Caja PDF

Comparta fácilmente sus documentos PDF con sus contactos, la web y las redes sociales.

Compartir un archivo PDF Gestor de archivos Caja de instrumento Buscar Ayuda Contáctenos

Robert Steven et al Etymological Dictionary of Greek (vols. 1 & 2) Indo european .pdf

Nombre del archivo original: Robert Steven et al Etymological Dictionary of Greek (vols. 1 & 2) Indo european.pdf
Autor: aa

Este documento en formato PDF 1.6 fue generado por PDFsharp 1.2.1269-g ( / Adobe Acrobat 9.0 Paper Capture Plug-in with ClearScan, y fue enviado en el 08/02/2018 a las 01:31, desde la dirección IP 189.166.x.x. La página de descarga de documentos ha sido vista 927 veces.
Tamaño del archivo: 24.6 MB (930 páginas).
Privacidad: archivo público

Descargar el documento PDF

Vista previa del documento

Leiden Indo-European

Etymological Dictionary of Greek

Etymological Dictionary Series
Edited by

Robert Beekes

Alexander Lubotsky
With the assistance of

Lucien van Beek







This publication has been made possible by the financial support of the Netherlands
Organization for Scientific Research (NWO).
This book is printed on acid-free paper.
Library of Congress Cataloging-in-Publication Data


Beekes, R. S. P. (Robert Stephen Paul)
Etymological dictionary of Greek / by Robert Beekes ; with the assistance of
Lucien van Beek.


p. cm. - (Leiden Indo-European etymological dictionary series; v. 1011-2)
Includes bibliographical references and index.
ISBN 978-90-04-17418-4 (hardback: alk. paper) 1. Greek language- Etymology­




Dictionaries. 1. Beek, Lucien van. n. Title.

Pre-Greek loanwords in Greek

PA422.B44 201O




Abbreviations and symbols............................................................................................... xlv
The Greek etymological dictionary A-A


.................... .......................................................



The Greek etymological dictionary M-O'

ISSN: 1574-3586
ISBN Set: 978 90 04 17418 4
ISBN Volume One: 978 90 04 17420 7
Copyright 2009 by Koninklijke Brill NV, Leiden, The Netherlands.
Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing,
mc Publishers, Martinus Nijhoff Publishers and VSP.
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.





......................................... . ......... ......... .......


. .

. .

...................................................................................... ... . ... . ...................



.................... .............................................................................................. ............




Whoever takes up the task of writing a new etymological dictionary of Greek, has to
depart from the existing dictionaries. The present dictionary, too, owes a great deal
to previous work in the field, especially to the excellent dictionaries of Hjalmar Frisk
and Pierre Chantraine.
Apart from compiling the first comprehensive etymological dictionary of Greek
in the English language and incorporating the most recent scholarly literature on
Greek etymology, there were a number of other reasons why a new dictionary
seemed to be a desideratum. In the preface to his dictionary, Frisk expressed doubts
on three points: 1. the laryngeal theory; 2. Mycenaean; and 3. the Pelasgian theory on
the Greek substrate language. Ironically, it is precisely on these three points that
substantial progress has been made in the last decades, so that we can now be much
more confident in these areas.
1. Frisk felt uneasy about the laryngeals. In the preface (p. vi) he wrote: "Fur die
griechische Etymologie fallt sowieso die Laryngaltheorie (... ) nicht schwer ins
Gewicht". I have been acquainted with the problems of the laryngeal theory since the
start of my academic career (see my dissertation, Beekes 1969), and I vividly
remember how the chaotic spectrum of theories and hypotheses discouraged many
people in the beginning.
Since the 1980'S, the situation has changed dramatically. When Bammesberger's
Die Laryngaltheorie appeared (Bammesberger (ed.) 1988), there had already been
general consensus on the main rules of development of the laryngeals in Greek and
in other Indo-European languages. It is absolutely clear now that the development of
the laryngeals is essential for understanding Greek etymology. Chantraine's
Dictionnaire etymologique de la langue grecque (DELG) often does not give
reconstructions with laryngeals either; as a consequence, many of the etymologies
still defended in his dictionary are clearly untenable within the framework of the
laryngeal theory. It must be admitted, however, that many of these deficiencies have
been remedied in the Supplement (DELG Supp.), which often contains very helpful
2. The study of Mycenaean has by now become an integral part of Greek studies.
The Mycenaean material was already accepted by Chantraine and incorporated into
DELG. I have tried to include all Mycenaean data with a reasonably certain
interpretation, provided that these data have a bearing on the etymological
interpretation of classical Greek. Personal names are generally excluded from the
discussion, as their interpretation is often too uncertain to base any conclusions on.
The task of incorporating Mycenaean data was not too difficult, since we have the
excellent Diccionario Micenico (1985-1993) by Aura Jorro at our disposal. Although




the Mycenaean material is limited, it is of great importance and should always be
taken ito account. The exact attestations of the Mycenaean words are usually not
cited, as they can easily be traced in Aura Jorro's dictionary.

The rigorous editing of the etymological sections of the dictionary was done by
Lucien van Beek. He integrated my own views with traditional etymologies and
recent insights. In those cases where a word can now be proven to be of Pre-Greek
origin, part of the old reasoning has sometimes been retained in order to illustrate
the flaws in the traditional approach, according to which practically every word is
bound to have an Indo-European etymology.


3. It is now clear that the Pelasgian theory, which started from the assumption
that there was an Indo-European substrate in Greek, has been a completely
unfruitful and wrong approach. Although Frisk doubted this theory, he nevertheless
conSistently referred to Pelasgian throughout the dictionary. This is a pity, because
the theory has yielded no positive results. Chantraine often used the vague terms
'acheen' or 'mediterraneen', without clearly identifying Greek substrate words in this
In the present dictionary, no reference to the Pelasgian theory is made anymore.
Instead, I have extensively used Furnee's 1972 book, who meticulously studied the
substrate material and concluded that we are dealing with loanwords from a single
non-Indo-European language. Unfortunately, this work has been neglected or
rejected by most scholars without due argumentation. In order to explain the
principles of Furnee's work and to present his conclusions, as well as my own
findings from recent years, I have written a special introduction to Pre-Greek (as I
call the substrate language), see pp. xiii-xlii. Throughout the dictionary, much
attention is paid to the Pre-Greek material, and one of my main goals was to
generate a collection of substrate words which would be as complete as possible. I
intend to publish a separate work, containing all certain or probable Pre-Greek
etyma, in the coming years.

The dictionaries of Frisk and Chantraine are different in their orientation. Whereas
Chantraine is more oriented towards the philological study of Greek (as follows
from the subtitle Histoire des mots), Frisk focuses on the Indo-European side of
Greek etymology. In fact, it may be fair to say that Frisk to some extent tried to
produce not an etymological dictionary of Greek only, but of Greek and Indo­
European at the same time. The main focus of the present dictionary is also
etymology, rather than philology.
I started working on the project in 2002. At first, the idea was to produce an
updated English translation of Frisk in the framework of the Indo-European
Etymological Dictionary project. While largely maintaining the philological part of
the entries, I modernized old reconstructions, added new ones from the literature,
and rejected older etymologies in the light of the substrate theory. Furthermore,
many new entries have been incorporated, most of them glosses by Hesychius, which
were gleaned from DELG, from Furnee's book and from the new 2005 edition of
Hesychius (part Ill, II-L:).
Gradually, I have come to the conclusion that a much more rigorous approach
was necessary: there is simply too much irrelevant and dated literature in Frisk's
dictionary, and many of his pre-Iaryngealist reconstructions are now useless. Also,
research interest in Indo-European studies has shifted considerably over the course
of decades. It was therefore decided to completely reorganize the etymological
treatment of the entries.

Structure of the entry

After the lemma, grammatical information is given between square brackets, for
instance, 8UpOflaL [v.] 'to lament, bewail', or £YKUTU [] 'intestines'. If it is
unknown (for instance, in a gloss), this may be indicated with a query.
The grammatical information is followed by the meaning of the word. For most
of the glosses, an English translation has been provided. Although this is a major
break with tradition in Classical Studies, I consider it to be convenient for specialists
in other Indo-European languages than Greek. Of course, in many cases a gloss can
be ambiguous, but I hope to have been suffiCiently prudent in the translations.
At the end of the first paragraph, I give the origin of the word (in abbreviated
form) between two arrowheads. The abbreviations must be understood as follows:
There is a good Indo-European (IE) etymology. The IE root is
reconstructed, and in most cases also the formation represented by the
Greek etymon. If there are no cognates, but the Greek word looks Indo­
European, a reconstruction has sometimes been proposed, too.
An Indo-European etymology exists for the entry concerned, but it is not
entirely convincing.
The word was coined in the more recent (pre)history of Greek, and
consists of one or more (pOSSibly) inherited elements; however, the
formation as a whole was certainly not inherited from IE.
The word certainly belongs to the Pre-Greek substrate language. The
reason for this decision may be indicated with (V), which means that
there are formal variants, or with (S) if the word contains a suffix
characteristic for Pre-Greek.
The word may be Pre-Greek (see above on (V) and (S)).
A loanword. The donor language is indicated in abbreviated form, e.g.
<!{LW Sem.� a loanword from Semitic.
A loanword from (one of) the European substrate language(s). Such
words are not reconstructible for PIE, but share similarities with words
from other European language families (Germanic, Italo-Celtic, Balto­
Slavic) that must be due to substrate influence.
<!{ONOM� An onomatopoeic word.
No good etymology exists, or the etymology is unknown.



The philological information is subdivided into sections in order to make the
presentation more transparent:



Inflectional forms and phonological variants.
Dialectal forms. Mycenaean is mostly given in the (approximate)
phonological transcription.
.COMP Compounds (only the most common or etymologically relevant
compounds are given).
ETYM Etymological discussion .






The Proto-Indo-European reconstructions

The reconstructions in this book follow some conventions which deviate from
common usage. Let me mention the most important ones:
a) PIE had no phoneme *a. Whenever *a appears in a reconstruction, the stage of
language should always be understood as post-PIE.
b) In lE reconstructions, vocalization of resonants and laryngeals is as a rule not
indicated, since the consonantal and vocalic allophones were not phonologized in
the proto-language. Thus, for the PIE pre-form of �a(vw , I write *gWm-ie!o-.
Whenever vocalization is indicated, i.e. *gw1jl-ie!o-, this is understood to be a post­
PIE development.
c) I follow Kortlandt's theory of Balto-Slavic accentuation, and adopted his
reconstruction of (pre-)glottalized consonants for PIE (see, for instance, on £KaTOV
and Tt£VT�KOVTa).
d) It should be noted that the term 'prothetic vowel' is used in this dictionary to
indicate the vowel (mostly a-) that may or may not be present in Pre-Greek substrate
words. In inherited words, a facultative prothetic vowel is not reconstructed any
more since it contradicts the laryngeal theory.
Bibliographical references

Within the limited amount of time available for this project, it proved impossible to
modernize all references and to check all reference works. It was necessary,
therefore, to make certain strategic choices. It was decided to concentrate on the
etymologically relevant publications and to adjust the philological treatment of the
material only sporadically.
The second editions of reference works, such as Lejeune's Phonetique historique
(1972) and Risch's Wortbildung (1974) have been systematically consulted. I have
generally maintained references to Chantraine Formation, as this book contains a
very concise and precise overview of the different suffIxed nominal formations in
In contrast to Frisk's dictionary, references to works on specific morphological
topics have been left out. For instance, for a derivation in -mJvT], Frisk often refers to
Wyss's 1954 book. Other such works, to which the reader can refer, are: Redard 1949


(-LTT]<;), B06hardt 1942 (-£u<;), Fraenkel 1910 (agent nouns), Benveniste 1948 (agent
and action nouns), and, more recently, Leukart 1994 (suffIx -nl<;, -a<;).
Furthermore, references to the dictionaries of individual languages have largely
been omitted. Most references to Walde-Hoffmann (Latin), Vasmer (Russian),
Fraenkel (Lithuanian), etc. are superfluous in a Greek etymological dictionary. It is
understood that the reader who wants to know more about the cognates in a given
branch will find his way to the relevant dictionaries. References to Mayrhofer's
KEWA have been retained in some instances, because it often contains more details
than the EWAia. The LIV2 has proven to be a very important work of reference for
all verbal roots, even if I very often disagree with details of their analysis.
References to Stromberg's Pjlanzennamen and Fischnamen have been maintained,
as well as to Thompson's Glossary of Greek fishes. Unfortunately, it has not been
possible to adjust all references concerning Greek religion to recent works such as
Burkert 1985.
Regarding the epigraphic material, no systematic check has been made of the

A new etymological dictionary of a language like Greek cannot be written in a few
years by just one person, without the help of others. Many people helped me on
various stages of the project.
First of all, I am greatly indebted to Lucien van Beek for editing, correcting and
proofreading the whole volume containing about 7500 entries over the course of
more than two years. Several others assisted him in this work, sacrificing many
weeks of their spare time: Alwin Kloekhorst, Guus Kroonen, Michael Peyrot, Tijmen
Pronk, and especially Michiel de Vaan. Needless to say, it is I who remain
responsible for all views expressed in this dictionary, and for any mistakes in it.
I am very grateful to Alexander Lubotsky, who proofread a large part of the
dictionary, and spent a lot of time and effort in formatting the manuscript. Dr.
Velizar Sadovsky (Vienna) has been so kind to write many macros for generating
indices and bibliography and to proofread some parts of the manuscript. I am
indebted to Dr. Thomas Olander (Copenhagen) for solving various font problems.
I would like to thank the students of our department - Kristen de Joseph, Marijn
van Putten, Simon Mulder and Alain Corbeau - for technical assistance. Kristen de
Joseph further copy-edited the manuscript. Marijn van Putten and Simon Mulder
helped compiling ilie bibliography.


A. Introduction
B. Phonology

1. The phonemic system of Pre.-Greek
2a. Characteristic sounds or sound groups: 1. au; 2. �; 3. �O ; 4. yo; 5. yv; 6. ov; 7. KT; 8. KX; 9 flY; 10. DU;
11. mp; 12. pO; 13. pKV; 14. pv (po, Vo); 15. a; 16. a�; 17. ay ; 18. OK, aT; 19. aT A; 20. Te; 21. <pe; 22. Xfl,
Xv; 23. '1'-; 24· w; 25. geminates
2b. How to recognize words as Pre-Greek?
3. Prothetic vowel
4. s-mobile
5. Consonant variation
5.1. Voiceless I voiced I aspirated stop; 5.2. Prenasalization; 5.3. Nasalization; 5.4. Labial stops I m I 1J
(a.TI, �, <p I fl; b.TI, �, <p I (F); c. fl I (F)); 5.5. Stops interchanging with 0(0) , with stop + oh or with a
+ stop; 5.6. Velar I labial I dental stops: labio-velars; 5.7. Dentals I liquids; 5.8. Simple I geminate; 5.9.
0- I zero; 5.10. K-, T- I zero; 5.11. V-, A- I zero; 5.12. Metathesis, shift of aspiration; 5.13. Secondary
phonetic developments; 5.14. Other variation.
6. Vowel variation
6.1. Single vowels, timbre; 6.2. Long I short; 6.3. Single I diphthong; 6.4. Rising diphthongs?
6.5. Secondary vowels (or syncope).
C. Morphology

1. Reduplication
2. SuffIxes
2.1. Introduction; 2.2. Survey of the suffIxes; 2.3. The material: -a�-(0-), -ay-, -ayy-o-, -ao-, -ae-o-,
-m-IE(L)-, -m(F)- o-, -m�-o-, -me-, -mv-, -mp-Coo), -aK-, -a A(A)- o-, -afl�-o-, -aflv-o-, -afl-o-, -av-o-,
-av-, -avo-, -avop(-o)-, -aveh-, -avv(-o)-, -a�-, -aTI-O-, oap, -ap-, -aa-a/o-, -aaa-o-, -aT-, -aup-a/o-,
-ax-, -a'l'-, -yo-, -yp-, -eo-, -e�-a, -eLP-O-, -eA-a/o-, -eAA-a/o-, -efl-O-, -eflv-(o-), -evv-a, -ep-a/o-,
-eT-O-, -wp-, -WT-, -T]�-a/ o-, -T]e-(o-), -T]K/X-, -T]A- o-, -T]v, -T]v-, -T]P, -T]P-, -T] a(a)-a/ o-, -T]T-(O-), -T]H-,
-T]'I'- o-, -e- o-, -ep - a/ o-, -L�-, -Lyy/K/X-, -Lo-, -Lova, -Le-, -Le-, -LK-, -LK-, -LA-, -LA-, -LAA-a/o-, -Lflv-a/o-,
-Lv-a/o-, -Lv-(o-), -LVO-, -LVe-(O-), -L�-, -LTI-O-, -La-a/o-, -LaK-O-, -LT-a/o-, -LX-, -KV-, -fl-O-, -V-, -�-, -OTI-,
-op-, -oaa-a, -O TT-a, - ouA-O-, - oup -, -oua(a) -a, -TIV-, -TIT-, _po, -py-, -po-, -pv-, -aK-, -0-0, -00-, -aT-,
-aTpov, -T-O-, -T T-, u� , -uyy-, -uo-, -uova, u e , -UL-a, UK , -UK-, -uA-, -ufl-, -ufl�-' -uflv-, -uv-, -UVO-,
-uveh-, -UVV-, -u�-, -UTI-, -up-, -up-, -ua-, -UT-, -uX-, - <pe-, -<p-o-, -WK-, -wA-, -wfl-, -WV-, -WTI-, -wp-,
-0000-) -WT-.




3. Word end
3.1. in vowel (a. -a; b. -L, -L�; C. -u, -U�; d. -W�;
3.3. in -�, -'I' (a. -�; b. -'1'); 3.4. in -v; 3.5. in -a�.


e. -w,


-W�); 3.2. in -p (a. oap;

b. -LP;


-op ;

d. -wp) ;

D. The unity of Pre-Greek

E. Pre-Greek is non-Indo-European

A. Introduction
The substrate language of Greek will be called 'Pre-Greek' in this dictionary; this is a
translation of the German ten;n 'das Vorgriechische'. No written texts exist in this
language, but it is known from a considerable number of loanwords in Greek.





