 18.4
kav- 7.8.2
kavyas 7.4.2
ks 11.20.1
ksya 11.26.1
ka
ma- 11.9
k-, k- 22.5
k-cit 22.5
k, km 22.5.1
word index 155
ksya, kna 22.5.2
krntti 26.9
krtu- 7.17,6, 11.16
krtu 19.5.18
krto 19.5.22
krtvas 19.5.22
krtve 19.5.24
krtv 19.5.25
krtau 10.6.1, 19.5.26
kv 11.1.3
kam-
ka
m 18.1.2
jms 18.1.5
jmay-a
18.1.9
kmi 18.1.9
ka
mas 18.1.14
1kay
ati 7.8.3
2kay
kti 11.21.6
kar 11.25.5: fn. 2.13
ktra- 11.21.6
gam 11.12.2
g-cha- 11.12.2, 11.20.4
gt- 7.17.5
jagmi- 30.1.e
gat- 33.3
gya- 10.1.2, 19.1.3
gv-
ga
m 19.2.11: fn. 3.5

gu 10.6.1, 11.12, 19.5.18
garh 11.25.3
grbhna
ti 26.5
grbhy- 11.1.4
-gy- 10.5.2
ghar
jghrkati 26.17
gharm- 11.12
cakr- 11.9
cak 18.4
catras, ctasras 21.1
catva
ras 7.3.1, 11.17.1, 21.1

-ca 11.9, 38.2
cit 22.5.1
citr- 7.9.4
cyautn-
chy 7.3.2
jagm- 30.1.e
jgh- 11.12
jani
ja
yate 26.11.b
janiyti 26.18
jt- 7.4, 11.24.1
jna- 11.25.4, 19.5.39
ja
nun 19.5.28
jihva
- 19.4
jihva
/ jihvy 19.2.7
ju- 11.20.5
juhu
- 19.4
juhu
bhi 19.4.11
ja- 11.24.1
j 11.16
jtr- 11.25.4
jya
- 11.1.3, 11.12.2
jryas- 11.24.1
tak
tka- 27.2.a
tkan- 11.21.6, 18.4
tanu
- 19.4
tanu
19.4.1
tanvm 7.17.5, 19.4.2
tanva
tanu
nm 19.4.9
tanu
bhyas 19.4.10
-tam 30.1.b
-tara 20.1.2
tvi- 7.8.2, 7.15.1
-tas 30.1
-tm 30.1.b
30.2.c
-tt 30.1.c
ta
n 7.9.3
tpya- 26.15
-ti 21.1
tirs 7.9.4, 23.13
tudti 26.7
turi
ya- 21.2
trti
ya- 21.2
tyjas- 7.10.1, 8.1
tri-, tisr 21.1
tryas 11.17.1, 21.1
tri
n, tri
, tisrs, tribh,
tribhys, tra
m 21.1
tv- 22.4
tvm 7.17.5
tva
m 7.6.2, 11.17.1
t 7.11.3, 22.1.5
tv 22.15
tvar- 7.13.3, 9.1, 11.14.4
-thas 30.1
dam siha- 7.6.1
dabh
dpsati 11.25.2, 26.17
dar
am 7.4.3
am 9.1
dr- 11.21.1
darh 11.25.5
da 11.13, 11.20.3
daam- 21.2
dasm- 11.26.2
dsyu- 7.17.1
dsyum 19.5.20
dasr- 11.26.3
dah
ati 11.25.6
d 11.13
ddmi 7.6.3, 26.2.a
ddan 11.10.1
dta
ras 7.3.1
dta
ram 7.4.6
da
ru- 8.2, 19.5.39

dr 19.5.40
durvacas- 11.25.1
du 11.25.1
drgh- 11.2.4, 11.12
di
vyati 26.11.a
deva
19.1.10
devi
- 19.2, 19.3.1
dvi 19.3.3
devya
s 19.3.4
devya
s 19.3.4
devya
19.3.7
devyi 19.3.6
devya
m 19.3.8
156 word index
devys 19.3.9
devi
19.3.1, 19.3.9
devi
19.3.10
devi
nm 19.3.11
devi
bhyas 19.3.12
devi
bhi 19.3.13
devi
bhym 19.3.9
devi
m 19.3.2
devu 19.3.14
dra
t 19.1.6
dyv-
dya
m 19.2.11: fn. 3.5