The study of Pre-Greek has had an unfortunate history. In the past century, it was
called 'Pelasgian' and considered a dialect of Indo-European. This idea fascinated
scholars, and research concentrated on this proposal. But the whole idea was clearly
wrong, and by now, it is generally agreed that the substrate was non-Indo-European.
Therefore, the term 'Pelasgian' can no longer be used. Frisk already had strong
doubts about the Pelasgian theory, but nevertheless, he often mentioned the
proposals of its adherents. Since all work following this line has turned out to be
useless, I decided to make no mention of the theory anymore in the dictionary.
When Frisk completed his dictionary in 1972, Furnee's book 'Die wichtigsten
konsonantischen Erscheinungen des Vorgriechischen', which was his dissertation
written under the supervision of F.B.J. Kuiper, had just appeared. It was an
elaboration of Kuiper's 1956 study on Greek substrate words, which opened a new
chapter in the research of the field. Furnee rejected the Pelasgian theory, too (see
especially op. cit. pp. 40-55).
Furnee's book met with fierce criticism and was largely neglected. In my view, this
was a major mistake in Greek scholarship. True, some of his identifications are
improbable, and his repeated claim that certain forms were expressive leads
nowhere. What remains, however, is that he studied a great number of relevant
forms and drew obvious conclusions from them. Pre-Greek words often show a type
of variation which is not found in inherited words. It is self-evident that this
variation must be studied, and this is what Furnee did. It has turned out (as Kuiper
had already shown) that this variation shows certain recurrent patterns and can be
used to recognize Pre-Greek elements.
. Furnee's book is not easy to use: every form is discussed at three or four places,
each time in a different context, so that it may be difficult to find out what his point
really is. On the other hand, his treatment is very careful, and there hardly any
obvious mistakes. I found a number of cases which he had not recognized (e.g.
mwX6c;), but this does not change the fact that his book was the best collection at the
time. Furnee worked on it for twenty years, and even now it is the only hand-book
on the subject. The short overview which follows below is based on Furnee's material
and on my own research of more than thirty years.'
Furnee went astray in two respects. First, he considered almost all variation to be
of an expressive character, which is certainly wrong: it is evident that the variation
found is due to the adaptation of words (or phonemes) of a foreign language to
Greek. We shall see below that many variants can be understood in this way.
Secondly, Furnee was sometimes overzealous in his search for inner-Greek
correspondences. Many of Furnee's discoveries are brilliant (see s.v. 80PUKVLOV for
an example), but sometimes he went too far: not every alternation necessarily points
to Pre-Greek origin. The author can hardly be blamed for his enthusiasm. He was
exploring new ground, and it can only be expected that he sometimes overplayed his

Several scholars were baffled by Furnee's proposals and hence rejected the whole
book altogether. His method, however, was correct and I have only filtered out the
improbable suggestions. In many cases, of course, we cannot be absolutely certain,
but this cannot be an objection. Except for a very small number of cases, Furnee's
material does consist of Pre-Greek words. His index contains 4400 words, and
taking into account that many of these words concern derivatives and variants, as
well as a few Indo-European words, I estimate that Furnee's book discusses some
1000 Pre-Greek etyma.2
In general, I have given only a few personal names and toponyms, and no
material of this kind from outside Greece and Asia Minor. The comparison with
Basque or Caucasian languages has not been considered in this dictionary, as this is
not my competence; it is likely that there are such connections, but this must be left
to other scholars.
My suggested reconstructions are not essential. One may ignore them and just
consider the variation itself. These variants are often explained as incidental
phenomena (assimilation, influence of other words, etc.), and such explanations may
be sometimes correct, but if we know that some variants frequently occur, we will
have to consider Pre-Greek origin. Existing etymological dictionaries often seem to
avoid the conclusion that a word is a substrate element. It is remarkable that
Chantraine was quite aware of the problem in his Formation, but in his dictionary he
often withdrew his earlier evaluation (which in my view was correct). It looks as if
substrate elements were not welcome there.
The relationship with Anatolian languages is a separate problem. A Greek word is
often called a loan from an Anatolian language, while it may just as well be borrowed
from the Pre-Greek substrate. It is generally accepted, on the basis of toponyms, that
there was a language which was once spoken both in Greece and in western Asia
Minor.3 In most cases, however, it is impossible to distinguish between substrate
words and loans from Asia Minor (the latter are from a later date). A word may have
been adopted through commerce, as often happens between two neighboring
countries, or starting from the time when Greeks settled in Asia Minor, probably as
early as the 15th century. From a methodological point of view, I think it is better to
consider such words as Pre-Greek, and to define them as loanwords from an
Anatolian language only when there is reason to do so. Still, it is clear that we may
often make mistakes here. A case in point is TOAUTtTj 'clew, ball of wool ready for
spinning'. The word is clearly related to Luwian and Hitt. taluppali- 'lump, clod'.
The Greek word is typical of Pre-Greek words: the structure CaC-up- (with a
appearing as 0 before u) and the absence of an Indo-European etymology (Melchert
Orpheus 8 (1998): 47-51 is not convincing) imply that the word is Pre-Greek or Pre­
Anatolian. On the other hand, 'clew' is not a word that is easily brought from
overseas; it is an everyday word that the speakers of Greek and Anatolian must have

Since Kuiper was my supervisor as well, I was acquainted with the book from the very beginning (see
my review in Lingua 36, 1975).

, Note that Furnee often adduces' new material that is not mentioned in the current etymological
dictionaries, mostly glosses froin Hesychius.
3 A point for further study is to establish how far to the east such related names can be found. It is my
impression that these names can be found as far south as Cilicia.



picked up not far from home. I completely agree with Furnee's interpretation (3533)
that the word was brought to Greece by settlers from Anatolia who spoke the
language, which, from another perspective, we call Pre-Greek. In other words,
TOAUTIT] is a loan from an Anatolian language, but this (probably non-Indo­
European) language was also spoken in large parts of Greece before the Greeks
(speaking an Indo-European language) arrived there.
It is essential to realize that substrate words are a frequent phenomenon. One may
regret this (for instance, from the Indo-Europeanist point of View), but this is
irrelevant; the existence of Pre-Greek words is simply a fact that has to be accepted.
To me, it is fascinating that in this way we can learn something about the oldest
language of Europe (including Anatolia), of which we otherwise have no evidence.
The 'Pelasgian' theory has done much harm, and it is time to forget it. The latest
attempt was Heubeck's 'Minoisch-Mykenisch' (discussed by Furnee 55-66), where
the material was reduced to some ten words; the theory has by now been tacitly
B. Phonology

The phonemic system of Pre-Greek

Voiceless, voiced and aspirated stops may interchange in Pre-Greek words, without
any apparent conditioning factors. This fact shows that voice and aspiration were
not distinctive features in Pre-Greek.4 On the other hand, the Linear B signs
(graphemes) for rjo, rja and tja show that palatalization probably was distinctive.
This is confirmed by the sign pte (e.g. in ra-pte-re Jhrapteresl with the agent suffix
-ter-), which must go back to an earlier pe. In the Pre-Greek material, such a
phoneme may underlie examples like 8CtTITU. One may wonder whether Kpoaa6cp80v
points to p > pt, which was realized with aspiration. Further, the signs two, twe, dwo,
dwe, nwa, swa, swi, point to labialization as a distinctive feature, i.e. tWo, tWe, dWo, dWe,
nWa, sWa, sWi. Note that palatal and labial forms of graphemes are found both with
resonants and stops, which is a phenomenon alien to Indo-European languages. The
existence of labiovelars is confirmed by qa-si-re-u
�aO"lAEUC;, etc. (see further
Beekes Glotta 73 (1995/6): 1 2f.) We may thus posit the follOWing systems:




Of course, it could be due to the fact that a different distinction was present in Pre-Greek (like fortis /
lenis, found in most Anatolian languages), but no obvious distribution pointing in this direction can be
discerned in the material.
5 Note that I distinguish between palatals of Pre-Greek origin, which are indicated by a superscript y
(e.g. k>,), and palatovelars ofIndo-European origin.



Of course, it is possible that one or more of the posited phonemes did not occur in
Pre-Greek (e.g., mY is a rare sound in the languages of the world).
We can now use this insight in explaining the surfacing Greek forms. Thus,
McpvT] 1 l5auxv(u)- can now be explained from a Pre-Greek form *dakwn-.6 In the
former form, the labiovelar yields a labial stop cp. In the latter, it is rendered by -uX-,
with anticipation of the labial feature, while the labiovelar turns up as a velar, possib­
ly by dissimilation from uklV• Again, note that aspiration is not phonemiC in Pre­
Greek. It is very important to note that we cannot predict how a Pre-Greek form will
surface in Greek: sometimes a stop turns up as an aspirate, sometimes as a voiced
stop (e.g. aiTIuc; 1 ucpap, see B 5.1. below). As a consequence, it may happen that there
is a large number of variants, but it may also be that there are no variants at all.
As a second example, we may also understand aux�v 1 Lesb. uflCPT]v from a pre­
form *ankwen. The latter form is directly understandable, with cp from the labiovelar.
The first form went through *anwken or *awnken, giving aux�v with loss of the nasal
(a development known from Armenian). Perhaps, a scenario *akwen > aux�v is also
possible, with a prenasalized form *ankwen (> uflCPT]v) beside *aklVen.7 Such
interpretations may be wrong in individual cases, but this is no reason not to try. On
the other hand, variation that is strange from an exclusively Indo-European point of
view becomes understandable in this way, starting as we do from a limited set of
The existence of palatalized phonemes in Pre-Greek may explain a number of
other developments. Thus, I assume that a geminate AA may continue Pre-Greek *1>'.
We know that lE *ly gave AA in Greek, but if a variant with single A coexists, we are
warned. For example, the name A.XLAAEUC; has a variant A.XLAEUC; with one A. And
although the latter only occurs in Homer, this fact points to Pre-Greek origin. The
variant was preserved because it was metrically convenient, it was not created for
metrical purposes. Of course, the fact that there was more variation at an earlier date
is what we expect. As far as the other palatalized resonants are concerned, anY may
have given atv, arY may have given alp (or also ELp with coloring of the vowel, see
section C2 below on the suffIxes), etc. We have -aLp-, -aLV- but no *-aLA- in Pre­
Greek words. This is confirmed by the fact that geminate AA is very frequent (Fur.
387), whereas geminate pp, vv and flfl are much less frequent, or even rare.
In a similar fashion, *asY may have yielded either -aLa- or -ua-, cf. KCt�aLaOC;,
which has a v.l. KCt�aaoc;. In rendering such a foreign word, the palatalization may
have been represented at one time, and may have been neglected at another. This
6 Although I assume that voice was not distinctive in Pre-Greek, I do write d- in this case, because only
8- surfaces in Greek. We must avoid losing information present in the Greek forms. Thus, my notation of

Pre-Greek forms is heuristic to a certain degree, and not always consistent with the phonemic system I
tentatively reconstruct here.
7 On prenasalization, see B5.2. below. As an alternative, an Indo-European etymology starting with the
root *h,emt- 'to tie, betroth', can be offered; see the dictionary (althoughI prefer the analysis given here).




phenomenon was the main cause of variation in Pre-Greek forms. The
interpretation is further confirmed by the parallel development of labialized
consonants. Thus, I suppose that arW resulted in -a(u)p- (see the section on the
suffIxes). In this way, we may understand KaAaupo'\l beside KOAOpO�OV from a pre­
form kalarw-op-. Another form which shows the remarkable interchange a/au is
apaoxuOe<; I aupoaxu<;. Here one might assume a pre-form *arwask-at-. Note that the
labial element would at the same time explain the 0 as a variant of a in both cases. A
similar mechanism must be at the basis of the etymon aAo�, aiJAa�, dlAa�, EUAaKCt,
which is hopeless from an Indo-European point of view. I assume that all forms go
back on Pre-Greek *alw-ak-. It gives aUAaK- through anticipation, aAoK- through
coloring. In this way, the first two forms, which are best attested, are directly clear.
Further, au/wlw interchange frequently, which explains dlAa� and EUAuKa; OAOK- is
not problematic either, as both la/'s were colored to [0] by the labialized resonant.
Only the Homeric accusative dlha is hopeless: it is the only form that has no vowel
between A and K, and therefore may be due to some accident of the tradition. This is
a problem that has not been solved yet.
I do not know whether a diphthong is allowed in suffrxes of the structure VC, cf.
the forms in -aLFo<;. Structurally, one could think of _ayw_, or even -awY-, but such
sounds are rather rare in the languages of the world. An instance of -aL- due to a
palatalized consonant is e�alcpvT)<; I e�arclvT)<; I acpvw (a brilliant combination by Fur.
158, etc.), which must contain -ap- (the palatalization was ignored in the last form).
Comparable to the development in e�arclvT)<; is KVW'\I I KIVWrc£TOV, from kln- with I
representing palatalization, cf. Beekes 2008. Likewise, I assume that mvuTo<; beside
rcVUTO<; points to *pnut-. Perhaps, we must interpret atwrcuw as *sYop- because of
Euawrcla. An interesting case is Alfllv9E<;· £AflIV9E<;, for which I assume *JYm- beside
*alYm- with prothetic a (see B3 below on the prothetic vowel).
A palatalized consonant could color a to e. A good example is Kl>TCapo<;, KUrcaLP0<;,
but also KurcEpO<;, Kl>TC£LpO<;, where we have all possible variants due to the palatalized
consonant. Compare further Ku�apvol next to KU�£LpOL. Likewise, we have (aKEhl<;
next to (EKEhl<; 'KOAOKUVTaL', where the interchange occurs after ( from earlier
palatalized fY. 8Lcpgepa beside 8L'\Iupa may have had -pfY-; £A(A)0'\l next to aA(A)u�T)<;
goes back to *aJYap-, with the common variation a / 0 before a labial. A clear example
is AaatTo<; with, next to it, AEatTO<; and AUaTaL, AUaTaUpo<;. It may be interpreted as
representing PG *lasYt-.
Kuiper Lingua 21 (1968) : 269-277 pointed out that the substrate language had
labiovelars. He especially pointed to 9aAuKp0<; next to e9uAV'\Ia, 9UArcW. I added a
few remarks in Beekes Glotta 73 (1995/6) : 12f. From Mycenaean, we have a-to-ro-qo
(av9pwrco<;) and qe-to (rcI90<;), Mo-qo-so (Mo'\lo<;), qi-si-pe-e (the dual of �lcpo<;).
Further there is A-i-ti-jo-qo (gen. Ai910rco<;), ocp9aAflo<; with the variants OKTaAAO<;
and omIA(A)o<;, which cannot be explained from Indo-European. Instead of �lcpo<;,
we would perhaps expect **'\I1cpo<;. So the developments are largely as those of Greek,
but not completely.
Pre-Greek probably had a Iyl and a Iw/. Initial ya- presumably often lost its y-,
but it may sometimes be represented by ia- as in '(afl�o<;, 'Iaawv. The ending -ula

may have been -uy-a (a Pre-Greek y may have had a different development from y in
inherited words). In the same way, -ala may derive from PG *-ay-a with a variant
-Ela, cf. I1T)vEAOrcEla. Perhaps, the y disappeared in some cases, giving yala beside ya
(see below on the suffrx -aL- I -E(I)-).
Initial w- was often lost (ava�), but wa- may also have been rendered by oa-, as in
'Oa�o<; beside Cret. Fa�o<;. The same holds for 'OlAEu<;, which has been considered to
be identical with the root of 'IAo<;) . We find ua- (which became ua-) in UUKIV90<;,
Cret. FUKlv90<;. Fur. 377 assumes a prothetic u- in the latter word, but this seems
improbable to me. Another example may be ua/£Ao<;. The differences are probably
due to the date at which the word was borrowed and depend on whether the Greek
dialect concerned still had a F at that time. Another treatment can be found in the
word for 'truffle', for which we find OUITOV, OlOVOV (also -w-), U8vov (also -w-), or
hov. These are probably all renderings of *wit-. (Fur. 184 again assumes a prothetic
vowel, PT- I OpT-, which does not seem to be the right solution. He further assumes
a variation *wit- I wut-, which also seems improbable to me, though the variation I I
u is attested.) Rather, u- is a form of OL-, with the -0- changed under influence of the
-1- (cf. Lejeune 1972: 174, and note that Greek did not allow -UI- before consonants; of
course, 01 became U in Boeotian in the 3rd c. BC; variation 01 I U is found in more
Pre-Greek words). This case nicely shows that variation in Pre-Greek words is due to
different rendering of the sounds of a foreign language, and therefore has to be taken
seriously. �puKaAov· porcaAov (H.) probably attests a development *wrak- > �paK­
(as Fur. 147 remarks on KaAaup0'\l: "Die landlaufige Etymologie <connecting> percw
... ist wohl ohne weiteres aufzugeben."). aopoa· rcaALOupou dOo<; 'sorb-apple' (H.)
continues *sorw- (cf. Lat. sorbus, Fr. sorbier, Fur. 230).
It seems that there was no initial aspiration in Pre-Greek. Furnee has a few words
with a-, e- (one or two with 1-; none with 6-, �-, w-). Several of these are doubtful;
best is alflaCYlu (alflol). One might conclude th,at the language had no h. This would
agree with the fact that aspiration is not a distinctive feature in the stops. However,
this conclusion is remarkable for �pw<;, "EAAT)vE<; and "HcpaLaTo<;, which we expect to
be Pre-Greek words (but note that Myc. a-pa-i-ti-jo does not have a2-). Of course,
aspiration may have been added secondarily in Greek in individual cases, cf. the
variation in acp9a I acp9a and eAEowvT) I eAEOwvT), which is a variant of OEAEOWVT).
However, Prof. Ruijgh pointed out to me that Mycenaean had toponyms (a2-ra-tu­
wa) and personal names (a2-ku-mi-jo) with initial h-; it also occurs in inlaut (pi-a2-1a,
ko-ri-a2-da-na); C£ further e-ma-a2 (/Hermahasl 'Hermes').
Originally, I thought that Pre-Greek only had three vowels: a, i, u. The Greek
words concerned often have E and 0, but this would not be surprising, as the three
vowels have a wide phonetiC range, and the phoneme lal may have sounded like [e]
or [0] in many environments. The main reason for me to assume this simple three­
vowel system was the fact that the system of suffrxes has a, i, u, but not e, o. We have
-ay-, -Iy-, -uy-; prenasalized -ayy-, -Iyy-, -uyy-; likewise -a9-, -19-, -u9-; and
prenasalized -av9-, -lv9-, -uv9-, but no forms with -Ey(y)-, -oy(y)-, etc. The only
(but as a variant of oAuv90<;), and
cases I noticed are 'P�aKov90<; and oAov90<;
flT)AOAOV9T) with a variant flT)A( oA)uv9T) .






Recently, I have become more inclined to assume a system with the usual five
vowels, because there seems to be a distinction between ilie two variations U I £ and
U I 0, on the one hand, and a stable, not interchanging u, on the other. This would
point to a system with a, e and o. On the other hand, it is diffIcult to explain why the
suffixes do not show the same variation that we find in the root vowels.
It is essential that the palatalized and labialized consonants colored an adjacent U
to £ and 0, respectively. On the effects of palatalized consonants see Beekes 2008: 4655. Fur. 340 has a rule U > 0 before 0, w, U (e.g. KUAUPOC:; I KOAUPOC:;); this can now be
understood as the o-like realization of lal before high rounded vowels in the
following syllable (see 15 .3 - 2) .
So, e and 0 originally were variants of the phoneme la/. It is difficult to establish
whether they had already become full phonemes in Pre-Greek. A good illustration of
the case is the name of Apollo. In Hittite, Appaliunas renders Apollon- (see Beekes
JANER 3, 2003) . We know that Greek originally had A1t£AA-, with -£- arising from
-a- before the palatalized P. The -0- developed only later in Greek, but I assume that
the Hittite form still shows the -a-. The Pre-Greek form was ApaPun-.
I have long doubted (and still doubt) whether there was phonemic vowel length
in Pre-Greek. Greek substrate words quite often only have a form with a long vowel.
Vacillation is sometimes found, as in 8plVUKTj beside 8pivu� (see B 6.2), and note
6Pplf.LOC:; beside Pplf.LOc:;, Pplf.LTj. Quite a different argument is the following: axupov
and 1tlLUPOV both mean 'chaff; it is therefore probable iliat they contain the same
suffix -up-; but in the first word the u is short, while it is long in the second.
Note that Tj often represents a (ya8uAAlC:; I yTj8-), and as our knowledge of the
relevant dialects is rather limited, we often simply do not know whether Tj represents
an older a or e. If we had not had Dor. oloapoc:;, we would not have known that it
contains an old a. Also, A�f.Lvoc:; represents Aaf.Lvoc:;. There are well-known Pre-Greek
words with Tj < *e, like 01t�AatOV.
I assume two diphthongs, ai and au. If there were no e and 0, we do not expect
other diphthongs. A diphthong w is rare (Fur. 353 Anm. 5; I found some 12 instances
in the whole of Furnee's material); it interchanges with UU. Fur. 339 Anm. 2) calls £l
"(in mehreren Fallen) nur eine Nebenform von at". Also, Ol is rather rare, and we
may find ou more often, but mostly interchanging with other vowels (see the remark
on the suffIx -oup-). See further section B6.1 on vowel variation.
Regarding the accentuation, I noted vacillation in: appuf.LlC:; I -f.Llc:;; UiyWAlOC:; I -lOC:;;
axupoc:; I -oc:;; axwp I axwp; KOpUOOC:; I Kopu06c:;; KOPUOUAOC:; I KOPUOUAAOC:;; f.LEOlf.LVOC:; I
f.L£Olf.LVOc:;; OlKUOC:; I 0lKUOC:;; UplOXOC:; I UplOOOc:;. Note also the almost identical forms
such as AUKU\jIOC:; I AUKO\jlOC:;. This does not imply that the language had no clear
stress: the Greeks who adopted a word could simply have been uncertain about it.
The phenomenon may, however, be important heuristically: such variation is very
rare in inherited words.

1. au: Of course, uu does occur in PIE words, but only when it derives from *h2eu
(mostly in initial position) or eh2u. Examples: PAUUO£C:;, PPUUKUC:;, ypUUKUAUC:;,
KUVUUaLpOV, KUOUUpU, LPUU�uvu; Auppuuv06c:;.
2. �: As is well known, *b was rare in PIE. In Pre-Greek words, it seems to occur
relatively often. Examples: apAupOl, apupPTjAOC:;, appuATj, aLuppuKLOC:;, PUPPlAOC:;,
80pupoc:;, KlPUAOC:;. It is frequently found word-initially. Of course, p may also go
back to a Pre-Greek labiovelar (i.e. labialized velar): e.g. PU01A£UC:;, Myc. qa-si-re-u.
3. p�: The cluster is possible in PIE words, but it is rare (see on p sub 2. above).
Examples: apo£AAov, apoTjpu, apoTjC:;, ,(poTjC:;, UULO-KUPOUAOC:;, KlPOTjAOC:;, KUPUPOU;
4. y& Cf. Fur. 3185. There is nothing against PIE *gd, but it is infrequent. Of
course, the group is reminiscent of po. Examples: ayouc:;, af.LuyOCtATj, YOOU1tEW (cf.
KLU1tEW), '(yoTj, KPlYOUVOV, AUyOTj.
5. yv: Example: iyvuc:; (iKVUc:;). On Xv, <pv, see the section on the suffIxes.
6. �v: The sequence is rare in IE words. Examples: aKlovoc:;, aAU1tUOvOC:;, apuXlovU,
A£1tUOVOC:; (AU-), Ol1tUOVOC:;; 'APlUOVTj.
7. KL: The group is regular in PIE, but in Pre-Greek it is found with variants; see
B5.5. Examples: apluKLoV, PUKLat, OlKLU.
8. KX: The group can hardly be of IE origin, but it is not frequent. I noted PUKXUp,
ACtKXU, OUKXUP, OUKXUC:;; BUKX0C:;, BplUKXOC:;, BUKXlC:;. The group -KX- is the geminate of
X. Cf. on 1t<p, L8.
9. �v: The group is certainly possible in PIE words, but it is also frequent in
Pre-Greek. Examples: af.L<pl-KEA£f.LVOV, FOlf.LVOC:;, ,(uf.Lvoc:;, pUOUf.LVlULTjC:;, KpTjf.LVOc:;,
AUf.LVU, AWpUf.Lvov, f.LEplf.LVU, poouf.Lvoc:;, OlyUf.LVOV, OlOplf.LVOV; ALUf.LVlOC:;.
10. ou: The diphthong is perfectly IE, but it is found several times in Pre-Greek. I
do not think that Pre-Greek had a diphthong -ou-, but it may have arisen from e.g.
-arw-, which often surfaces as -oup-. Examples: O£VOOUKTj, OKlOUpOC:;, oLpou80c:;,
LUYXOUpoc:;, LOU<POC:;, <puvooupoc:;, <pOUOKOC:;, XAOUVTjC:;.
11. 1t<P: The group can hardly be of PIE origin, but it is rare in Pre-Greek words,
too. Like in the case of KX, it is the geminate of <po Examples: apXl�U1t<PTjC:; (?); L:U1t<pw
12. p& On a morpheme boundary, the group is possible in PIE. Examples from
Pre-Greek: ayEpou, KU1tUPO£UOat, KUpOUf.LUATj.
13. pKV: A rare group, perhaps there is even no reason to speak of a group.
Examples: apupKvu, P£PKVlc:;.
14. pv (variants po, vo): Examples: Kl01pVlC:; (-vo-), aXEpou (-vu), OKU1tEpOU. See
the section on the suffIxes.
15. A ° occurs both word-initially and between vowels, where it has disappeared
in most inherited words. Initial: OUpUHU, OUYUPlOV, OUVU1tLlV, OUVOUAOV,
ouppu<p8£iv, O£KOUU, 0lpuvTj, OlyUf.LVOV. Intervocalic: ayuouAAlC:;, ayxouou (eyx-),
u'i8ouo(o)u, uif.Lu01u, U'(OUKOC:;, aA£loov, opoooc:;. After resonant: aAooc:;, PUAOUf.LOV,
Y£AOOV, yEV01f.LOC:;, f.Lup011t1toc:; (-U1t1toc:;).

2a. Characteristic sounds and sound groups

In Pre-Greek words, we find some sounds or clusters that are rare in PIE words. In
brackets, I give the variants.



16. O'�: The group is hardly known from inherited words (o�£vvul-u is
problematic). Examples: ao�oAo<;, 810�Tj, 'Ao�no<;. -o�- may continue Pre-Greek
_sgW_: Myc. ti-qa-jo may stand for IthisgWaiosl elo�a10<;.
17. ay: Again, this group is hardly known from lE words. It may sometimes
continue -tYg-, as in Cq.lU0y£ACt, AOy£AaTa<; (see 5.5). Examples: CtAlOY£W, uOylvTj,
cpuoyuvov, Ctoyuv8Tj<;, 1tloyl<;.
18. O'K, aT: These groups are well known from lE, but mostly in word initial
position. See section B5.5. Examples: �£OK£pOl, �U0TU�, KUOT£POl, AUOTUy£l.
19. O'TA: Though the cluster contains nothing that could not be lE, it occurs more
often in substrate words. Examples: aOTAlyy£<;, 0TA£yyl<;.
20. Te: The group can hardly be of PIE origin. In Pre-Greek, it is a variant of TT
and 00 (see 5.5). Sometimes, it is clearly the geminate of 8: AT81<; beside A8�vTj.
Further examples: iTS£AU, KOT8u�0<;, IIn8uAoL
21. cpe: The cluster is possible in inherited words. Example: vUOKucp80v.
22. X!l, XV: Rather rare in lE; Fur. 110 assumes that the nasal caused the aspiration.
Examples: 8uuXfl0<;, 8uuxvu-, OUUXflOV.
23. Frisk gives some seventy lemmas with '/1-. Many words are clearly Pre-Greek,
and there are no convincing Indo-European etymologies. That many of these words
are of substrate origin is also clear from the fact that there are variants with 0-.
Apparently, Pre-Greek did not have any difficulty with ps-, as Greek has so many
words with '/1-. Originally, I thought that all words with 'i'- were Pre-Greek, but this
thesis cannot be maintained. Among the non-substrate words, 'iJUAAU originally did
not have *ps-, and 'i'- for cp8- is secondary (see Lejeune 1972: 39); the verb 'i'�w may
well be non-lE.
24. w: Of course, W is perfectly lE, but it also occurs in Pre-Greek words.
Examples: CtflUKpWTl<;, av8pw1To<;, Ctvwvl<;, Ct1TOCPWAlO<;, Ctppw8£w, CtOKUAW1TU<;,
FMKwv8u<;, Ct0flWA£lv, �UAAWT�, KMowPl<;, AWPUflVOv.
25. Geminates (see also B5.8 on single I geminated consonants): Indo-European
had no geminates. Of course, geminates arose in Greek, but they are not very
frequent. I doubt whether Pre-Greek had geminates, but several occur in Pre-Greek
words (Brixhe 1976: 95 states that there were no geminates in this language). As
Pre-Greek had palatalized phonemes, I wonder whether [Y was (often) represented by
AA in Greek. In a similar vein, perhaps nY might be represented as vv, and rY as pp,
but this needs further investigation. For 00 and TT see B5.5. Unclear are 88, KK, 1T1T,
and flfl (a palatalized mY is a rare sound). Some further examples:

StopS8: 88: a88at, Ci88t�
KK: CtKKUAO<;, �£A£KKO<;, AUKKO<;(?)
1T1T: aypl1T1To<;, AOU1T1tl<;
TT: �ITTuKo<;, AUTTU, KUTTO<;, flUTTU�O<;, fl£TT£<;, fl1TTo<;; IIlTTuKo<;.
Liquids: AA: CtAAU�Tj<;, CtAA01TlTj<;, CtfllAAUKUV, �UAAWT�, �8£AAU, �lAAlV, 1TUT£AAU
flfl: KA£flflu<;
vv: Ctyuvvu, �A£VVO<;, Ylvvo<;, AUXUVVU; L'lIKTuvvu

We also have to recall the instances OfKX, mp, Ta (see above).



pp: CtppU�UKU, �IPPTj, �lPPO�, KUppOV
Sibilant 00: u'(8ouo(o)u, acpPloou, yloou.
2b. How to recognize words as PrecGreek?

This appears to be relatively easy. A first indication is that a given word has no lE
etymology. Often, there is variation which is impossible to explain in Indo-European
terms. Therefore, the discussion of these variants is essential. Then, there are
numerous suffixes that are typical for Pre-Greek (see the list below). The meaning
may also provide an indication. The words concerned are often names of plants or
animals, or part of viticulture. Frequently, the words are sexual terms.
If we have some of the above features, it is quite clear that we are dealing with a
Pre-Greek word. The origin of the word is then indicated �PG� in the dictionary. In
many cases, we do not have enough data and can only suspect that the word might
be Pre-Greek (the origin is then indicated as �PG?�).
3. Prothetic vowel

Pre-Greek had a prothetic vowel, e.g. CtOKUAUCP0<; beside KUAUCP0<;' In most cases, the
vowel is Ct-. The numbers (Fur. 368ff.) are as follows: U ± 90, 0 10, £ 5, l 3, U 0, Tj 6, at 2.
Note that, generally speaking, U may interchange with 0, £, and at. Indeed, we have
cases where prothetic 0 interchanges with u, and the same holds for £ (e.g. £iKA- I
UiKA-, £'i'lU I Ct'i'lU). Although not all other cases can be explained away, it seems that
the phenomenon originally only concerned u. Examples: CtyuauAA1<; I YTj8uAA1<;;
CtKlpl<; I K1PPl<;; CtKOPVOl I KOPV0'i'; CtXpu8ufluAU I Xpuflu801Aat; CtVUplTTj<; I VTjplTTj<;;
CtOKUAU�O<; I (O)KUAU�WTTj<;; Ctxuvw'i' I KUVW'i'.
4. s-mobile

A large number of words shows an initial 0- before a consonant, which is absent in
practically identical variants. It occurs before a stop or m (so not before r, 1, n); the
stop is mostly voiceless, sometimes aspirated; see Fur. 390f. Examples: Y£A£VO<; I
0X£AlVO<;, (O)KlOUcpTj, KlK£pO<; I OKlyKO<;, (0)Kop8uATj, �UTUAO<; I 01T-, 1T£A£80<; I 01T-,
CPUTTUYTj<; I 01T-, 8plyKO<; (TplyX0<;) I 0TPlYXO<;, T01T£lOV I OTU1T1T£lOV, (0)fl�plV80<;,
(O)flUpatvu. A prothetic vowel may appear before an s-mobile (Fur. 3908):
CtOKUAU�O<; I OKUAU�WTTj<; I KUAU�U<;, Ctocpupuyo<; I ocpupuyo<; I cpupuy�, CtOKUAUCP0<; I
5. Consonant variation
5.1 Voiceless I voiced I aspirated stop

Furnee's conclusion was that 'Pre-Greek' was a non-Indo-European language, with
no recognizable cognates. This implies that the phonemic system may have been
different from that of Indo-European. Thus, he found that the stops show variation
between voiced, voiceless and aspirated, so that there presumably was no phonemic
distinction between voice and aspiration in the language. As there is no reason to
assume that this is a recent phenomenon, it strongly suggests that the language was
non-Indo-European. For example, mwxo<; belongs to a root ptiik- I p tok- also seen
in mw�, -KO<;. Since such a variation is hardly understandable in Indo-European




terms, the word must be Pre-Greek. Furnee's discussion of this variation runs from
p. 115 till p. 200. Even if we allow for some mistakes, it is clear that there is abundant
evidence for this phenomenon.

K / aa
T / aa
9. t / ss
10. t / st
T / aT
The analysis of these variants is not easy, and I mainly present the data here. A
question that needs to be explained is why exactly s or t are involved in the given
The most complicated instance is 5b, where we find KT/aK. In fact, the most
complicated phenomenon contains most information, and can be solved best. In this
case, one expects a cluster with k, i.e. a consonant before or after the k. One of the
two expected clusters must have undergone metathesis. As Greek did undergo a
metathesis TK > KT (and no metathesis of aK or �), we may assume that precisely this
phenomenon was operative here. Thus, for an earlier stage we may reconstruct an
interchange aK/n. This interchange can be easily explained by assuming a
consonant, probably unknown to Greek, which resulted either in a or in T. In my
interpretation, this must have been a palatalized dental, i.e. /F/. For instance,
afluaYEAu / afluy8aAll was probably *amutYgala, represented first as *amusgala or
*amudgala, the latter yielding *amugdala. A less clear example is Asklepios, who was
called A(l)aKAumoe; or A(l)yAumOe;. It could be that the name was *AtJklap-, giving
*A(i)sklap- or *A(i)dglap-. In the latter form, metathesis did not operate because
**Agdlap- was not tolerated in Greek; the dental was then simply lost. Needless to
say, it often happens that only one variant is found. The strange feature or phoneme
may also be dismissed altogether, as in 8lK£lv next to 8laKOe; and 8[KTUOV.
One might suppose that all variants in this group are due to a palatalized dental,
but this is not evident, as consonant clusters are rather rare, and as there are no
suffixes beginning with a consonant (except n, r, etc.). We may be unable to
determine what exactly happened in each case.
Type 4 is treated by Fur. 2633• Since Pre-Greek did not distinguish voice and
aspiration in stops, these often vary; so if we speak of kt or KT, this also includes
realization as X8, such as in flopox80e; below. If we consider the variation with labials,
as in pt/ps, it is clear that we are dealing with a labial followed by a dental. The dental
could also appear as s, so it is clear that the phoneme concerned was a palatalized
dental, which I note /F/. This means that we are dealing with a group ptY• In the same
way, with a velar we have ktY•
The example 8t<p8EpU next to 8t\jlapu is well-known and clear. Furnee further
gives yvuflmoue;· XUAlVOUe; (H.) beside YAufl\jlOl' XUAlvol aTofluTOe; (H.) and
compares mlAov with Dor. \jIlAOV. His example 6moe; 'cooked' next to o\jlOv is less
Among the forms with a velar, there is no problem with flopox80e; / flopo�oe;. The
best known example is 'Ep£x8£ue; (also 'Eplx8£ue;) next to Ep£xa£e; on Attic vases. I
have no opinion on 'Eplx8ovloe;; it may be a Graecisized form, and in this case it is
unimportant for Pre-Greek. See further the ethnonyms L'luTuA£-mol, L'lllAo-mlle;,
fUAll-\jIol, Au8£-\jIOl and Tpuvl-\jIOl. Other forms are less clear.

5.2 Prenasalization

Before a stop, a nasal may be present or not in Pre-Greek words. E.g. Kaxpue; /
Kuyxpue;, KOpU<p� / Kopuflpoe;, aUAaPll / aUAaflPll, etc. The phenomenon is extremely
frequent, but its precise origin is not known (prenasalized consonants?).
5.3 Nasalization

A consonant is replaced by a homorganic nasal: Kl8u<p£U£lV / KlVU<p£U£lV, <PAll8<:ovTu /
5·4. Labial stops /

m / l}
There are three interchanges: labial stop / fl labial stop / F and fl / F'
Labial stop / ,.. (Fur. 203-227). Examples: appVAll / lipfluAU; papplTOe; /
papfllTOe;; KUfllV8le; / KUplV8le;; AUKapUe; / AUKaflue;; flUaTU� / PUaTU�; aKoAUfloe; /
aKoAUpOe;; <papfluKov / <poppuv-ra; a<papuyoe; / aflapuyoe;.

Labial stop / F (Fur. 228-242). Examples: T€811nu, 8anoe; / 8uuflu; KOpUAOe; /
KuuuA6e;; Kuaaupae; / Kuauupu; Kpaflpoe; / KpUUpOe;.

,.. / F (Fur. 242-247). A difficulty here is that Greek did not preserve a F in most
cases, so that we often just find zero, and the F can only be reconstructed. This gives
rise to a certain degree of uncertainty. Perhaps, we have to reckon with the
possibility of a development 1j > b. Examples: Puauflvl-aTlle; / Puauv-lue;; KplflvOV /
KPlVOV; flE8tflvoe; / F£8lflvoe;; alyuflvoe; / alyuvoe; (also alyuvvOe;). The evidence
comprises 8 or 9 words in - flvoe;. It is found six times word-initially: e.g. fl�AOV /
�AOV; flov8uA£UW / 6v8uA£UW; note flEpO\jl / Mpo\jl (e'lpo\jl), where the latter forms
could continue *a-F£po\jl / *e-F£po\jl with a prothetic vowel. Note further Kuufloe; /
KUflllXu, which perhaps continues *KuF-ufl-, *Kufl-llK-.
5.5 Stops interchanging with a(G), with stop +

a/T or with G + stop
This kind of variation is quite complicated. I distinguished no less than 10 (or even
15) different types9• They may be represented as follows (C = consonant):
l. C / Ct
2. C / Cs
3. C / sC
4. Ct ! Cs
5. Ct ! sC
6. Cs / sC
7. Cs / ss
8. sC / ss

a. labials
n / nT
n / \jI
(n / an)
m / \jI
(\jI / an)

U aa

b. velars
K / KT
K / aK
KT / �
KT / aK
(� / aK)
aK / aa

9 Since the word 'i'LTn'tKLov / 7lLCnaKLOv 'pistachio' is probably an oriental loanword, there are no good
examples for an interchange aa / aT.