draps- 11.21.5
drhvan- 5.4
dro 19.5.40
droh
drhyati 11.25.6, 26.11.a
dva
, dva
bhym 21.1
dvita
8.1
dviti
ya- 11.14.1, 21.2

dvas- 11.14.1
dvis 11.14.1
dvi 8.1, 11.14.1
dhnvan- 18.6
dhar
dhryati 11.13
dhartr- 9.1
dh 11.13
ddhmi 7.6.3, 26.2.a
dhatt 11.11.2
-dhi/hi 30.1.c
-dhyai 33.1
dhruv- 11.1.3
-dhvam 30.2.c
npt- 11.10.3
nptar- 11.10.3
nmas- 7.8.3, 11.3.1, 11.28, 18.7
nar-
nram 18.1.2
nre 10.2.2
nars 18.1.5
nar 18.1.7
nr 18.1.12
nras 18.1.14
nr n 18.1.15
nara
m 18.1.17
nr bhyas 11.1.4, 18.1.18
navam- 10.3.3, 21.2
nvna
m 21.1
1na 11.21.1
2na
anti 11.21.2
nas 22.1.5
nahyte 26.11.b
na
bha-ndiha- 11.18
na
man- 7.6.2, 11.31

nvj- 7.3.3
n 23.20
ni, nir 23.1
nej
nenikt 11.25.6
ndiha- 11.24.2
ndyas- 11.13
nma- 21.4
-ntm 30.2.c
nycam 18.2
pakth- 21.2
pac
ata 11.9
pca 11.3.2
pacna
m, pac-t 21.1
pnth-
pnthm 18.1.2
pnths, paths 18.1.1: fn. 3.2
pars 7.9.4
pri 23.15
pa- 7.12.1, 11.21.5
paca
23.18
pav 19.5.33
pm s- 7.6.1
pa
dam 18.1.2
pa
i 11.20.5
pitr- 11.10
ram 7.4.4
tr 9.1, 10.2.2
pit-7.14.1
pitrvya- 8.2
pbati 26.8
pi
vas- 7.14.2
prandhi- 11.3.2
pur- 7.12.1, 8.2
purs, 23.17
pura
23.16
pyati 7.16.2
pu
rva 21.2
pe 11.17.4, 11.20.5
pr tan- 11.2.3
prtans 19.2.11
prth- 11.2.1, 11.17.3
pr 11.18, 23.19
prti 23.14
pratham- 21.2
pra
7.6.1
*pra
k- 7.6.1
prac
pr chte 11.20.4, 26.16
pran- 11.21.2
priy- 11.1.3
priym 7.14.3: fn. 2.6
priya
s / priya
sas 19.2.11
bandh 11.3.2
bhu
-
bhu
19.5.27
bhv 19.5.27
budhn- 11.13
bravi
ti 11.31
am 10.3.3
bhakt- 11.16, 33.3
bhga- 11.12
bhaj 11.25.6
bhar
mai 30.1.a
ati 8.1
an 31
an 11.10.1, 11.4
bhavi 7.8.3
am 10.3.3
ati 26.6
bhvemahi 10.1.1: fn. 2.7
bhvema, ta 10.1.1: fn. 2.7
bhyr 30.1.b
bhiaj- 26.13
-bhi 16.7
-bhym 15.3
word index 157
-bhys 16.6
bhri 19.5.3
bhra
trvya- 8.2
m-
ma
m, md 22.1.1
mhyam 22.1.1, 22.1.3
m, me
maku
7.12, 11.24.2, 7.12.1,

11.21.4
majjn- 11.12, 11.24.2
mtsya- 11.7, 11.20.2
mdhu- 7.12.1
mdhv 19.5.25
mdhun 19.5.25
mdh 19.5.27
man
mnya- 13.1, 26.10
-mna- 30.2.e
mnas- 7.8.3, 7.9.3
i 11.26.1
mntra- 76.1
many- 7.8.3, 10.3.4
mrka- 11.2.2
mark- 11.2.2
mrtya- 11.2.3
mrtyeu 19.1.21
maha
ntam 7.5.1
ma
39.2
ma
na- 9.1, 11.13