8c. C / ss



. xxvii



There may have been series with three forms, with kt I ks, pt I ps and also k or p. I
can only mention 'ApaXEloC; I Apu�'lC; next to 'Apayoc;, and perhaps, next to 8l<pElepa I
8l,\,upa, the verb M<pw (together with &'\'-), for both cf. Fur. 263.
Above, we assumed that a labial or a velar could be followed by a palatalized
dental !tY/. If this is right, we can also postulate that this consonant (labial or velar)
was followed by a normal dental. Of course, this yielded pt and kt. I assume that the
second consonant of this group (the dental) could have been dropped, which yielded
single p or k. This explains the type TI (r)oAEfloC; (Fur. §so) and � p oYXoc; (with
prenasalization) beside �poXEloc; (Fur. §Sl).
I will shortly review the 10 (15) types (I call the labials la, etc., the velars 1b, etc.).

phenomena discussed here. Fur. 253 has the heading T, 8, El I a(a), (. I think this
should be reformulated as T (8, El), H (TEl) I a (�), aa, i.e. T with its usual variants 8, El;
or the geminated H (with its expected variant TEl, which is the Greek form of
geminated ElEl), interchanging with a or aa. If the � was [sdl , it does not fit in well. As
to its interpretation, it could represent single *tY, which was rendered H or aa, or
single a, T (the variant � would then fit in, but one would also expect a variant a-r).
Examples (Fur. 2S3ff.): KlHOC; I Klaaoc;, KPOTlOV I Kpoaao<pElov, fluPTlv'l I flupalv'l,
TEUTAOV I aEuTAov, T1A<p'l I alA<p'l, yu80C; I yu�ac;, aaflWAelV I a8flWA�.
I think that the phoneme rendered by aa, Att. H (called the foreign phoneme or
Fremdphonem) was a palatalized velar, which I write as kY, cf. Beekes JIES 37 (2009):
191-197. This would be parallel to the development of inherited velar + yod, which
gave aa, Att. H, as in <puAuaaw, <pUAUHW. This interpretation is confirmed by
EluAaaaa, EluAaHa, where we have a variant 8aAuyxav· EluAaaaav (H.). Here we see
that after the nasal (prenasalization is well known in Pre-Greek), the palatal feature
of the consonant was dropped. This resulted in a velar (here realized as an aspirate).
The variant shows that we may be dealing with a velar in cases of aa I H. We can
also compare KOAu fl�Ulva beside KOAu �8U1va, which had pY; again we see that the
palatal feature was lost after the inserted nasal.
There is a third representation. We know that the name of Odysseus was
'OAuaaeu-, 'OAUHEU-. This means that it probably had a palatalized velar, *kY. But we
also find OUAl�EUC; (Ibyc. ,apud Diom. Gr. p. 321 K, Hdn. Gr., Plut.), a form which was
at the basis of Latin Ulixes. This form was taken from a Western Greek dialect,
probably Doric. Therefore, a third representation of the foreign phoneme is -�-.
lOa. T I OT may be from *tYt giving a-r or, with loss of the t, *tY > aa. Examples
(Fur. 301ff.): �aAAwT� I �aAAaual lov; flUTPUAAOC; I flUa-rPUAAOC;; flUTlC; I fluaTa�;
TIaT1A'l I TIaaT1A'l.
As we saw, it is very difficult to determine what exactly happened in each case; on
the other hand, it is clear that almos.t all variation can be understood if we start from
just a few assumptions.

la. TIT may represent a single phoneme pY, as we saw in Bl. Examples: (Fur. 31Sff.):
YVUTI- I yvum- (yvuTIn-); KOAUfl�Ulva I KOAu �8U1va; Kl�aAoc; I Kl�8'lC;; AUTI'l I Auma;
without variants note Kp oaao<pElov, aappu<pElelv.
lb. KT is most probably explained like sb, discussed above (so 1b is a part of sb).
Examples (Fur. 319ff.): iipaKlC; I apuKT'lv; floyew I floXElew; TIEAEKUV I aTIeAEKTOC;;
aKaKla / KUKTOC;.
2a. '\' may result from *ptY. It is remarkable that there is no 2b. K I �, as � is
unproblematic in Greek.lO
3a. TI I OTI, b. K I OK: Both may represent *tYp, tYk. Examples: Ella�'l I Ell�lc; (Fur.
2922), �eKoc; I �eaKEpOl; '(XAa I '(aKAUI; flUKEAAa I fluaK'l (�uaK'l); flUKOC; I fluaKoC;;
<pUKEAOV I <puaKwAoc; (Fur. 29Sff.).
4a. TIT I '\', b. KT I � were discussed above and may continue *ptY, ktY; they may
belong together with 2a. Examples: 8l<pElepa I 8l,\,upa (Fur. 263 Anm. 3); xaAu�8lKOC;
I XUAU'\'OC; (Fur. 318, 324); flopoXEloC; I flOPO�OC; (Fur. 263 Anm. 3).
5b. KT I OK was discussed above. Examples: afluayeAa I afluy8CtA'l (Fur. 301 Anm.
2); 81aKoc; I 81KTU(OV) (Fur. 279, 319).
6a. '" I OTI, b. � I OK. Fur. 393 simply considered the interchange as due to
metathesis, which, of course, is possible. *sp, *sk may represent *fYp, tYk. Examples
(Fur. 393): aaTIlvElLOv I a,\,lvEllov; 6a<puc; I '\'DUI; '(aX1ov I i�uC;; <pouaKoc; I <po�OC;.
7b. � I 00. If � represents *ktY, the k may have disappeared in other cases (which
did not give �) after which *tY became aa. Examples: KPl�OC; I Kplaaoc; (Fur. 13059);
al�8a I �lfl�a (Fur. 317); Tpau�ava, Tpw�avov I Tpauaavov (Fur. 28672); i�UA'l I iaUA'l
(iaaeAa, iTEleAa); OUA1�'l C; I '08uaaEuc;.
Sb. OK I 00 can be explained parallel to 7b: *tYk > aK or, with loss of the k, *tY > aa.
Example (Fur. 300): uplaxoc; I u p laao c;.
9a. T I 00. This is the well-known element that yielded aa I H. Furnee does not
discuss it under this heading, because he gives only one phoneme ('letter') and its
variants; for instance, he discusses aK I KT under K I KT. The situation is also different
here, as we are able to discern a distribution among the Greek dialects, and attribute
the different renderings of these loanwords to dialectal developments. Still, the fact
remains that a foreign element was rendered in different ways, as with all other


I have some difficulty with Furnee's section XI (Fur. 323-329). My conclusion is that a variation C /
although some instances remain difficult to explain otherwise.

Ca cannot be proven,

5.6 Velar I labial I dental stops: labiovelars

There is limited evidence for variation between velar and labial, between velar and
dental, and between labial and dental, and between all the three classes (Fur. 388ff.) .
We find:

K I TI, � Kh, 8 TIh
X / <p / El
X / <p
<p / El

y/ �/8

It is remarkable that the variants mostly agree in voice I aspiration. Since examples
of this phenomenon are not particularly numerous, this may be an indication that
the words concerned are not of Pre-Greek origin, but due to borrowing from a
different substrate, for instance. Examples:

y I �: �puKaAov I POn:aAOV; yHTIW I �AeTIW; XUAlC; I <paAlKpov
K I T: aaKuv8'lC; I aaTuv8'l C;


Y / 0: YUAaTllov / aoaATollov
n / T: panalV£l / paTalV£l
p / 0: aUIlPaAov / aUvOaAov
cp / 8: yvucpal / yvu80e;
y / P / 0: yEcpupa / pEcpupa / O€cpupa.
It is tempting to assume labiovelars to explain these cases, but some cases may have a
different origin (thus, ppuKaAov / ponaAov could be due to dissimilation in the first
variant). On the existence of labiovelars in Pre-Greek, see above on the phonemic
5.7. Dentals / liquids

There are some instances of variation between dentals (including n) and liquids (1,
r). This variation is incidental. Examples (Fur. 387f.):

0 / A: apAapOe; / poapol (Fur. 33027), McpVT] / AUCPVT], 'Oouaa£ue; / 'OAuaa£Ue;. Cf.
Myc. gen. da-pu2-ri-to-jo /daphurinthoio/ / AapuplV80e;, KaAuIllv8a / Myc. ka-da-mi­
ta. The interchange 0 / A and the fact that Linear B has signs for da, de, di, etc.


(which Lejeune explained by assuming a specific, unusual sound d) might point to a
dental fricative le.
8 / A: 8uma / AaHa
v / A: VLTpOV / ALTpOV

h. 0 / p: alpoa / �IIlPPaL
v / p: PA�XVOV / PA�XPOV

A / p: a�T]ple; / a�T]Ale;, Kplpavoe; / KAlpavoe;, Kpwlla� / KAwlla�.

5.8. Simple / geminate

Except for a few isolated cases, we find this interchange in v / vv, but more notably
in A / AA. On T / H and a / aa see above sub 5.5. Cf. Fur. 386£. Examples:
v / vv: aVT]80v (also T) / avvT]80v (also T); TT]Il£vle; / T�p£vva. In this context, note the
suffIx -uvV-.
A / AA: paA(A)�v; 8UALe; / 8uAAle;; anEA£80e; / anEAAT]�l; llaKEAT] / lluK£AAa (this
probably derives from PG *-al.Ya-). Note yda(a)ov, auplaa / auplaaa, and the case of
A8�vT] / AT81e; / ATTlKOe;.
5.9. (J- / zero
We discussed a / zero before consonant under s-mobile above, section B4.
An -s- from Pre-Greek is normally maintained. The only instances that I know of,
where it may have disappeared, are (cf. Fur. 241): auplXOe;, aUplaaoe; / UplXOe; (also
-laKoe;, -laxoe;, -laaoe;); aupuAAae; / UpUAAT]e;; aay�vT] / Cypr. ayuvcl; amuT] / inua.
Perhaps 'EAAue; beside L:£AAol belongs here, too. Another instance could be amov,
which is cognate with Lat. pirum which points to -pis-.

K-, T- / zero
There are instances where a velar or a dental may be absent in initial position (Fur.
391, and 13159). Dentals may also be absent in inlaut. Examples:





K / zero: KuvoapOe; / av8pa�, KaAlvO€ollaL / aAivow, KOYXVaL / 0YXVaL, Kav8�AloV /
y / zero: ylvvoe; / ivvoe;, but this form may be a late development. As an explanation, one could think of a uvular q.
T / zero: Tuyxoupoe; / ayxoupoe;, T�yavov / �yavov, Tlcpuov / '(cpuov (with l in LSJ);
o / zero: OeA£OWVT] / £A£OWVT] (also t-).
Loss of a dental in inlaut: vETwnov / vlwnov, i80uAie; / '(ouAle;, aaloapoe; / aalapoe;.
V-, A- / zero
v- and A- can also be absent (Fur. 391f): vucp8a / acp8a (also a-). AaL\/IT]pOe; / ai\/lT]poe;,
Aalln�vT] / an�vT], AaTIl£v£la / aTIl�v, Perhaps, it concerns palatalized nY, lY, which
are pronounced very 'light'.


5.12. Metathesis, shift of aspiration

There are instances of metathesis. It mostly concerns p, sometimes A. The consonant
jumps to the other side of the vowel or the consonant: KlpaOe; / KplaaOe;, Kpl�Oe;;
TEPlllv80e; / TpElll80e;. Cf. T£PIlIAaL / Tp£IlIAaL; apm� / anpl�; KEopona / KEpoona;
vu8pa� / vup8T]�. In most cases, it cannot be determined what the original configur­
ation was. In a case like i::ppwe; / £upwe;, where p may stand for (or continue) F, I
would think that the F was anticipated. It may concern an original rW.
The cases of an / \/I and aK / � are discussed in 5.5 above.
Shift of aspiration is found in some cases: 8plyKoe; / Tplyxoe;, a8payEvT] /
avopuxvT]. In the case of cpuwT] / nu8vT] the metathesis seems to have occurred in the
later history of Greek (Beekes 2003).
5.13 Secondary phonetic developments
1. We may assume secondary phonetic developments, either in Greek or perhaps
already in the original language. One might consider:

po- > PA-: poapol / apAapOl. For this case, cf. 5.7b 0 / A.
po > pp: PO€AAlOV / PpEAALOV (Fur. 308)
yo- > 0-: yoounoe; / oounoe;
ov- > yv-: ovocpoe; / yvocpoe;
KIl- > 11-: KIlEA£8pov / IlEAa8pOV
\/1- > an-: \/I£VOUA- / anovOUAT]? See 5.5.6 above.
\/1- > a-: \/IEcpae; / adcpa; \/IIHaKoe; / aIHaKoe;; cf. 'Â¥ancpw, L:ancpw.
2. a > 0 before u in the following syllable. The a was probably pronounced a little
higher before the u, and was realized as [a], which resulted in o. Examples: a�ouYYla
> 6�UyylOV, KaAupT] > KOAUPOe;, *aKapap- (KupaPOe;) > aKopopuAOe;, OOPUKVlOV for
* o (a)puKv-.
5.14 Other variation

There are a few instances of isolated and puzzling variation. I mention just one, the
word for 'night', where we have \/IEcpae;, KVEcpae;, ovocpoe;, �ocpoe;. I think that in some
of these cases, the solution may be found in a cluster. Carian, for example, allows an
initial cluster kbd-. Such clusters would have been simplified in Greek. In an
inherited word, we have the parallel of Lat. pecten, Gr. KT£Ie;, which is supposed to




continue *pkt-. If we assume a cluster *kdn- in our example, it may have been
reduced to kn- or, with loss of the first consonant, to dn-. Thus, the process is the
same as the reduction yo- > 0-, see 5.13 above. Such variant simplifications are typical
for loanwords. In this way, we could connect two of the words; but I see no way to
connect the other two.

e I W (Fur. 115). Example: apy£To<; I apKw80<;.
eL I m: see m.
2f. H I 'l (Fur. 339 Anm. 2). Examples: KeL8LOV (xe[nov) I K�8LOV, XHpaflo<; I X'lpaflo<;.
2g. w I e: see e I w.
2h. w I au: see au.
2i. e I 'l (Fur. 35842). Examples: evu <H po v I �VU<HpOV, flEpflepo<; I flEPfl'lpa, '/faKeAov I
�aK'lAov, fl�OW I flEOW (flE(W); II'lAayove<; I IIeAayove<;.
2j. 'l I L (Fur. 171"4). Examples: �A�Tov I �A[Tov, aK�vo<; I aK[vap, '/f'lflu8LOV I


6. Vowe! variation
6.1 Single vowels (timbre)

The vowels show many variants. I will discuss them in the following order: first a,
then e and 0; and within each of these groups first the short vowel, then the
diphthongs, then the long vowel (and the long diphthongs, but these hardly occur).
Note that a variation x I y is not repeated under y.
the vowel a.
a I e has 80 occurrences in Furnee's material (347). Examples: ayxouaa I
eyxouaa, apuao<; I epuao<;, yaALv80L I YEALv8OL, (aKeh[<; I (eKeh[<;, Ka[aTa I
KmE-ra<;, KaflTCo<; I KEflTCOp, KaXpu<; I KEYXP0<;, aavou� I aevOouK'l.
lb. a I o. This interchange also occurs frequently. Fur. 339 mentions that he found 80
instances. Examples: u�ouyy[a I 6�UYYLOV, uppw8Ew I 6ppWOEW, ypa�LOv I
yo�p[m, �TC[aAo<; I �TC[oAo<;, Ka�a� I Ko�aKTpa, KaAu�'l I KOAu�o<;, AUKa'/f0<; I
lC. a I m (Fur. 336ff.). Examples: uKpmcpv�<; I uKpaTCv�<;, uaUCP'lAo<; I aiaucpLo<;,
AaYfla-ra I Aa[Yfla-ra. The L here is due to the following palatalized consonant.
Id. a I au (Fur. 30237). Examples: KavauaTpov I KaVa<HpOv, flvaaLov I flvauaLov; aAo�
I aDAa�. In the last example, the u is probably due to the following labialized
phoneme lw.
le. a I w: KAa8o<; I KAwva�.
If. m I H (Fur. 352 Anm. 4, 339 Anm. 2). Examples: Kmp[a I KHp[a, KUTCmpo<; I
KUTCHpO<;, Ama[ I Aelm. Both m and eL are due to the following palatalized
19. au I w (Fur. 353 Anm. 5). Examples: AauKav['l I AWKav['l, TCETaupov I TCETWpOV;
aDAa� I eUAaKa.
lh. au I w, o (Fur. 30132). Examples: Kaaaupa(<;) I Kaawp[<;, 8aufla I 8wfla, aauaa� I
aWaLKe<;, �auKaAov I �WKO<;, KaAaup0'/f I KOAAWpO�OV I KOAAOpO�OV.
li. a I m (Fur. 338). Examples: A�8apyo<; I Aa[8apyo<;, A'lKaw I AmKa(w, TC�yavov I
lj. <;t I a. Examples: A<;t8o<; (AD8LOv) I 11.6.80<; (A�8Lov).



e I a: see under a.
2b. e l L (Fur. 355ff.). Examples: �A[Tu� I �HTue<;, E�[aKo<; I i�[aKo<;, OETCa<; I Myc.
dipa, eVTu�ov I '(VTU�O<;, KeAAov I K[AAL�, KLAA[�a<; I KeAA[�a<;, KUTWO<; I KUTLao<;,
Hacpo<; I A[aTCo<; (cp). The e was not phonologically distinguished from i, and they
were phonetically close.
2C. elL I u (Fur. 35455). Example: Kexpafl0<; I K[XPafl0<; (KLYKpafla<;) I KuXpafl0<;.
2. the vowel




the vowel o.
0 I a: see a.
3b. 0 I L (Fur. 19137). Examples: aKov o<; I aKLvo<;, i�p[KaAOL I 6�p[KaAa, 'OvoyALv I
3C. 0 I u (Fur. 358ff.). Examples: oAov80<; I oAuv80<;, aKoAo�pEw I aKoAu�po<;, aKuT'l
I -KoHa, KUOWVLOV I KOOWVW, KupaEa<; I Kopa[<;, TCpuTavL<; I TCpoTavL<;, TOTCelOV I
aTUTCTCelov. 0 and u were phonetically very close, and not distinguished
phonologic-ally (cf. on elL).
3d. 0 I ou (Fur. 359). Examples: �pOKO<; I �pOUKO<;, KOAoTEd I KOAouTw (also -Au-,
3e. 0 I W (Fur. 279). Examples: yvoTEpa I yvwTEpa, KOAAWpO�OV I KOAAOpO�OV,
cpaa[wAo<; I cpaa[oAo<; (also -ouAo<;), wpuyye<; I opu�, -yo<;; waxo[ I oax'l'
3f. OL I u (Fur. 127). Example: xpaflaoolAm I uxpaOafluAa (uKpafluAa).
3g. OL I o u (Fur. 358). Examples: KOAouT[a I KOAOLT[a (KoAOTEa), '/f0UOLOV I '/fo[8'l<;?
3h. ou I U (Fur. 12029). Examples: KTUTCO<; I yOOUTCEW, Kpouvm I ypuvo<;.
3i. ou I W (Fur. 133). Examples: flwKaoflm I flouK�(eL; AOuTC'l<; I AW�'l� (Fur. 148).
3j. W I 'l. Example: 8pwva� I uv8p�v'l.
3k. W I U (Fur. 30235). Examples: (WyLO<; I (UYYLO<;, uaawTCo<; I iaaUTCo<;, Aw�euw I
31. 0 I e. Example: yopyupa I yepyupa



L I u. There is some variation between L and u, but I do not know how to interpret
it. Examples (Fur. 364ff.): aiauflvaw I aiaLflvaw; uv8p[aKo<; I av8puaKov; �[0'lv I
�uoo[; �PLKO<; I �pUKO<;; (uyaaTpov I a[YLaTpov; KLVWTC£TOV I KuvouTCe<;; KU�eaL<; I
K[�LaL<;; flapalTCTCo<; I flapuTCTCo<;.


u I e. Example: yupya8o<; I yepya80<;.

The behavior of the diphthongs may be summarized as follows:
and (vice versa) eL I m
aul w, w
w l au
OL I u, ou
ou I u, OL, w
All this variation is understandable in terms of adaptation of a three-vowel system.




/ short:
One may doubt whether Pre-Greek had a distinction of long and short vowels (see
Bl). We do find 'l and w, however, but not very often, and the latter has several
variants. On the other hand, the variations W / 0 and 'l / £ are not very frequent
(although in this case also the difference in timbre may have been important,
depending on the Greek dialect). Variation between long en short L and u is frequent,
especially in suffIxes: y�8uov / yu8La, KU�£<JL<; / K[�'l<JL<;, 8i�L<; / 8[�L<;, Kp[IlVOV /
Kpillvov, 8piva� / 8pLVUK'l; '!''lllu8LOv / '!'LIlU8LOv, O'1tOV
K&pa�o<; / Kapull�Lo<; (cf. K'lpa<p[<;), <p£vaK[�w / 1t'lV'lK[�W 'deceive'; ny'lv(-) /
ny£v(-); yvoTtpa / yvwTtpa.
There is some evidence for short vowel + CC alternating with long vowel + C: e.g.
IlUKO<; / lluO'Ko<;; AuplO'a / AupLO'O'a; see B 1 on - L�, -u�.

C. Morphology

6.2. Long

6.3. Single vowel / diphthong:
There are several instances where a diphthong varies with a Single vowel. They can
be found above (6.1). Most frequent is a / UL, but this is due to the effect of a
following palatalized consonant. We further find a / au, £ / £u, and ou / u and OL / u.
In two cases we find diphthong alternating with a long vowel: UL / a, £L / 'l. Examples
were given above.
6.4. Rising diphthongs?

Relatively frequent in Pre-Greek words are sequences of a more closed vowel
followed by a more open one, sequences that are not found in lE. They would be
rising diphthongs if they formed one syllable, but in fact we may have to do with two
syllables. Examples are:
-w-: O'£aywv (<JL-, O'u-)
-La-: �aTLUK'l' 8[aO'o<;, 8p[all�0<;, O'[aAov, <pLUA'l, <pLap6<;. Note <JLaywv (0'£-, au-)
-LU-: iuy�
-ua-: �puaAL�wv, yuaAov (yu£-), Kuall0<;, 1tUaAo<;, 1tuavov, O'uayp[<;
-u£-: yutALOV (yuaAov), 1tU£AO<; (1tua-)
Remarkable, too, is the sequence -wu- in m vu(y) �, Ilwu<;.
6.5. Secondary vowels (or elision)

Sometimes, words show a vowel that is absent in nearly identical forms. It mostly
concerns vowels between a stop and a resonant. It is often not clear whether the
presence or the absence of a vowel is secondary. See Fur. 378-385. Examples: �puYXLa
/ �apuyXLa; OOPUKVLOV for *OpUKV- in O'Tpuxv-; O'Ktp�oAO<; / O'Ktpa<po<;; Kvu�a /
(O')K6vu�a; O'K6pooov / O'K6poov; Tov80pu�w / Tov8pu�w; Apmu[a / 'Ap1tULa; KVW'!' /
KLVW1tHOV / KUVOU1t£<;; Kopu�avT£<; / Kup�avT£<;.


1. Reduplication

Some forms seem to have reduplication, though we often cannot demonstrate this.
Most frequent is partial reduplication, where only the first consonant and a vowel
are repeated. The vowel is mostly £ or L.
Examples: �t�pa�; �t(Il)�po<;; yuyyallov; yaYYA[ov; yayypa[va; y[yapTov;
y[yYAull0<;; K[KU�O<;; O'tO'u<po<; / �[au<po<; (cf. O'o<po<;); 1l£lla[KuAov (also IlL-); V£Vl'lAO<;;
O'tO'£AL(<;); O'[aupa (also -upva); 1l£Il�pu<;(?); perhaps K[Kull0<;; KlXPall0<; (also K£-, KU-,
KLyK-); o£vopuw. Also the names KtKPO,!,; IImup'l80<;; TLTap�<JLo<;; AtA£y£<;. With
prenasalization we find T£v8p'lowv, T£v8p�v'l (cf. av8p'lowv, 8pwva�). In these
examples, I neglect the fact that there may (or may not) be prenasalization.
Other reduplication vowels are found in: AaAaIlL<; (cf. AalAa'!'), KOKPU<;, perhaps
also Y'ly�AL�.
Intensive reduplication in: 1l6Pllopo<; (lloPllupala), Ilapllapuy'l.
More difficult to judge are ytAyL<; next to aYAi<; (perhaps from *Y£-YA-, a-YA-),
KtpKa next to aKpL<; (if from *K£-KP-, a-Kp-). Also M£Il�Alapo<; beside BAlapo<; (cf.
1l£Il�pu<;), M£Il�AL<; = MtAo<;, also MLllaAAl<;.
A completely different type is perhaps found in allulla�u<; (cf. alla�l<;), and
perhaps also allaIlL8uo£<;.
2. Sufftxes
2.1 Introduction

It appears that most suffIxes have the same structure. They contain a consonant; if
this is a stop, it can be prenasalized, i.e. -�- or -Il�- ' -8- or -v8-, etc. The stop has its
usual variants, like � / 1t / <p, etc., although mostly one of these is predominant. The
suffIx usually starts with one of the vowels of the language, mostly a, L, u (we find £
or 0 only rarely, e.g. oAov80<; beside oAuv80<;). Thus, we may find e.g. ayy - Lyy uyy; av8 - Lv8 - uv8, etc.
A different structure is present in suffIxes containing -v- (mostly followed by a
vowel) directly after the root-final consonant: e.g. KUOVO<;, mO'uKva, 1l0AUXvov,
<p£vaKvl<;, O'a-rapvl<;. In this way, the groups -pv-, -ov-, -KV-, -Ilv- in Pre-Greek words
probably originated. In the case of -Ilv-, we often find a vowel again: -allv-, -Lllv-,
-UIlV-. The groups -Ilv- and -pv- are especially frequent. They are very important, as
they are found in Etruscan, which for the rest shows little agreement with Pre-Greek;
-Ilv- is found as far as in Cappadocian (see Beekes BiOr 59 (2002): 441f.). Perhaps,
the groups -avv-, -LVV-, -uvv- arose in this way, too.
Other consonants are found in suffIx-initial position, too: e.g., -p-, -0-, -y-, rarely
-A-. Examples: ,!,uopo<;, KupL8pa, 1tavaypl<;, <puAaKp0<;; O'Ka1ttpoa; Aa8apyo<;; OVLyALV.
It is often possible to determine to which series the Pre-Greek consonant
belonged. Thus, -ULV- could render -anY-, while -a/Y- seems to have resulted in -aAA­
(or -£AA- with coloring of the vowel). Likewise, -£LP- could r:epresent -arY-. This
thesis would be nicely supported by the segment -aup-, if this represents -arW- (e.g.
aupoO'xuo£<; beside apaO'xuo£<;, if this form had * -arW-) . Cf. B1 above.




Another type of suffix has a followed by a dental: KCtvaa80v (-aTpov), AataTPOV or
another stop Ev8puaKov, Q1JPoaxCt<;, Kavvap[aKa; these forms may have been partly
adapted to Greek suffixes (-TpOV). See below on the suffix -aT-.
A form such as -WT- is deviating; we do not often find a diphthong before the
consonant. Does it stand for *-aut- from -atw? Cf. -aiy- in EAatOV, where we may
suspect ayW or awY (but it may be part of the root). See further section B1.
Not seldom do we find a long and a short vowel with a suffix (= consonant), e.g.
L8 - l8, UK - UK. In the case of up, one might again think of urY > uir, although rY is a
rare phoneme (like mY).

4. -ao-: upaaxCt8e<;, JleJlppCt<;, anupCto-.
5. -a8-0-: uanCtAa80<;, yupya8o<;, aTr1Jpa8OL. TN 'Y PVCt8LOV (Epidauros).
6. -at-I-e(L)- before a vowel: There are words in -ata I -e(L)a, such as ypuJl£a I
ypuJlela (also ypuJlcia) I ypuJla[a (note the hesitation in the accentuation). I
suggest that the suffix was -ay-Ca), which was pronounced as [-teya] or [-eya] (we
saw that n often varies with at). The speakers of Greek identified the suffix with
Gr. -at- or -eL-, but the -y- could also be lost. In this way the three variant forms
can be explained. Further examples are KOAOLT£a I KOA(0)uT£a, Kopxup£a
(KOpKoopua in H. is probably an error); KwoeLa I Kwoea (note the short a), beside
KwoULa I Kwo[a (these are not entirely clear to me, but cf. AJlCt8ULa I AJlCt8na).
Furthermore, *-ay-a is likely to be the same suffix as -£[(1 which makes feminine
names, e.g. AJlCtA8na, IIfjveAOneLa, 'I<pLJl£ona (note that in Myc. Ipemedeja, the
-j- is preserved, cf. Ruijgh 1957: 1553). Of course, many place names end in -eLa:
KaoJle[a, KaAaupna, Kepuvna, M[ona, LKeAepoe[a, AepCtona, etc.
The final was often adapted to -a[ii after the dominant type, which is derived
from the adjectives in -a'io<; (see Chantraine, Form. 91): type uvaYKa[fj; cf.
ppUKTa[a, oLpKa[a, mpa[a.
We also find -eta used in nouns: oaupela, �aAe[a, KOuAUpCtTna.
Nouns with -£0- are very rare; we find: YWAeo<;, eLAeo<;, KOAeov, VLKUA£OV,
au<peo<;(?), <pwAeo<;. It may further be found in 'OK£avo<; < *-kay-an-, note the by­
forms 'OYfjv, 'Oyev-.
Beside -ata, -£la, we may expect thematic -at-O-; we find it e.g. in O[pKatOV,
an�AatOV, ,/,L<palov; ypa,/,a'io<;, *aKapapato<; (reconstructed by Fur. 169).
7. -at(F)-o- (see Fur. 23322, 25532): Partly from -atFO-; it is often impossible to
establish whether a form had a -F- or not. See also 6. above. Examples: uKuAalov,
upato<;, payalo<;, paAatOV, O[pKatOV, EAatOV (Myc. era3 / rawo), JlCtTatO<;, Jlwaa'iov,
a[patov; AXatF0<;. TN AaTunCtAata (Fick: 58).
8. -atp-o-: TN IIeppatpo[ (Thess.).
9. -at8-: TN LUJlat8a (Thess.), IIepat8ci<; (Arc. deme), KeAat8el<; (Thess. deme),
Kuvat8el<; (Arc. deme).
10. -atV- (Fur. 171"7): aKatVa, -ov, POA[TatVa, YCtyypatva, KOAupOatVa (also -uJlp-),
Kopu<patVa, Jlupatva, aJlupatva, Tp[atVa.
11. -atp-(0-) represents -arY-: Kunatpo<; (also -eLpov, -fjpL<;, -epo<;), JlCtxatpa.
12. -aK- (Fur. 15 8 64): UPUpTCtKfj, aDAa�, panCtKfj, 86va� I owva� I oouva�, 8pLVCtKfj
(8plva�), 8wpa� (also -fj�, -iiKO<;), KauvCtKfj, 8UACtKfj, Tr1aTCtKfj, <pCtpJlaKov. TN
ZCtpii�, -fj� (Lac.).
13. -aA(A)-o- (Fur. 25428, Beekes 2008): upupaAAo<;, aiy[8aA(A)0<;, KopuoaA(A)o<;
(also -6<;), nCtpOaAo<;. TN KaaTaAla (Phoc. source), <DCtpaiiAo<;, LTLJJl<piiAO<;
14. -aJlp - o - (Fur. 184): ol8upaJlpo<;, 8p[aJlpo<;, ,(aJlpo<;, KapCtJlpa<;, a�paJlPo<;.
15. -aJlv-o-: o[KTaJlvov, pCtoaJlvo<;, a<p£voaJlvo<;. TN L£oaJlVo<; (Crete).
15. -aJl - o - : apTaJlo<;. TN K[a(a)aJlo<; (Cos), II£pyaJlov, KwyaJlo<; (Lydia), KuaJlov
(Kydon.), yopaJlo<; (Kydon.).

2.2 Survey of the suffIxes

In principle, we find one of the three vowels of the language followed by a
(prenasalized) consonant: a, i, u + (mlp, (olT, (olK. The groups actually found are, in
Greek letters (forms in brackets are rare or less frequent):




















So, we do not find: 1. VNn and 3. VN<p, 7. VNK, 9. VNX (except for oaAayxav).
In the same way, we find vowel + C. The consonant may have the normal
variation: plain, voiced, aspirated. A palatalized consonant could color a preceding
and/or a following lal to [e] , which may also appear as n. This phenomenon is often
seen in languages with palatalized consonants, such as Russian and Irish. Thus, we
find -arY- represented as -atp- (-np- is also possible). A palatalized -[Y- may be
rendered as a geminate -AA-.
If a labialized consonant followed or preceded an a, this vowel may have been
perceived as (an allophone of) 101. For example, -arW- may be represented as -aup-,
with anticipation of the labial element, but also as -oup-, in which case the a was
The suffixal consonant may be geminated; as there is frequent variation between
single and geminated consonants in the language, there possibly was no opposition.
Vowels could be either short or long; in suffixes, a long vowel was quite frequent.
A long Lt was sometimes represented as w.
2.3 The material

The examples are mostly taken from Furnee, to whom I refer for details. Words can
also be checked in the present dictionary. Variants are given in brackets. I added
geographical names (TN) from Fick 1905, and some more material, with references.
1. -ap-(0-) (Fur. 107): uypCtKapo<;, UA(A)CtPfj<;, UaKCtAapo<;, uTT£Aelap0<;, KCtvvapo<;,
KOAAap0<;, Jl£a(a) apov, AaTpapo<;, JlCtTTapo<;. TN KaTTCtpLO<; (Rhodes, Fick 47),
KCtaTapo<; (Caria).
2. -ay-: apnay- (cf. Chantraine 1933: 397ff.), ACtTa�, Ol'j pa�.
3. -ayy-o-: a<pCtpayyo<;.
3a. -ayx-: oaACtyxav.





16. -av-o-: TN 'Iup8avo<; (HN Crete, Elis), ATIL8avo<; (HN Thess.), 'HpL8avo<; (HN),
A v8avo<; (Fick: 18).
17. -av-: y£vnav�.
18. -av8-: aayuv8Tj<;.
19. -av8p-: y£Aav8pov. TN T�Aav8po<; (Fick 51) , TUllav8(p)0<; (Pamph.), MUlav8(p )0<; (Fick: 53), C!>oA€yav8po<;.
20. -avS/T- (Fur. 19135; 2167" -aVT- unless otherwise stated): aAl�a<;, aaKuvSTj<;
(aKuvSa<;), KLAAl�a<; (but K£AAl�aT-), 6Kpl�a<;, <puAavS/To<;, l\�avT£<; (Fick: 69,
etc.), M€AavS/T-, IldpavS/T-, [[yavT£<;, Kopu�avT£<;. TN Ba�puvnov (Chios).
21. -avv-: TUpavvo<;.
22. -a�- (cf. -L�-, -o�-): ulla�a, aTpu<pa�u<;, aa�u�a<; (also -KT-), TN KupTupna�ov
23. -an-o- (Fur. 23531) : upvano<;, yauaano<;, Ilovano<;. TN MwaaTILo<; (Crete, Fick:
24) .
24. -ap (Fur. 13475), mostly neuters: '(KTap, Ku8ap, V€KTap, aKlvap, oii<pap; adj.
lluKap; animo oap, Mllap (gen. - PTO<;), cf. Myc. dama beside duma.
25. -ap- (Fur. 25736) : uaKapo<;, �aaaupa, yu8apo<;, ylyyAapo<;, KlaSapo<;, Kuaaapo<;,
Awxupa, <puAapa. Also al8apo<;? TN l\mapa (Crete, Lycia), IlCnapa (Lycia),
M€yapa (Fick: 75) , AAAapla (Crete).
26. -aa-alo- (Fur. 15i7): Kupnaao<;, Kallaao<; (Ku�ULao<;), nuyaaa. TN Kup�aaa
(Crete), Il�8aaa (Mess.), Ilayaaal (Thess.).
27. -aaa-o-: TN 'PUTLaaao<; (Crete), Kpuaaao<; (Crete), MUKuATjaao<;, Ta<pLaaao<;
(Fick: 32) .
28. -aT-: anuTTj, �AaKuTTj. TN KalpaTo<; (Crete), MlAaTo<; (Crete, Pick: 27).
29. -aup-alo- (this may continue -arW-): (a)<pavpo<;, <pAavpo<;, (a)llaupo<;, uyAaupo<;,
STjaaupo<;, Kaaaupa (-a<;), AUaTaUpo<;, n€Taupov (£u). TN 'Enl8aupo<;.
30. -ax-: �oTpaxo<;, KUIl�axo<;, a£Aax0<;.
30a. -a'/l-: AUKa'/l0<; aKLv8a'/l0<;.
31. -y8-: perhaps unpLy8a [adv.] .
32. -yp- (cf. on -p-): navaypl<;, auaypl<;.
33. -£8-: TN T€v£80<;, A€�£80<;, perhaps in AaK£8alllwv.
34. -£�-a: see below sub 73. on - La-.
35. -£LP-O- (may continue -erY-, -arY-): a'(Y£Lp0<; Kun£Lpov, aU�£Lpo<; (aan€p8Tj<;);
36. -£A-alo- (cf. the next): alluay€Aa, aa<p68£Ao<;, �PlK£AO<;, 8pu'/l£Aa, (tTIL)�U<p£AO<;,
iTS€Aa, KU�£Aa, perhaps 8uan£lln£Ao<;.
37. -£AA-a/o- (cf. 36.) : aKpoan£AAo<;, �uT£AAa, �8€AAa, nUT£AAa, n€AAa.
38. -£11-0- (Fur. 15142) : iUA£Il0<;, KoaA£llo<;, n(T)oA£Il0<; (if not lE), S£A£�OV.
39. -£Ilv-(0-) (Fur. 15144): ull<pL-K€A£llvov, KapT£llvl8£<;. TN L€A£IlVO<; (Pick: 95) .
40. -£vv-a (I wonder whether nY could give vv): T� �£vva. Cf. �A€vvo<;. Cf. Lat. (from
Etruscan) (doss-)ennus, Porsenna.
41. -£p-a/o-: 8L<pS€pa, aaK€pa (also -Tjpa), KaaalT£po<;. TN 'OA£pO<; (Crete).
42. -£T-O- (Fur. 1154) : KUL£TO<;, Kaln£To<;, lluan£Tov, v€n£To<;, TTjAuy£TO<;? TN 3un€TTj
(Att.), TaUY£To<;.