msya- 11.28
ms-
ma
s 18.1.1
ma
sam 7.5.1, 18.1.2

msya- 11.28
mitra
vu 35.3
mh- 11.20.5
medha
- 11.24.2
mrty- 81, 11.17.1
mriyte 26.12
y-, yt, ysya, ysys, y, ya
n,
ya
s ca 22.3
yac cid 38.3.b
yaj- 7.10.1, 11.21.1
yajya- 7.10.1
yt 38.3.a
ytra 38.3.d
yth 38.3.a
yd 11.10.2
yada
38.3.d
ydi 38.3.c
yam- 7.14.3
ya
vat 38.3.d
yukt- 7.17.3
yuga
16.1.1
yuvm, yuva
m, yuva
ku, yuv
22.1.3
yym, yumbhyam 22.1.4
yumd 11.10.2
r/lghyas- 11.3.2
ragh-7.8.2, 11.12
rtha- 11.2.1, 11.17.3
ratheha
- 10.2.2: fn. 2.9, 40.2

ray- 7.3.2, 10.1.2
raym 19.5.2
-re 30.2.a
reh 11.2.1
rhmi 11.2.1
rrihat 11.2.1
rocas- 10.3.1
rcate 11.2.1
rikt- 8.1
lhmi 11.2.1
vdhar- 9.1
vm sat 7.9.3
vaktr- 11.17.2
vak
vakya- 26.15
vakyti 26.18
aukat 11.20.5
vac 11.20.5
avci 25.2, 32.1.c
uvktha 11.17.2
vca- 27.2.b
skt- 11.10.2: fn. 2.11
vcas- 7.9.4, 18.7
vaym 10.12
va
mi 11.21.1
vaki 11.20.5
vai 11.21.1
-vas 30.1.a
vas 22.15
vas
te 11.20.1
vsu- 7.12.1
vsu 19.5.21
vsave
vsau 19.5.26
vsnm 11.29: fn. 2.14
vsyas- 11.1.1, 11.28
vsyn 18.7
vah
vhanti 7.8.1, 25.1.1
vkat 11.25.5
va
k- 13.1, 18.1.1
va
cam 18.1.2
vcs 18.1.5, 18.7
vca
18.1.8
vgbh 18.1.19
vgbhyas 18.1.18
va
m 22.1.3
vy- 7.3.2, 10.3.4
v 23.23
vidva
n 18.8
vidva
sam 18.8
vim at- 21.1
v-
v 18.1.1
vam 18.1.2
vis 18.1.5
vi 18.1.7
via
18.1.8
vi 18.1.9
vas 18.1.15
via
m 18.1.17
vibhys 18.1.18
vva- 7.9.3
vvebhyas 19.1.19
vvasmai, vve, vvem
20.2
vivvasu- 5.4: fn. 2.1
vga- 10.2.1
ved
vttha 11.20.2
vitt- 11.20.2
158 word index
vda- 27.2
vohar- 11.25.3
vo 19.5.29
vr ka- 11.2.2
vrjna- 7.9.1
vrat- 8.2
vrdh 8.2
vr ddh- 11.11.2, 11.24.3
vrnt 11.2.4
am sa- 7.9.3, 11.3.2
atm 11.20.3, 19.1.11, 21.1
ay
te 7.8.1, 25.1.1
yna- 7.8.3: fn. 2.3
viha- 7.8.2, 7.15.1
s 11.20.1, 11.20.3
ia- 27.2.a
star- 11.20.3
pre 19.2.8
pti- 7.16.1
rav 11.20.3
ra
vi 25.2, 32.1.c
van-
va
18.1.1
va
nam 18.1.2
nas 18.1.5
ne 18.1.7
va
n18.1.10
va
nas 18.1.14
nm 18.1.17
ryas- 11.1.3
rha- 10.2.2: fn. 2.9
11.16
s, tt, sa
22.2.3
sakr t 21.3
sakman- 11.16
skhy-
skh 11.16, 19.5.1
skhyam 19.5.2
skhy 19.5.9
sac 11.26.1
ate 7.9.1 8.1
ante 7.14.3
sak- 7.14.1, 26.2.a
sacan- 7.9.1
sc 23.24
sattr- 11.17.4
satym 7.14.3: fn. 2.6, 7.15.2
satra
23.26
sna- 27.2.a
sapt 11.26.1
saptat- 21.1
sapttha- 21.2
sm 23.27
srva- 7.17.5, 11.1.3, 11.26.1
sarvtti- 8.2
sah 23.25
shas- 7.9.4
sahsra- 19.1.11, 21.1
sicti 26.9
sm 22.1.6
sukrtu- 11.16
sumy- 10.5.2
suvit- 11.30
skt- 11.10.2: fn. 2.11
srya
yai 19.2.6
skndha- 11.20.1
skambh- 11.4
star- 9.2