42a. -W- as in nom. -£u<;: �aaLA£u<;; several PNs like AXLA(A)£U<;, '08uaa£u<;.
43. -wp-: see -aup-.
44. -WT- (Fur. 173, 1817) : �aaKWTal, KpaTWTal.
45. -Tj�-a/o-: KUpTj�O<;. TN Kuv8Tj�a, T€v8Tj�a, ToppTj�o<; (all in Lydia).
46. -TjS-(0-): TN IlmupTjSo<;, LWUPTjSO<; (Fick 67) , KLKUVTjSO<; (Pagas.), IlupvTj<;,
-TjS- (Att.). Cf. -as-.
47. -TjK-, -TjX- (Fur. 199, 2457°) : �� PTj�, Mv8Tj�, AW�Tj�, n�ATj�, TpU<pTj�, <P�ATj�; KUIlTj� /
48. -TjA-O- (Fur. 1155): a�poKTjAO<;, a�up�TjAO<;, aau<pTjAo<;, �UKTjAO<;, KU�TjAO<;,
Kl�8TjAO<;, V£VlTjAO<;, TPUXTjAO<;, <puaTjAo<;.
49. -�v (Fur. 172"8): aTIl�v, aTTay�v, aux�v (ufl<PTjv), �aA(A)�v, 80SL�V, taa�v,
Kaflaa�v (-aao<;), KU<p�V, a£Lp�v, aWA�v, Tay�v, n��v, perhaps a8�v. TN eTjp�v
(Crete, Fick: 25), TpOL��V, Apa8�v (Crete).
50. -Tjv-: yAa�p�vTj.
51. -TjP: anlvSTjp; AiyAaTjp? TN IloSTjp£u<; (Crete), KUSTjpa.
52. -Tj P- (Fur. 204'°): u�8Tjpa, uv8Tjpa, ai'/lTjp0<;, aaKUATj poV, aaKTjpu (-€pa); ill�Tj PL<;,
KunTj pL<;, A£�Tjpl<;.
53. -Tja(a)-a/o- (cf. -aaa-): TN MupnTjaaa (M Paros), MUKaATjaao<; (Boeotian, Fick
80) ; Ap8TjTTO<; (Att.),'YflTjTTO<; (Att.).
54. -TjT-(O-) (Fur. 172"8): aA(A)u�Tj<;, KUVTj<;, A€�Tj<;, lluaSATj<;, TunTj<; (Myc. tepa).
Perhaps also UV(V)TjTOV (also -SOY, -aov)? TN MaaTjTa (Fick 71) .
55· -TjTT-: see -Tjaa-.
56. -Tj'/l-o-: TN A,(8Tj'/lo<; (Euboea), faATj'/l0<; (Thracia).
57. -S-o- (see Chantraine 1933: 368, and cf. -vSo<;): �p€vSo<;, KavSo<;, an€A£So<;,
58. -Sp-a/o-: KupLSpa, llupaSpov; HN Al�TjSpa. On -aSpov see Fur. 30339:
Kuv(v)aSpov, anUAaSpov. Cf. on -aTp-ov.
59. -I�-: tpuSI�Tj.
60. -Iy-: llaaTIy-, n£fl<pLy-.
61. -Lyy/K/X-: UaTALyya<;, �punyyol, EAflLYYO<;, SplyyO<; (also - LYK-, -LYX-), SWIlLy�,
-yyO<;, OALyyO<;.
62. -I8- (cf. -Lv8-, Fur. 3247) : �aA�l<;, y£AYl<;, KTjAl<;, KpTjnl<;, a<ppayl<;.
63. -L8-va (probably a combination of two suffixes, cf. on -v-): apuXL8va (cf. upaKo<;,
64. -IS-, -LS- (cf. -Lv8-): UYAL<;, -IS-, aiYLSaAo<;, aiyLSo<;, YUALSOL, �AlSLO<;, KUALSO<;.
65. -LK- (cf. -LX-, Fur. 226102): KUALKOV, KupvLKa, A€(l)KpLKa, IlUplKTj (later "[), VWpLKOV,
65a -IK-: C!>OlVLK£<;.
66. -LA-, -IA-: aiYlAw'/I, KovlATj, fl€aTILAOV, llaplATj, lluaT1ATj, (a)naTlATj (-IA-)
naaTlATj, aTpo�IAO<;. TN LKav8lATj (Cos).
67. -LAA-a/o-: upyLAAO<;, uIlLAAa, upLAAa, UaLAAa, pO�LAAO<;.
68. -Lllv-a/o- (Fur. 2467') : 1l€8LIlVO<; (also -l-), ll€pLllva, al8pLllvov.
69. -Lv-a/o-: UKLVO<;, anoALvov, (�a)�uKLvOV, yoaaUTILvov, Konvo<;, 6�lva. TN
MupLva (Lemnos), LIKLvo<; (Cyclades).





70. -lV-(0-): KUlllvov, TtU'rlvTj, PTj-rIVTj, aEAlvov, CPO�lVO':;, CPOpIVTj. TN LaAaIlLv-.
71. -lVO- (cf. -lv8e and -lO-, -l-r-): KU�lVOl':;, uAlvoov. TN Kpaualvowv (R), TIUplVOO':;
72. -lv8-(0-) (c£ -lVO-): alYlv80.:;, cwTtlv8LOV, Aa�uplv80.:;, Allllv8£.:;. TN K�PlV80.:;
(Euboea), KOplV80.:; (Fick 74).
72a. -l�-: KO-rl�l':;, KUVI�£l':;, aOpvl�a.
72b. -lTt-O-: TN EU plTto.:; .
73 · -la-alo -: upmaa (-£(a), Ku-rlao .:;. TN Auplaa, K£Oplao.:;, KTjcplao.:; (-lao.:; = -laao.:;,
Fick 25, 61).
73a. -laK-O-: aA8laKov, l�laKo,:;, llaplaKo,:;, uplaKo,:; (and variants).
74· -l-r-alo- (cf. -lO-, -l8-, Fur. 163): �Up�l-rO':;, �oA(�)l-rOV, Tt0pcpl-rOV. TN Lu�pl-ra
75· -lX- (cf. -lK-): Up<JlX0':;.
76. -KV- (probably a combination of -v- with a preceding consonant; see sub 78 on
-v-): u�apKva, OOPUKVLOV, mauKva, CPlOUKVTj.
77· -11-: TN Au-rll0':; (Caria), TIu-rIl0':;.
78. -v- (Fur. 13265), where a preceding velar may become aspirated: apuxvTj, oauxva-,
KEPKVO':;, KUOVO':;, KUAIXVLOV, Tt£AIXVTj, aa-rapvl':;, u-rvov I uovov, 'I'uovo,:;;
Ku�apvo.:; . TN Ku8vo.:; (Cyclades).
79 -o�- (cf. -l�-): Ilopo�o,:; (also -X8-).
80. -OTt- (Fur. 1 07) , often there is a variant with -a�-: £1..(1..)0'1', KaAaup0'l', -Tto,:;
(-Ocpl':;), KOAAO'l', aKuA0'l'. TN KopoTtTj (Thess.), Ka<J<JloTtTj (Core.).
81. -op- (see also the section on word end): uxopa (-upa), AETtOpl':;.
82. -oaa-a, -OH-a: TN E pllwvoaaa (Chios), A(LOHTjVO':; (Lydia).
83. -ouA-: cpaalouAo,:; (-wAo,:;) ?
84. -oup- (may contine - a rW-) : '(vooupo.:;, KU�OUpO':;, AlyyouPlov (also 1..0 -, AU-) ,
TtaAloup0':;, Ttuvooupa, auyoupov, -ruyxoupo.:;. TN AUKoaoupa (Are., the oldest
town of all; Pick: 93).
85. -oua(a)- (Fur. 19755): uyxouaa (also £-), a'Leoua(a) a (also alowaaa) , Kuoouaa.
TN AKloouaa, KTjAouaa (M K�Awaaa).
86. -TtV- (this may rather be a suffix -v- after a root): 8£puTtvTj, 0IlTtvTj.
87. -m- (this suffIx probably consisted of one phoneme P): Ilapumov,
TtEaau(ll)mov, auvaTtTlv.
88. -p- (Fur. 12437; 21562): �uAaypo.:;, y�Alypo,:;, alypat; 'Ioaypo.:; (= Lye. idakre?). See
also the suffIxes -pv-, -py- and -yp-.
89· -py-: Au8apyo.:; (also -at-, -Tj-).
90. -po-: TN Kuapoa (Caria).
91. -pv- (Fur. 48'26, 21562): aKapvuv (aKupva�), KU�£pVUW, AlTt£PVEW (also Alcp-),
aKETtapvo.:; . We also find variants without -v-: alaupva I alaupa, KU�£pvuw I
KUIl£P�Vat, aaTapvlo£,:; I aa-rapl8e.:;, KI<Jlpvl':; I Kla<Jlpl':;. Therefore, the cluster
probably arose by addition of the suffix -V-. Note that -rn- is found in Etruscan
and already in Cappadocian (Fur. 48126). See also the suffix -p-. TN <DaMaapva
(Crete), AEpva, AAlKUpVa (Aet.).

92. -aa: There are several words in -aa: 8E'I'a, ol'l'a, KO'l'a (K0'l'la), KU'l'a (Kull'l'a),
perhaps AU'l'a.
93· -aK-: uplaKo,:; (-X-, -aa-).
94. -a- (Fur. 25427, in several cases this does not seem to be a suffix, but rather the
end of a root; cf. on -aa-, -la-, -ua-): uAao.:;, KU�atao.:; (also -aa-), Iluao.:;, mao.:;,
cpupao .:;. TN TIplavao.:; (Crete).
95. -aa-: KUTtaa<Jl':;, KUTtUplaao.:;, aUplaaa.
96. -<JT-: aAu�aa-ro.:;, 8£lll<JT- (cf. Myc. temitija I timitija), A£Ttaa-r�, TtAa-ruvla-ro.:;.
TN Kupu<JTo,:;, <Data-ro.:;.
97. -a-rp- (cf. -8p-): aAu�aa-rpov, OETtaa-rpov (also 1..- ), £vua-rpov (also �-),
(uya<J-rpov, AaL(a)-rpov, alyla-rpov.
98. --r-: uacpaA-ro.:;, u-rpaK-ro.:;, ucpAaa-rov .
99. -H- (see 5.5 on H I aa): KUPlHo l, TtpOKOHa; <DEp£cpaHa.
100. -u�-: £v-ru�ov, 8opu�0.:;, lV-rU�O':; (also - ou�-) , alAAu�o,:;, aKoAu�o,:; (also -11-) ,
aKoAu�pa (-o�-); c£ 6X80l�O,:;.
101. -uyy-: Mpuy�, Ttlauyyo.:;, aTt�Auyy-, cpupuy�.
102. -UO-: allu,:;, -00':;, Ellu,:;, Ko pu06.:;, TtTjAallu,:;, XAaIlU':;.
103. -uova: TN KaAuovu (Cos).
104. -u8-, -u8-: ayvu.:;, A�Ku80.:;.
105. -Ul-a: uyula, KwoUla; 'ApTtUla. TN Klvoula (Crete, also KlVOUTj, Fick 18, 24).
106. -UK-: UIlTtU�, lOU�, aKapoulluK-ro.:;. TN Napu� (Locris).
107. -UK-: OOlOU�, K�PU�, -UKO':;, KapuK(K)Tj, auvou�.
108. -uA- (Fur. 205'4): ap�uATj, OUKTUAo,:;, Kav8uATj, KPW�UATj, IllllalKuAov (also 11£-) '
acpovouAo,:; (also aTt-).
109. -uA-: (a)KopouATj, acpovouATj (also aTt-). TN KapoalluATj (Mess.).
110. -UAA-: LI�uAAa.
Ill. -ull-: YEpaullov, Y1YYAull0':;, -8EAullvo,:;. TN KUPUllat (Crete).
112. -Ull�-: 'Leull�o,:;, KOAull�o,:;.
113. -ullv- (cf. Fur. 24366 on -umn- in Etruscan and Cappadocian): aiaullvuw,
alyullvo,:;. TN P l-r/8ullva (Crete), Aupullva (Locr.).
114. -uv- (see also the suffIx -uvv-): �o8uvo.:;, <JlyuvTj (cf. -uvv-), Myuvo.:;. TN fop-ruv
115. -uvv-: alYuvvo,:;, !llK-ruvva. Cf. on -uv-.
116. -uvo- (cf. -uv8 I -r-): B£P£KUVOat. TN Kalluvolo,:; (Rhodes).
117. -uv8/-r -: �oAuv80v, oAuv80.:;; B £P£Kuv8/-rat . TN ZUKuv80.:; (+88), Tlpuv,:;.
118. -u�-: TN 'OA6 cpu�o.:; (Athos).
119. -UTt-: '(aauTto.:; (older uaawTto.:;), llupauTtTto.:;, o iaUTtTj, -roAUTtTj .
120. -up-: a�aupo.:;, uxupa (also -opa), (EcpUP0':;, llauKupov, M8upo.:;, 6vupl(e-rat,
aaTupo,:; . TN EAupo.:; (Crete), TEyupa (Boeotia), Nlaupo.:; (Cos).
121. -up-: uyKupa, avuyupo.:; (also 6vo-), yEcpupa, AUcpupov, TtATjlluPl':;, Ttl-rUpOV.
122. -ua- (on -U<JTpOV see - <JTp ov) : upuao.:;.
123. -u-r-: mvu-ro.:;, VTjTtU-rlO':;. TN AaYlVUTtu-rOV (Crete), KOAAU-rO':; (Crete).
124. -ucp-: KEAucpo,:;.
125. -uX-: �o (a)-rpuxo.:; . TN MoauXAov (Lemnos).







126. -cp8-: Kpoooocp80v, A6.Kacp80v, lloAocp80c;, va( 0)Kacp80v, oappucp8civ.
127. -cp- (on -acpco- see Chantraine Form. 263): Ctpy£AOcpOl (also -IA-), Ilampocpoc;
(also -rroc;), O£PlCPOC;, o£oucpOC;.
128. -WK-: TN K08wKLOm (Att., Fick 70).
129. -WA-: CtrrocpWAloC; (?), cpaOKwAoc;. TN KLllwAoc; (Cyclades).
130. -WIl-: �apwll0C;.
131. -wv- (Fur. 30339): CtAKUWV, �'LWV, oavowv, <JlVOWV, OXaowv.
132. -wrr- (a variant is -ourr-): 8ullaAw\jI, aiYLAwrr-, Klvwm:Tov (Kuvoum:c;). TN
Eupwrroc; / a (Crete), Ka<JowrrTj (Epirus).
133. -wp- (Fur. 2ll50): CtXWp, Ctllavwp, �laTwp, IXwp, AdTWp. TN IHAWpOC; (Chalc.,
Fick 22).
134. -woo (see -ouo( 0)-): TN LlIPCPWOOOC; (Euboea), IIlowoooC; (Caria, Fick 26).
135. -WT- (Fur. 28383; 384'32): CtOKaAa�WTTjC;, -KauOWTOv, Kl�WTOC;, Kpall�wTov,
OiOrrWTTj. TN 8wrrpwToL

derives from stems in * -oi-; I assume that Pre-Greek words secondarily joined this
inflection. Words in -wc; are masculine: Ctxapvw(c;), � PWC;; MLVWC;, TaAwc;.


3. Word end

Word end provides an interesting situatipn, as some original finals of the Pre-Greek
language may have been preserved. Of course, Greek endings must be removed,
notably -oc;, -ov. Thus, -IOV, -uov may often continue original -I, -u: cf. Myc. dunijo
next to duni. The words in -v80c; have replaced almost all of those in -v8- (as in
3.1. words ending in a vowel

A short -a can only come from *-ya < * -ih2 in inherited Greek words. In all
ot;her cases, we may be dealing with a Pre-Greek ending -a that was originally short.
It is often difficult to see whether -a is short or long; the material requires further
study. Examples: a�apKva, uyavva, Cty£poa, Maha (?), a'LKouoa, a�oTjpa, aKapa,
aKop-va, uKooTIAa, CtKTapa, aAapa, Ctllouxpa, apoa, CtoTayava, acp8a, �pouKa, yooa,
YOAa, 86.�a, 86.poa, oaA6.yxa, 8ama, Ilooa, pOIlI�a, oaTTa, oopvI�a, ooua, K£oporra
(also X-), etc. Note forms in -ua, like apua, and in -evva. Note, further, oaAallavopa,
oKoAOrrevO pa.
For words ending in -oa, see the list of suffIxes.
a. -a.

h. -L. lE words (neuters) in -I are very rare in Greek. Examples of Pre-Greek words in
-I: (aKn, KOPI, olvam, Tayupl (Tayuplov), uKapL We may assume that many words
ending in -tOY, -uov originally ended in -I, -u. Final - IC; is frequent, too.
c. -U. Ct�apu, KOVOU, IlWAU. For -uov, see the foregoing. Final -UC; is also found several
times: ayouc;, aprruc;, CtTpacpa�uc;, �l8uv, �A£TUC;, -1l£VOUC;, IllllapKuc;, IlWUC;, rrTjAalluc;,

d. -EUC;. Though the ending may also be inherited from lE, in many words it is clearly
of Pre-Greek origin, e.g. �a<JlAeUC; (Myc. qa-si-re-u), AXIA(A)euc;. I withdraw my
considerations in FS Kortlandt on this point.
e. -wo K08w, Kop8w, 1l0TW, TlTW, fenw. The suffIx also makes feminine names in -w:
ATjTW, �arrcpw. It is usually assumed that the original inflection of all words in -w

3.2. words ending in


aoap(ov)?, CtKxavTap, �aKap, EALllap, Kuoap, V£KTap, VWKap, m:ALap.
h. -01'. Examples: CtOlYOP, CtKKop, KaKKop, K£lllloP (also -Ilrr-), rrLoop,
c. -up. Examples: \jIl8up, (Dor.) llapTup.
d. -wp. Examples: axwp, ixwp, K£AWP, \jIo8wp.
a. -a l' .

-� (stem in -K-) is found quite often:

3.3. words with a nom. in -� or



-a�: a�a�, CtKapva�, all�a�, Ctvopacpa�, av8pa�, �uoTa�' llaA�a�, 0auoa�. ava� has
a stem in -KT-.
-a�: cp£va�, pa�, ocp��
-Tj�: 86.vOTj�, ��PTj�
-I�: MOI�, CtV8£PI�, Ctrrpl�, KOAI�, apm�
-o�: �£�pO�, �lPPO�
-ou�: �pou�
-u�: �AlTU�, yopTU�; IIvu�, �TU�.
Note acc. �aneKa; ace. �plyKa.
h. -\jI: A6.Tpa\jl, AalAa\jl, aAI\jI, KOpl\jl, aiYlrro\jl, KOnO\jl, Il£po\jl, YU\jl, llovw\jI.
Monosyllabic: XP£Il\jl.
4. words in -v:

�an�v, Kap�av; ny�v. KlVOUV, Ilooouv, pw8uvec;.

Ct�ac;, a�Aac;, a8pac;(?), Ctllcplac;, CtOKaAwrrac;,
CtOKWvoac;, CtTTayac;, �a86.c; (�aTac;), �aoKac;, �uac;, KaAa�ac;, Kaoac;; A8allac;,
With a stem in -aVT-: CtAl�ac; (-VT-), AUKa�ac; (-VT-) etc.; see the suffix section.
With stem in ao-: Ctxpac;, �ouvlac;, rrpTJllvac;; see the suffIx section.

5. words ending in -ac; (a-stems):

D. The unity of Pre-Greek
The material itself shows that we are largely dealing with one language, or a group of
closely related dialects or languages. Of course, we cannot demonstrate in each and
every case that the words that are non-Greek belong to this same language. The bulk
of the known non-Greek words, however, seem to fit the general picture of the Pre­
Greek substrate. For example, KOT8u�0c; / Kooull�oC; does not only show the element
00 / T8, well-known from geographical names, but also the suffIx -u�- with
prenasalization. The pair KPOO<JlOV / Kponov also shows the element 00 / T, but
Kpoooocp80v has a suffIx added that is also typical for this language. The word
oaAayxav next to 8aAaooa (-TTa) again has the suffIx 00 / TT, but also
prenasalization. aOTAlY� / omAlY� has both the typical (prenasaJized) suffIx -Iyy- and
variation a / o. In ll�plv80c; / 01l�Plv80c; we have the 's-mobile' and the well known
suffix, while Il£PIlIC;, -18ac; has the variant without prenasalization, and 0ll�plyyec; has



a different Pre-Greek suffix. In a(f.l)�puTTOl / �pUTTO<'; (�puO'O'o<,;) we have a
combination of a prothetic vowel and prenasalization.
Other languages may well have existed in the area. Thus, it is not certain that
Hieroglyphic Minoan reproduces the same language as Linear A. Further,
Eteocretan has not yet been connected with other elements and seems isolated.
Another matter is that (non-Indo-European) loanwords from old Europe may
have entered Greece, cf. Beekes 2000: 21-31. Moreover, these may have already been
adopted in Pre-Greek, as is suggested by £pE�lV80<.;, which has a Pre-Greek suffix,
but a root which is attested (with some variation) in other European languages.
Sometimes, elements from other IE languages may also have been adopted at a very
early date, such as 1tEAEKU<';.
However, I think that it is methodologically more sound to start from the
assumption that non-Greek words are Pre-Greek. Only when there is reason to
assume that they have a different origin, should we consider this option.
E. Pre-Greek is non-Indo-European
Our knowledge of Indo-European has expanded so much, especially in the last thirty
years (notably because of the laryngeal theory) that in some cases we can say almost
with certainty that an Indo-European reconstruction is impossible. A good example
is the word yva80<.;. In order to explain the -a- of this word, we need to introduce a
h2. However, a preform *gnh2dh- would have given Gr. *yva8-. One might think that
assuming *h2e would remedy the problem, but *gnh2edh- would yield *yuvu8-. The
conclusion is that no Indo-European proto-form can be reconstructed, and that the
word cannot be of Indo-European origin. Another example is the word KpT]f.lv6<.;
'overhanging bank', for which a connection with KpEfluflUl 'to hang (up)' used to be
evident. However, we now know that most long vowels go back to a short vowel plus
a laryngeal, and that long vowels cannot be postulated at random. In this particular
case, there are simply no conceivable formations that would contain a long root
vowel. This morphological objection is strengthened by the fact that there is no trace
of the expected root-final -u- < *-h2- (as in KpEfluflUl < *kremh2-). Positively, one can
say that landscape terms are frequently borrowed from a substrate language. The
inevitable conclusion is that the word is Pre-Greek.



Classical Armenian
Cuneiform Luwian
Church Slavic
in glosses


Hieroglyphic Luwian
Low German
Middle Breton
Middle Cornish
Middle Dutch
Middle English
Middle High German
Middle Irish
Middle Low German
Modern Dutch
Modern English



Modern French
Modern High German
Modern Irish
Modern Norwegian
Modern Persian
Modern Swedish
Middle Persian
Middle Welsh
New Phrygian
North-West Greek
Old Albanian
Old Avestan
Old Breton
Old Cornish
Old Church Slavonic
Old Czech
Old English
Old Frisian
Old High German
Old Icelandic
Old Irish
Old Latin
Old Low Franconian
Old Low German
Old Lithuanian
Old Persian
Old Phrygian
Old Polish
Old Prussian
Old Russian
Old Saxon
Old Swedish
Old Welsh


Russian Church Slavic
Swiss German
Tocharian A
Tocharian A and B
Tocharian B
Vulgar Latin
Young Avestan


ace. to

et al.

according to
anno domini
before Christ
1. conjecture
2. conjunction
for example
and others
any PIE laryngeal

vel sim.

personal communication
personal name
passive past particple
sub voce
or Similarly
namely, to wit



Only the most common authors and works are mentioned here. Please refer to LSJ
for a complete list.



AB = Anecdota Graeca, v. I
A.D. = Apollonius Dyscolus
A.R. = Apollonius Rhodius
Aesch. Alex. = Aeschylus Alexandrinus
Agath. = Agathias
Ale. = Aleaeus
Alem. = Aleman
Amm. Marc. = Ammianus Marcellinus
And. = Andocides
Ant. Diog. = Antonius Diogenes
Ant. Lib. = Antoninus Liberalis
Antip. = Antipater
Antiph. = Antiphanes
Ap. Ty. = Apollonius Tyanensis
Apollon. = Apollonius
App. = Appianus
Ar. = Aristophanes
Ar. Did. = Arius Didymus
Arc. = Arcadius
Arch. = Archias
Arist. = Aristoteles
AscI. = Aselepiodotus or Aselepius
Aselep. = Aselepiades
Asp. = Aspasius
Ath. = Athenaeus
Aus. = Ausonius
B. = Bacchylides
Bacch. = Bacchius
Call. = Callimachus
Call. Corn. = Callias Comicus
Call. Hist. = Callias Historicus
Callin. = Callinus
Cerc. = Cercidas
Cercop. = Cercopes
Corn. = Comedy, Comic
D. = Demosthenes
D. C. = Dio Cassius
D. Chr. = Dio Chrysostomus
D. H. = Dionysius Halicarnassensis

L. = Diogenes Laertius
D. P. = Dionysius Periegeta
D. S. = Diodorus Siculus
D. T. = Dionysius Thrax
Dam. = Damascius
Din. = Dinarchus
Diog. = Diogenes
Dioph. = Diophantus
Dsc. = Dioscorides Medicus
E. = Euripides

EM = Etymologicum Magnum
Ep. = in the Epic dialect
Epic. = Epicus
Epig. = Epigenes
Epin. = Epinicus

Et. Gen. = Etymologicum Genuinum
Et. Gud. = Etymologicum Gudianum
Eub. = Eubulus
Euc. = Euelides
Eup. = Eupolis
Euph. = Euphorio
Eust. = Eustathius
Eustr. = Eustratius
Foed. = Foedus
Gal. = Galenus
Gramm. = Grammarians
h. Ap., ete. = Hymnus ad Apollinem, ete.

h. Hom.


hymni Homerici

H. = Hesychius
Halie. = Halicarnassus
Hdn. = Herodianus
Hdt. = Herodotus
Herael. = Heraelas
Herod. = Herodas
Herod. Med. = Herodotus Medicus
Hes. = Hesiodus
HId. = Heliodorus, Scriptor Eroticus
Horn. = Homer
Hp. = Hippocrates
Hsch. = Hesychius




Hymnus, Hymni

Il. = Iliad
Is. = Isaeus
J. = Josephus
Leon. = Leonidas (two epigrammatists)
Leonid. = Leonidas Medicus
Lyc. = Lycophron
Lyr. = Lyricus, Lyric poetry
Lys. = Lysias
Lysim. = Lysimachus
Mel. = Meleager
Meliss. = Melissus
Men. = Menander
Moer. = Moeris
Mosch. = Moschus
NT = Novum Testamentum
Nic. = Nicander or Nicias
Nic. Dam. = Nicolaus Damascenus
Od. = Odyssey
Orac. = Oraculum
Oratt. = Oratores Attici
Orchom. = Orchomenus
Pall. = Palladius or Palladas
Parm. = Parmenides
Ph. = Philo
Phan. = Phanias
Phil. = Philippus Epigrammaticus
Philet. = Philetas
Philipp. Corn. = Philippus Comicus
Philol. = Philolaus
Philonid. = Philonides
Phld. = Philodemus Philosophus
Phlp. = Philoponus
Phoen. = Phoenix
Pi. = Pindarus

PI. = Plato
Plb. = Polybius
Plin. = Pliny
Poet. = Poeta, poetica
post-Horn. = post-Homeric
Prise. = Priscus Historicus
Quint. = Quintilianus
Rhet. = Rhetorical, Rhetoric
Ruf. = Rufus
S. = Sophoeles
S. E. = Sextus Empiricus
Sext. = Sextus Philosophus
Srn. = Symmachus
St. Byz. = Stephanus Byzantius
Str. = Strabo


Tab. Herael. = Tabulae Heraeleenses

Th. = Thucydides
Them. = Themistius
Themist. = Themistoeles
Theo Srn. = Theon Smyrnaeus
Theoc. = Theocritus
Theod. = Theodorus
Thgn. = Theognis
Thphr. = Theophrastus
Ti. Locr. = Timaeus Locrus
Tim. = Timotheus Lyricus
Tim. Corn. = Timotheus Comicus
Tim. Gaz. = Timotheus Gazaeus
Tim. Lex. = Timaeus Grammaticus
Trag. = Tragic, Tragedy
Tryph. = Tryphiodorus
X. = Xenophon
Xenoph. = Xenophanes
Zen. = Zenobius
Zon. = Zonas





becomes by regular phonological development
reflects by regular phonological development
is replaced by way of analogy
replaces by way of analogy
the origin of the word (see preface above)
see also the entry
see s.v.
meaning uncertain or appurtenance uncertain
a reconstructed item of a proto-stage
spelled with the symbol X
the phoneme X
phonetically pronounced as X

copulative prefix (a a8pOlGTLKOV). <l IE *sm-�
.VAR E.g. in un:u�; by Grassmann's dissimilation or by psilosis also a-, which was
analogically extended: aAoXOe;, aOeA<pOe;; am:80e; 'even', a�lOe; 'rich'. A form like
Horn. aKOlLle; was not aspirated because the Attic redactors of the text did not know
the word, so they followed the Ionic pronunciation (cf. �ALOe; next to �£ALOe;) .
ETYM From lE *sm- in Skt. sa- (sa-naman- 'with the same name'), Lat. sem-, sim­
(sim-plex), from PIE *s1J1-, a zero grade to *sem in Skt. sam 'together', � £te;. Within
Greek, cf. also � of.Loe;, � uf.Lu. From the meaning 'together, provided with', the so­
called a E1tLLaLlKOV (intensive) developed, e.g. a-£Ovov· n:OAU<p£pvov 'with much
dowry' (H.), cf. � £8vu. It has been supposed that in some cases a comparable a­
arose from *1}- (the zero grade of *h1en 'in'), e.g. aA£Yw (Seiler KZ 75 (1957): 1-23), but
the alleged instances are probably all wrong.

a- 1


privative prefix (a aL£pTjTLKOV). <l IE *n-�
.VAR Prevocalic av-.
ETYM The antevocalic form av- arose because of a following laryngeal (Indo­
European roots always started with a consonant), which led to vocalization of the
nasal: *1}-HV- > *anV-. In Greek, the loss of initial consonants (*lj-, *s-) disturbed
the original distribution: thus alGOe; « *a-pGOe;) next to the reshaped aVlGOe;.
Sometimes this led to analogical forms, like a-o(oe; next to original av-o(oe;. In the
Myc. PN a-u-po-no IAhupnosl, the a- stands before h-, like in classical aun:voe;. In
aoplGLOe; (to opoe; < *worwo-), we see that the initial F- was originally retained before
*0 as well.
If the second member began with laryngeal + cons., this yielded Greek VTj-, va-, vw­
like in V�yp£LOe;, vw06e; < *n-h�r-, *n-h3d-. These adjectives were later reshaped, e.g.
into avwvuf.Loe; (see Beekes 1969: 98-113) In avCt£OOe;, aVCt-£AltLOe; and avCtn:VeDaLOe;,
which seem to show avu- 'un-', the last two are analogical, and the first may stand
for av-££8v-. On pleonastic a(v)-, cf. � a�£h£poe;.
a(v)- was originally limited to verbal adjectives and bahuvrihis, both in Greek and in
other lE languages; see Frisk 1941: 4ff., 44ff., Frisk 1948: 8ff., Wackernagel 19201924(2): 284ff., Wackernagel 1920-1924(1): 282f., and Moorhouse 1959. In other
languages, we have e.g. Skt. a(n)-, Lat. in-, Go. un-, all from PIE *1}-. The sentence
negative was *ne, in Lat. ne-scio, ne-fas, etc., but this use is not attested for Greek
(not in � v£n:oOee;).

a- 2

a- 3

<l PG(V) �



.ETYM In Pre-Greek substrate words, a "prothetic vowel" occurs, e.g. aaTaxu<.; next to
muxu<.;. In other words, the prothetic vowel may be present in the reflex of a
substrate word or not.
Without a doubt, a phonetic process was at the origin of these variations. It is rather
frequent, see Fur.: 368-378. The vowel was almost always a- (only very few
exceptions can be recorded, see Pre-Greek).
a- 4 in Anatolian place names, e.g. ATImao<.; / I1mao<.;, 'A9ufl�pa / eUfl�pa. � LW�
.ETYM The origin of the phenomenon is unknown, and could be different from that
of the Pre-Greek prothetic vowel. It is probably due to adaptation from a non-lE
a 5 interjection (ll.). � ONOM�

.DER a�W [v.] 'to sigh, groan'.
.ETYM Onomatopoeic; see Schwyzer: 716.

aaaTOe; Mg. unknown. In vuv flOl ofloaaov a. LTUY0<.; uowp (3 271) 'inviolable'?,
a£9Ao<,; a. (<p 91, X 5) 'infallible'?, KUPTO<.; a. (A. R. 2, 77) 'invincible'? � ?�
VAR For the varying length of the vowel, see LfgrE s.v.
.ETYM Comparable to � aT'l and � aaw 'to damage'. Cf. perhaps au�aKTOl' a�Aa�£1<.;
'undamaged' (H.). Note that the privative prefix is a-, not av-.

liac5a . £vo£la. AUKWV£<.; 'want, lack (Lacon.)' (H.). � ?�
.DER aao£1v· aTIop£1a9m, amT£iv 'to be in distress, abstain from food' (H.), see on
� ao'lv, We also find aao£iv· 6XA£1v, AUTI£ia9m, aOtK£iv 'to disturb, be vexed, be
wronged' (s.v. � avMvw, � �8U<.;).
ETYM The forms and meanings are uncertain; see DELG for an uncertain suggestion
by Frisk. Latte assumes aoo = a�a, contrary to the alphabetical order.


aaw [v.] 'to damage', med. 'act in blindness' (ll.). � IE? *h2euh2-�
.VAR Beside pres. aUTm (T 91 = 129) < *aF&'£Tm only aor. aaaa (contracted aaa) <
*aFaaa, med. -ufl'lv, pass. aaa9'lv. With -aK-: aaaK£l' <p9dp£l, �AUTtT£l 'destroys,
damages' (H.); difficult KaTt�aaK£' KaTt�Amjt£v 'damaged' (H.), for *KaT-a�aaK£?
.COMP aWl<ppwv (wrong for aam-) 'damaged in mind' (ll.), cf. aam<popo<.;· �AU�'lV
<ptpwv 'bringing damage' (H.); avaT(£)l 'without harm, with impunity' (A.).
.ETYM For PGr. *awa-je!o-, we may mechanically reconstruct an lE root *h2euh2-, but
there are no known cognates. A verbal noun * awa-teh2-, with which compare auuTa
(Alc.), yields � a1''l 'damage, guilt, delusion'. Not connected to � WT£lA�.
li�a TPOX0<.; � �o� 'wheel or screaming' (H.). � ?�
.ETYM In the sense of 'screaming', Specht compared a�a with Hom. au£ [ipf.]
'called'. See further the glosses a��p£l' �O£l 'sings' and a�taa£l' e1tlTI09£1, 90pu�£1
'yearns for, makes noise' (both H.); lastly also a�wp with the meaning �o�. See
� aU8�, � adow.
li�ayva pooa MaK£oov£<.; 'roses (Maced.) (H.). � ?�
.ETYM See Kalleris 1954: 66-73; Belardi Ric. ling. 4 (1958): 196 .


a�aK�e; [adj.] � ?�
.VAR Only Aeol. a�uK'lv <pptva [] (Sapph.), explained as �auXlov Kal TIpq.ov
'quiet and gentle' (EM).
.DER a�uK'laav (0 249) '�aUxaaav' (?) and a�aKl�ofl£vO<'; 'quiet' (Anacr.). Further
a�aK�flwv· aAaAo<,;, aaUv£To<,; 'unspoken, not understood' (H.), and a�uK'1To<,;·
av£1tl<p90vo<.; 'without reproach' (H.).
.ETYM Perhaps it belongs to � �u�w (�t�aKTm, �U�l<';) 'to speak' .
a�aKA� => afla�a.

aa�w [v.] 'to breathe with the mouth wide open' (Arist.). � ONOM�
.DER aaaflo<.; (Arist.).
ETYM Probably onomatopoeic. Another suggestion by Solmsen 1901: 284 relates it to
� a'lfll. Cf. perhaps � a�w 2, from � a.

a�aAfj aXP£1ov, AUKWV£<.; 'useless, foolish (Lacon.)'. o l o£ vw9pov 'bastard' (H.). � ?�
.ETYM Fur.: 167, 348 compares a�£AAov· TaTI£lVOV 'vile, low' (H., Cyr.) and a�aAl<';'
flox9'1pa eAala 'worthless olive tree' (H.), but not � a<p£A�<,; .

aavOa [f.] £100<'; evwTlou TIapa AAKflUVl W<'; AplaTo<puv'l<'; 'a kind of earring in Alcm.,
ace. to Ar.' (H .). � ?�
ETYM Schulze 1892: 38 explained it as *aua-av9a to the root of oU<'; < *h2ous-, but
since the suffIx is unclear (cf. oLvuv9'l), this analysis cannot be substantiated. Cf. also
Bechtel 1921, 2: 366.

a�avT«O'lv · avu�amv 'mounting' (H.). � GR?�
.ETYM For earlier *a(fl)0uvTaamv [] ? See Schwyzer: 50 .




liaTITOe; [adj.] In X£ip£<.; aaTtTOl (Hom., Hes.), later of K�TO<'; (Opp.), perhaps
'invincible'. � ?�
.ETYM Aristophanes read *CtETtTou<.;, which is also unclear. Meier-Briigger's
explanation of � aTtTo£TI�<'; does not convince me. Cf. � a£TtTo<.; and � eu<p9'l.
li(a)TOe; => a'lTo<,;.
aaoX£TOe; => £Xw.


li�aAle; => a�aA�.

li�a�, -KOe; [m.] 'board for calculating or drawing' (Cratin., Arist.). � PG?(v)�
.ETYM Etymology unknown. The assumption of a loan from Hebr. 'iibiiq 'sand, dust'
(Lewy 1895: 173) is semantically weak (rejected by E. Masson 1967: 97). On the
meaning, cf. Bruneau REGr. 80 (1967): 325-330; see also Kratzsch WZHalle 23 (1973):
126, who defends the connection with Hebrew. Borrowed as Lat. abacus. Kuiper
compared afluKlov, a�a�. AUKWV£<.; (H.), but Fur.: 221 doubts this. Yet, if the
comparison is correct, the word is Pre-Greek (variation �/fl).
a�apuJTav . yuvmKl�ofltV'lV, Ka9mpofltv'1v KaTafl'lVlOl<';. KUTIplOl 'being made to play
a womanly role, being cleansed or purified during the menses (Cypr.)' (H.). � ?�
.ETYM Related to � aa�aplXl<';, with Cyprian loss of s-?



li�apKva [f.] . Aq..lO<; 'hunger' (H.). � PG?�
. .
. ,
ETYM Fur.: 122 connects it to flapyo<; ,mad, gluttonous , whIch IS not convmcmg.
The formation with -va after -K- suggests Pre-Greek origin.

a�ap-ra[ => u<pap£u<;.
li�ap-ro<; · anAT]u-ro<; 'greedy', o l o£ aflapyo<; 'greedy' (H.). �PG?(v) �
VAR u�ap-rla. unAT]u-rla 'greediness' (H.).
ETYM Fur.: 217 connects it to fluPY0<;, which seems difficult formally; or do we have
to read * a�apyo<;? If so, the word clearly shows a prothetic vowel, which points to
substrate origin.
a�apv [n.] . 6plyavov <TO ev> MaK£Oovl<;t (or MaK£Oovla<;?) 'oregano (Maced.) (H.).
� LW�
.ETYM Related to uflupaKov 'origanum', acc. to Fur.: 210. Cf. also �apu· TLV£<; fl£v
<paUl Elufllafla £uwOe<; 'ace. to some, an odoriferous i�cens�' (H.); see L�tte. The
, as havmg
suggestion of Kalleris 1954: 75f., who assumes connectIOn WIth �apu<;
strong (heavy) odor' (with prothetic a), is untenable.


li�S£Uov [adj.] . Tan£lVOV 'low, abased' (H.). � PG(v) �
.VAR Hesychius also has a�£AAov with the same mg.; Latte reJects It as a corruptlon
of the other form, which is unnecessary.
ETYM See Fur.: 167. If the variation �O / � is real, it is a Pre-Greek word, which is
likely anyway for a form with �o.

li�SlJpa [f.] . Kat a�8La � EluAauua (EM 3, 8). � PG?�
.ETYM Fur.: 309 connects it with the TN 'A�oT]pa.
a�SlJ<; [m., f.] . flUUTl� nap' 'InnwVaKTL 'whip (Hippon.) (H.). �PG(v) �
ETYM Probably a foreign word in Hipponax; see O. Masson 1962: 170 (fr. 130). Fur.:
388 convincingly compares U�PlUT�V' fluuTlya (H.). The word is Pre-Greek because
of the cluster �o (see ibid.: 318), with -�p- perhaps representing earlier -�O-.

li�£u; . EX£l<; 'vipers; you have' (H.). � ?�
. ,
ETYM Perhaps Illyrian, with � from IE *gwh, as m vl�a· Xlova. It IS unlIkely, ho\Vever,
that EXl<; should be reconstructed as *h,egwh-i-. Maybe the gloss is just Lat. habes (see
Pisani Paideia 10 (1955): 279).

li�£Uov => u�aAT], a�OeAAov.
a�£AL£po<; [adj.] 'simple, stupid' (Ar.). � ?�
.ETYM Wackernagel GGN (1902): 745ff. connected it with �£h£po<;, assuming that an
original mg. 'morally good' developed into 'too good, simple', with an unclear prefix.
This is unlikely; see Osthoff MU 6 (1910): 177 and Hatzidakis Glotta 11 (1921): 175f for
different analyses.
a�£p�lJAOV [adj.] . noAU, EnaxEl£<;, fl£ya, �apu, UXUlUTOV, flUTmov 'much, heavy or
burdensome, great, empty, rash' (H.). � PG(v) �
.VAR U�U�T]AOV (H.) is glossed in the same way; also -T]TO<; (EM).