sth 11.20.1
sneh
snihyati 11.25.6
sp 18.1.1
sprh 11.20.1
smd 11.26.2, 23.22
smsi 11.26.2
-sva 11.31, 30.2.c
sv- 11.30, 22.4
svsmin, sv, sva
nm 22.4
svapty- 8.1
svpna- 9.1, 11.30
svr- 7.9.1, 9.1, 11.30
svs 18.5
svinn- 11.13
srm- 11.26.3
srtas- 11.26.3
han
ti 11.12.2
han-
ha
18.1.1
hnam 18.12
ghns 18.1.5
ghn 18.1.7
hnas 18.1.14
ghns 18.1.15
harm
i
y- 8.1
hav
hvyati 11.14.2
hsta-
hst 19.1.10
htayo 19.1.13
hstbhy 19.1.14
hstai 19.1.19
hastn- 18.4
h
jahmi 11.24.1
h 38.2
hit- 33.3
him- 11.24.1
hotr- 11.24.1, 18.5
hvar
ate 11.14.2
Umbrian
bum 19.2.11: fn. 3.5
tursa 19.2.3
tutas 19.2.4
vitlaf 19.2.11
Topical Index
ablaut 11.9, 13.2, 14.4, 14.5.1, 17, 18.1, 18.1.5, 18.1.9,
18.1.14 18.2, 18.3, 18.4, 18.4, 18.5
accent 7.14.4, 11.2.2, 11.22, 13.2, 17, 22.4;
displacement 11.2.2, 13.2, 19.1.6
hysterodynamic 17, 19.4, 19.4.3, 19.5.1,
19.5.2, 19.5.7, 19.5.18, 19.5.22, 19.5.24,
19.5.33
proterodynamic 17, 18.1.5, 19.5.4, 19.5.17,
19.5.22, 19.5.33, 25.1.1
agent 23.24, 36.3
agreement 20, 35
Aktionsart 25.3
allomorph 13.2, 13.4, 14.1.1, 30.2.e
analogy fn. 2.2, 11.11.1, 11.11.2, 11.21.2, fn. 2.14, 16.5,
18.2, 18.3, 18.8, 19.1.6, 19.1.17, 19.3.5; fn. 2.3,
3.2
intraparadigmatic 7.4.4, 11.12, 11.27, 18.6,
18.6; fn. 3.2
anaptyxis 7, 8, 9, 11.14.1
archetype 2,4, 4.2.5, 4.2.7, 5.2, 5.2.3, 5.2.4, 7.1,
7.7, 11.1.3, 11.4, 11.23.1
aspect 25.3
assimilation 11.1.4, 19.1.3, 19.1.16
Bartholomae 11, 11.11.1, 11.20.2, 11.24.3
Brugmann 7.4.6, 13.2, 18.4, 18.5, 18.7, 18.8, 19.5.2,
19.5.39, 21.1, 26.15, 28, 32.1.c
causative 27.14, 26.15
(relative) chronology 3, 11.1.4, 18.1.2, 18.8, 19.1.1,
21.3, 22.2.1, 25.1.2, 35.3, 40.1, 40.2
comparison 18.7, 20.1.1, 20.1.1, 21.2, 23.11, 36.2,
37.4.b, 38.3.b
compound(ing)/composition 18.1, 19.5.18,
19.5.39
vowel 5.4, 7.9.4, 7.13.2; fn. 2.1
ending
14.1.2, 14.3, 14.4, 14.8, 16.3.2, 18.1.3, 18.1.4,
18.1.9, 19.1.8, 19.2.1, 19.2.3
blending 15.4.2, 19.1.13
OAv./YAv. dialectal differences 1, 7.2, 11, 11.11.1,
14.5.1, 18.3, 19.1.6, 19.1.7, 22.1.5, 30.1.a
degrees of adjectival comparison 18.7, 20.1.1
comparative 18.7, 20.1.1
superlative 5.4, 7.8.2, 7.15.1, 11.20.3, 11.24.2,
20.1.1, 21.2
devoicing 11.16
dissimilation fn. 2.14
enclitic 7.2, 8.1, 11.3.2, 11.28, 16.1.2, 16.6, 19, 19.1.1,
19.1.6, 19.1.8, 19.1.9, fn. 2.3, 3.8
enlargement 19.1.7
epenthesis 8, 9.1., 9.2.1, 10.5.3, 10.6.3, 11.1.4
fricativization 11.1.3, 11.8, 11.11, 11.14.3, 11.15,
19.5.30, 19.5.37
absence of 11.8, 11.10.3, 11.13, 11.14.3, 11.17.4
Gathas 1.4.1.1, 41, 41.3
gathicism 7.11.1, 11.18, 11.23.1, 14.6, 19.5.22
hypergathicism 11.26.3
Geldner 2; fn. 1.3
heteroclisis 13.3, 18.6; fn. 3.4
Hoffmann 2, 4.1, 5.1
inchoative 26.16
indeclinable words 21.1
infix 25.1.1, 25.1.3, 26.5, 26.9, 32.3; fn. 3.1
Iranian
(Proto-)Iranian 1, 6, 7, 7.14.5, 11, 11.2.2, 11.7,
11.8, 11.11.2, 11.13, 11.15, 11.20.5, 11.26.1
Eastern 4.2.2, 7.3.2, 7.3.3
laryngeal(s) 7.4, 7.14.5, 11.2.4, 11.15, 13.4, 14.7,
16.3.3, 17, 18.4, 19.1.8, 19.2.3, 19.3.3
stems in 18.1
lengthening 7.4.3, 14.7, 16.3.2, 16.5, 18.4, 19.5.18,
19.5.20, 21.1
loss
of d 21.1, 21.2
of 11.2
of i 19.5.1, 19.5.6, 19.5.8
of #k 21.2
of 19.1.3
of N# 18.4, 19.2.11
of r# 18.5
of t 11.10.1
of #u 22.1.4
of aspiration 11.7
of dental 11.11.2
of laryngeal 19.3.3
160 topical index
manuscripts 1, 2, 4.2.7, 4.2.8, 5.2, 5.2.2, 5.2.6,
5.4, 7.7, 11.1.1, 11.4, 11.22, 11.26.3, 38.1
pure, impure fn. 1.2
rediscovered fn. 1.3
Mazdayasnian fn. 1.1
motion-suffix 18.8, 19.3, 19.4, 20, 30.1.e
nasalization 7.7, 7.9.3, 11.3, 19.1.16, 19.5.14,
19.5.33
nom. pro voc. 19.3.3
nom.pl. pro acc.pl. 18.1.14, 19.1.16
participle 18.3, 18.8, 19.3, 30.1.e, 30.2.e, 33, 33.3,
39.3
passive 23.24, 25.2, 26.11.b, 32.1.c, 30.2.b, 30.2.c,
33.3
Persian
Old 1, 4.2.4, 6, 11.1.3, 21.4, 31, 37.2
Middle 2, 11.2.3
postposition 11.31, 14.5.1, 16.8, 18.1.9, 19.1.7,
19.1.9, 19.1.21, 19.2.15, 19.3.14, 19.5.26
preverbs 7.14.3, 7.15.5, 11.13, 11.14.3, 11.17, 23,
25.1.2, 31
prohibition 37.2; fn. 4.1
Prolegomena 2
(liturgical) pronunciation 4.1.2, 4.2.4, 4.2.6, 7.1,
9, 10; fn. 2.13
prothesis 21.1, 22.1.4
punctuation 5.4, 38.1
root 13.1, 17, 18.1, 18.4, 20.1.1, 25.1, 25.1.1, 25.1.2,
25.1.3, 28, 29, 32.1.c, 33.1
recharacterization 16.1.1, 19.1.15, 22.2.1
reduplication 11.13, 26, 26.2.a, 26.2.b, 26.8,
26.12, 26.17.b, 27, 27.2.b, 28, 32.2
Rckverwandlung fn. 2.4
ruki 11.20.5, 11.25.1, 14.1.1, 22.1.6
shortening 7.2, 7.3.1, 19.1.6, 19.1.18, 19.3.11, 19.5.2,
19.5.13, 19.5.16, 21.1, 22
Stang fn. 3.5
stem 5.4, 13.1, 13.3, 13.4, 17, 25.1.3, 25.3, 30, 31,
33.1
suppletion 22, 22.1, 22.2.1, 22.2.3, 22.2.4, 22.2.5
svarabhakti 9
tmesis 25.1.2
transcription of Avestan 5.1; fn. 2.11
vocalization
of laryngeal 7.14.5
voicing 11.11.1, 11.24.2, 11.26.1, 16.2
Vulgate 4.2.6; fn. 2.12
YAv. features in OAv. 4.1.3, 7.8.3, 7.9.4, 7.13.1,
10.1.2, 10.2.2, 10.2.3, 10.4, 14.6, 18.1.7, 19.1.9; fn.
2.1, fn. 2.6
Zarathustra 1, 4.1.1, 41, 41.3

Martínez de Vaan (2014) Introduction To Aves Tan

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Martínez de Vaan (2014) Introduction To Aves Tan

Uploaded by

Copyright:

Available Formats

Introduction to Avestan

hypothetically expected form

conjecture not present in

and s ), emerged. 7. Since 1288ce, the recent manuscripts

free; this letter then must have been used

who tells the truth (Y 31.12), fromthe adv. r right, cor-

ras); cf. also the ending of

reflects 1. an preceding h (< *-sa

-; cf. 5.2.3) or (i.e. NC; cf. 11.3.2);

hm(gen.pl.f. of the rel.pron. ya-; cf. Skt. ya

h (2.sg.aor.mid.sbj. of d give / put <

t (3.pl.aor.mid.sbj. of d); YAv. maza

#, which derives from the IIr. sequence

(gen.sg.f. of dem.pron. ata- this); OAv. dana

< *-s (cf. Skt. -o; the loc.du.

is maintained by analogy: gen.sg. (mazd-) mazda

sca; nom.pl. (dana-) dana

sca; acc.pl. (ssn-) ssna

m); OAv. mm(acc.sg. 1.pers.pron.;

m); camm (loc.sg. of caman- eye); nmm, nmanm

ma); YAv. dadmi (1.sg.pres.act.ind. of d give

ignorant (nom.sg.; Skt. vidvm s-); PIE *u ei d- > IIr. 1*u ai d-

haiin (3.pl.pres.act.inj. of rah move away from); or (b) before

; Gr. ?) OAv. spaniia

, YAv. spainiiahm (nom.sg.

, YAv. has sometimes restored original am and mostly restored original an

; cf. also fn. 6 below),

n); ams g (acc.pl.m of ams a- immortal); vsp g (acc.pl.

(i.e., in a palatal environment): cf. iiiejah- aban-

); pouru- much (Skt. pur-, OP paruv); vohu-

or before the sequence

vas-); 3. as the YAv. outcome of

r, influenced by the accent of the voc. p tar, versus the OAv.

, cf. 7.14.3 and fn. 6

. From there, it was introduced

m(acc.sg. of riuua- third); YAv.

: irixta- (adj. from ric

ru-); hauruuatt- wholeness (Skt. sarv-

has a special development in YAv., namely, the u was

tr vya-); tiriia- uncle <*p()turi a-

na-); xvaf na- dream (Skt. svpna-); f ri (dat.sg. of ptar- father;

weapon (neut.; Skt.

sun (stem xvan- neut.; Skt. svr-).

iiaoana- act, deed (v.l. s

iiaona-; cf. further 11.23);

1./3.sg.perf.act.ind. of vid know versus 2.sg. vist; cf.

, which was in turn

-); YAv. pairi.uruuata- he

van-; cf. 11.3.3).

(< Ir. *i ), was employed.

t); YAv. sraiiah- (comparative of srra- beautiful, excellent;

under 11.23.1; Ir. *hi - > x ii- under 11.28;

man-, Lat. nmen); nmah-

five (Skt. pca, Gr. ); YAv. rj-ita- (superl.

ma-, cf. Lat.

), developed to Ir. * > Av. c:

, Skt. ca, Lat. que, Gr. (Mycenaean -qe) < PIE *k

) and the PIE resonants *n , *m , *r , *l ), a single variant of the

d); ymat (abl.pl.

< *dad-tai ). These forms show the way in

ten, Skt. da < PIE

ya-. 2. The Ir. cluster *ju (< PIE *

e), Skt. skh (dat.sg.

ignorant (nom.sg.; Skt. vidvm s-); PIE u ei d- > IIr. 1u ai d-

) and the PIE resonants n , m , r , l ), a single variant of the

ya-. 2. The Ir. cluster ju (< PIE

- < IIr. mas-dhaH- < PIE mn s-dheh

loc. -i -au -Su, -hu

< IIr. -s < PIE -eh

2 (< --s < -aH-s; 7.5.2); r-j, ratu-fr. Root