.ETYM The variation points to a Pre-Greek word. Fur.: 374 compares MoGr. �Up�T]AO
'abundance' .

a��p [m.!f. ] ? . o'LKT]fla aToa<; EXOV, Taflelov. AUKWV£<; 'house provided with store­
houses, treasury (Lacon.)' (H.). � ?�
.ETYM Unknown. Not identical with � u�p 'air' (as in Frisk, who compares MoSw.
vind 1. 'wind' 2. 'bottom') .
a�[UlOv => ufllAAaKav.
a�lv [ace.m.!f.] . EAUTT]V, ol 8£ n£uKT]v 'silver fir; pine' (H.). � LW?�
.ETYM Comparing Lat. abies, Mayer KZ 66 (1939): 96f. assumed that PIE *ab- 'tree'
occurs in several Illyrian and Iranian names, such as 'A�m, 'A�pOl, A�lK� = 'YAata (St.
Byz.). If this is the case, is the root from a non-IE language in Europe?
a�moplov [n.] 'latrine' (IGR I, 599, Istropolis, Scythia Minor). � LW Lat.�
.ETYM From an unknown Lat. *abitorium 'latrine', in turn from abire (DELG Supp.):
a case in which a Latin word is known only from Greek.
a�AaMw<; [adv.] . �o£w<; 'sweet' (H.). � ?�
.ETYM A connection with *�Aa8U<; 'powerless' leaves the initial vowel unexplained
(see on � �AaOei<;); the semantic development is also problematic. Connection with
� uflaAouvw 'to make weak' would point to *h2mld- (with -aA- analogically after the
full grade *Ufl£AO-), but in this case one would expect *ufl�Aao- for our word; there is
yet no etymology.
a�AaPOl [m.] . �uAa 'wood' (H.). � PG(v) �
.ETYM Fur.: 370 compares �oapol' opu£<;, o£vopa 'trees' (H.); the word is Pre-Greek
(with prothetic vowel and �o alternating with �A). It is less probable that � �upu£<;
also belongs here .
a�AlJxp6<; => �AT]XP0<;.
a�Aon£<; [adj.] . u�Aa�£<;. Kp�T£<; 'sound, undamaged (Cret.), (H.). => �Aumw.
a�oA£i<; [m.!f.] . n£pl�OAat uno LlK£AWV 'garments (Sicilian) (H.) .
.ETYM Probably related to � U�OAAT]<;.
a�oA£w [v.] 'meet' (A. R., Call.), = UVTL�OA£W. � GR?�
.VAR u�oA�um· unavT�um 'to encounter' (H.).
.DER U�OAT]TU<; 'meeting' (H.).
.ETYM Cf. ��OAOV �flap' KaEl' 0 unaVTWUlV ei<; TauTov, � £uKmpov, l£pov (H.), which
may have its length for metrical reasons, or alternatively have arisen by
decomposition; cf. � en��oAo<;. The U- could be copulative (see � a- 1).
a�6UlJ<; [m.] 'kind of coat' (imperial period). � LW�
VAR Also u�oAAa (Peripl. M. Rubr.).
.ETYM LW from Lat. abolla (Varro). Cf. � U�OAel<;.

a�pa!l[<; a fish, a kind of mullet (Opp.). � LW? Eg.�

VAR Also a�paf.LLC;, a�pa�LC; (PLond. ined., IIIP).
ETYM The fish was salted in Egypt (Ath. 7, 312b) . Fur.: 220 thinks that the form with
� is due to a recent assimilation. It is either Pre-Greek, or a loan from Egyptian.



a�plo"nlv => a�8TjC;.
a�p6c; [adj.] 'graceful, delicate, pretty' (Hes.); mostly of young girls and women. "!l ?�
.VAR Fern. a�pa 'favorite slave' (not a Semitic loan, E. Masson 1967: 98) .
.DER o.�p6TTjC; 'splendor, luxury', o.�pomJvTj 'id.'; denominative o.�PUVOflaL [v.] 'to live
a delicate life', act. 'to treat soft-heartedly'.
ETYM No etymology. Not related to ��Tj 'youthful power', which has Tj- < *e « *eh,).
On the feminine substantive, see Francis Glotta 53 (1975) : 43-66.

a�pOnl�(() [v.] 'to miss'. "!l GR�
.VAR Only in a�pOTu�ofl£v [aor.subj.] (K 65) .
.DER a�p6m�lC; (H., Eust.).
.ETYM Perhaps formed on the basis of �fl�POTOV, the aorist of � o.flapTuvw. Could it
be an artificial archaism of the Doloneia (DELG)? On -�p- instead of -fl�P- ' see
Schwyzer: 277. Is it metrically conditioned?
a�p6TovOV [n.] 'wormwood' (Thphr.). "!l ?�
.VAR Also 0.-.
.ETYM Unknown; probably a loanword, perhaps from the Greek substrate.
Connected with o.�p6c; by folk etymology. From Akk. (a)murdennu 'flower with
thorns', ace. to Bailey TPS 1955: 82.
a�pvvu [] 'mulberries' (Parth. apud Ath.). "!l ?�
.VAR Also a-.
ETYM Unknown.

a�pvTO[ [m.] . £XLVWV SaAaaaLwv d80c; 'kind of sea urchin' (H.). "!l PG(v)�
.VAR Also afl�puTTOl' d80c; £XLVWV SaAaaaLwv'id. ' , �pUTTOC; (Ar.), �puaaoc; (Arist.).
.ETYM The formal variation (prothetic vowel and prenasalization) is typical of Pre­
Greek substrate words.
a�vc')6v [adj.] . �aSu 'deep' (H.). "!l PG?(v)�
.ETYM von Blumenthal IF 49 (1931) : 175 considers it to have an Illyrian ongm
("bottomless", related to �uS6C;). If the connection with �uS6C; is correct, it is not
necessarily Illyrian, but could also be a substrate word with prothetic a-. See � �uS6C;.
a�up�'lAOC; => a�Ep�TjAOC;.
a�vpTaK'l [f.] a sauce of leek, cress and pomegranate seeds (Pherecr.). "!l PG?(v)�
.ETYM Defined as uTC6Tplflfla �ap�aplK6v 'a foreign dish' (Suid.). Theopompus
wrote: ��£l 8£ M�8wv yaiav, EvSa TCOl£iTm a�upTuKTj 'he will arrive in Media, where
the a. is made' (Suid. 17 Kock), so perhaps it is an Iranian loan. But the structure a­
�UpT-UK-Tj is reminiscent of Pre-Greek words; for the suffIx, cf. �aTlUKTj, KauVuKTj,
maTuKTj. See also Fur.: 15864•

a�vO'O'oc; => �uS6C; .
ayu- intensifying prefix, e.g. aya-KA£�C; 'of great renown'. "!l IE *meg-h2- 'great'�
.DER Verbs: ayaflaL [v.] 'to admire, envy', ayaoflaL (Hes.), ayaLoflaL (Od.), ayu�w 'to
have too much' (A. Supp. 1061) . Nominal derivatives: ayTj 'admiration, envy' (11.),
ayuaflam (S. Fr. 885) , aya(a)aLC; (H., EM). See also � ayav.
.ETYM The same stem as flEya-, continuing a PIE zero grade *ytlgh2-. It has a
counterpart in Av. as-, e.g. as-aojah- 'with great strength' (from a zero grade *mgs-);
see Schindler 1987: 345. See � ayuHoflaL, � ayav, � ayavaKTEW, � ayau6c;, � flEyac;.
ayu8[c; 1, -[c')oc; [f.] 'ball of thread, clew' (Pherecyd.). "!l ?�
.ETYM No etymology. Under ayaS6c;, Frisk connects Skt. gadhya- 'was festzuhalten
ist'. Cf. � ayaSLC; 2.
ayu8[c; 2, -[c')oc; = aTjaaflLC; H.; aTjaaflLC; = aTjaafl� 'a mixture of sesame seeds, roasted
and pounded with honey', an Athenian delicacy given to guests at a wedding. Note
the expression ayaSwv ayaSL8£c; 'quantities of goods'. "!l PG?(v)�
.ETYM Belardi Ric. ling. 4 (1958) : 196 compared yUSla. aAAaVTla 'sausages (vel sim.)'
(H.); see � aHac;. If this is correct, the word is Pre-Greek, because of the prothetic
vowel. Fur.: 370 also compares � y�Suov.
ayu86c; [adj.] 'good, fit, noble' (11.). "!l IE?, LW?�
.VAR aKaS6v ayaS6v 'good' (H.); XUaLOC;' XpTjaT6C; 'good, useful' (H.). Dor. xU'(OC;
'noble, good' (long a) .
DIAL Cypr. a�aSoc; must probably be read ayaSoc;, see Egetmeyer Kadmos 32 (1993) :
145-155 ·
.ETYM The older comparison with Gm. forms like Go. gaps, MoHG gut, MLG gaden
'to fit' etc., and Slavic words like OCS godbn'b 'pleasant', goditi 'be pleasant' and Ru.
g6dnyj 'useful' should be forgotten, as these require a root *i'edh- (LIV2 s.v.), from
which Skt. gadh- 'to take, seize' derives. ayaS6c; is considered to be from a European
substrate by Beekes KZ 109 (1996) . An Indo-European attempt by Pinault MSS 38
(1979) : 165-170, who derives the word from *mgh2-dhh,-o- 'made great', or 'whose
deeds are great' (Ruijgh 1991b) . I find this semantic development difficult. Moreover,
as Pinault admits, a suffIx -dho- is rare (see Chantraine 1933: 366) . Finally, there are
forms like aKaS6v and XUaLOC;; if these variants are reliable, the word could be Pre­

ayulOC; [adj.] epithet of a sacrificial calf in the Labyadai inscr. (Schwyzer: 323) . "!l ?�
.ETYM Unknown. Connected with � aya-, based on comparison with ayaiov·
btl<pSovov 'liable to envy' (H.); see Buck 1955: 245.
ayuU[c; [f.] 'dwarf iris, Iris attica' (h. Dem.). "!l ?�
.VAR Also mse. (H.). On 74, 31 see DELG.
.DER ayaHLC;' UCtKlVSOC; � SpuaHLC;, � avayaHLC; 'hyacinth, plantain, pimpernel'
.ETYM No etymology; see DELG, Andre 1956 s.v. anagallis and Stromberg 1940: 78.
ayaUoflal [v.] 'to be proud, exult in' (11.). "!l ?�

aY UAoxoV


.DER ayaAfla 'glory, delight, honor; statue'. In later language, ayuUoflaL is replaced
by ayaUluo flaL, -lUW after the verbs in - lUW; thence ayaUlacn<;, -lafla. ayuUlo<;'
AOloo po <; 'slandering' (H.), ayaUlu�o flaL AOlOOpciG9aL, TapavTlvOl 'to slander
(Tarent.) (H.); Fur.: 370 compares yapplwfle9a, but there is no support for this. The
plant name ayaUI<; (h. Cer., Nic.) probably does not belong to ayuUOflaL.
.ETYM Looks like a denominative of *ayaA6<; , but such a form is unknown.

ayuAoxoV [n.] 'eagle-wood, Aquilaria malacensis' (Dse.). � LW�
ETYM One suspects an Oriental loanword; cf. Schrader-Nehring 1917: 39f. On
Pahlavi 'wlwg < *agaluk, see Henning BSOAS 11 (1943-1946): 728.

aya�aL - aya- .
Aya�E�vwv [m.] the Greek commander before Troy (11.). � GR�
·VAR Att. vases AyaflEGflwv, also AyaflEflflwv, -flEV(V)WV (Nachmanson Glotta 4
(1913): 246) .
ETYM Since Prellwitz BB 17 (1891): 171f., a pre-form *Aya-flEo-flwv has been
assumed, with the root of flEOOflaL. The development -ofl- > -vfl- > -flv- is known in
various Greek dialects (other examples in Lejeune 1972: 771, where also on the
development to -Gfl-).
Kretschmer Glotta 3 (1910-1912): 330£ connected the second part with flEVO<; and
flEV£lV (which von Kamptz 1982: 181 and 209 finds improbable), explaining -Gfl- as a
kind of popular assimilation.

ayav [adv.] 'much, too much' (PL). � IE *mg�eh2-m�
.DIAL Aeolic or Doric in origin, which explains the long a.
.ETYM The old accusative of the adjective � flEya<;. The form is important, as it points
to the type nom. *CeC-C, acc. *CC-eC-m, which I assume is the original
hysterodynamic inflection in Indo-European (see Beekes 1985: 103f.). Cf. also � aya-.
ayavaKTEw [v.] 'be indignant or irritated' (Hp.). � GR?�
.ETYM Expressive formation in -aK'rEW like uAaKTEw (to UAUW), so from original
*ayavuw? Cf. ayuvTjflaL' aGXuUw, ayavaKTw (H.). Pinault RPh. 65 (1991 [1993]):
196-198 derives it from *aya vaKTo<; 'pressed too much', from � VUGGw. Uncertain.

ayuvva - axuvTj.
ayavo<; [adj.] 'mild, gentle' (11.). � ?�
.ETYM No etymology. The connection with ayaflaL or yuvo<; [n.] 'splendor' (Bechtel
1914) is semantically unconvincing.
aya1tuw [v.] 'to receive with friendship; to like, love' (11.). � IE? �
.VAR Also aya1tu�w (11.). Retrograde ayu1tTj '(Christian) love' (late, especially LXX
and NT).
ETYM Pinault RPh. 65 (1991 [1993]): 199-216 assumes it derives from *aya-1t<l- 'to
protect greatly', referring to similar expressions in Sanskrit; c£ � efl1tu�o flal. The
Christian use may have been influenced by Hebr.'ahaba 'love'; see Ruijgh Lingua 25
(1970): 306.

ayyepuKo floV


ayaplKov [n.] name of several mushrooms (Dsc.). � GR�
.ETYM Perhaps from the TN Ayapla (Sarmatia)? See Andre 1956 s.v. agarieum and
Stromberg 1940: 122.
ayaO'uAA[<; [f.] plant that produces aflflwvlaKov, Fecula marmarica (Dsc.). � PG(v) �
.ETYM Fur.: 254 connects it with � yTj9uUI<; (Dor. ya-), which is Pre-Greek.
Comparing the latter with ayaGuUI<;, we note the prothetic vowel and the
interchange 9/0'. See Andre 1956 s.v. agasyllis.
ayauo<; [adj.] 'admirable, noble' (11.). � ?�
.ETYM Ace. to Schwyzer IF 30 (1912): 430ff., Aeolic = aya-F0<;; perhaps related to
� ayaflaL. An expressive gemination of the F has been proposed (e.g. Ruijgh Lingua
25 (1970)). But since a suffIx -yo- is doubtful in Greek (cf. Chantraine 1933: l24: "le
suffixe etait mort en grec"), the word may rather be Pre-Greek.
ayaupo<; [adj.] not quite certain, perhaps 'proud' (Hes., Hdt.; rare). � ?�
.ETYM Was � yaupo<; reshaped after ayauo<;? A pre-Greek origin should also be
ayyapo<; [m.] 'Persian mounted courier' (X.). � LW�
.VAR Rarely as an adjective, e.g. ayyapov 1tUP 'signal fire' (A. Ag. 282).
.DER ayyap� Lo <; = ayyapo <; (Hdt.), substantivized ayyap� Lov 'institution of the
ayyapOl' (Hdt. 8, 98, with a description of it). Denominative ayyapeuw [v.] 'press
into service' (Ev. Matt., pap., inscr.); thence ayyap£uT�<; 'impressed laborer' (pap.
VIP) and ayyapda 'service' (pap., inscr.), plur. ayyapciaL 'cursus publicus' (inscr.
IIIP); ayyaplKo<; (pap.). Hell. and late by-forms are eyyapeuw, -EW, -la, by folk­
etymology after the preverb ev-.
.ETYM The exact source is unknown. Not from Akk. agru 'hired man'; see Eilers Ilf 5
(1962): 225; Happ Glotta 40 (1962): 201. On the realia see Rostowzew Klio 6 (1906):
249ff. and R. Schmitt Glotta 49 (1971): 97-100 (who defends an Iranian origin).
Mancini Glotta 73 (1995): 2lO-222 reconstructs a form OP *angara- > *ayyapTj- as the
basis of ayyap�LOv (Hdt. 8, 98), which must be the oldest Greek form. Extensively on
this word Brust 2005: 17ff.


ayyEXo<; [m.] 'messenger' (11.). � LW�
.DIAL Perhaps Mye. a-ke-ro.
DER Denominative aYYEUw [v.] 'to convey a message'. Did ayyeAITj<; [m.]
'messenger' (Horn.) arise from a false interpretation of the genitive (T�<;) ayyeAITj<;?
See Leumann 1950: 168ff. Thence � ayyeAITj 'female messenger' (Hes. Th. 781). From
ayyeAla: ayyeAlwTTj<;, -WTl<; '(female) messenger' (h. Mere. 296); from ayyeAo<;:
ayyeAlKo<; 'of a messenger' (late); from aYYEUw: ayyeAfla 'announcement' (E., Th.),
ayyeATlKo<; 'premonitory' (late), ayyEAT£lpa 'female messenger' (Orph. H. 78, 3; not
.ETYM The connection with Skt. angiras-, name of mythical beings, has now been
abandoned. Perhaps an Oriental loan, like � ayyapo <;.






liyy01tTJVla [] . Ta nvv fleAlaawv K'lpla 'honeycombs' (H.). <!!! LW�
o ETYM From Iranian angupen, see Bailey BSOAS 20 (1957): 51 (does not belong to
� ayyoc;).
ayyo� [n.] 'vessel' (ll.). <!!! PG(v)�
oDIAL Probably Myc. a-ke-ha [pl.] .
oETYM Possibly a Mediterranean loanword (cf. Chantraine 1933: 418), as kitchen
utensils are often borrowed. Fur.: 275, 3072 compares ay8uc;· ayyoc; KP'lLlKOV, which
is quite tempting.
liyyoupa [f.] . pu�, O"La<puA� 'grape, bunch of grapes' (H.). <!!! ?�
oETYM Cf. MoGr. aywpoc;, ayoupoc; 'unripe, green, young man' and ayoupl8a 'unripe
grape', from a-wpoc; 'unripe, green', with the spirant as a transitional sound;
secondary nasalization as in MoGr. Cret. ayyoupoc; 'young, youngster' and ayyoupl
'gherkin'. The MGr. and MoGr. word was borrowed as MP angur 'grape' and Eg.­
Arab. aggur 'gherkin'. See Kretschmer Glotta 20 (1932): 239f.; also Cocco Arch. glott.

ital. 54 (1969): 98.

liyyoupo� 'cake'. =youpOC;.
aydpw [v.] 'to gather' (ll.). <!!! IE *h,ger- 'gather'�
o DIAL Myc. a-ke-re, a-ko-ra /agora/, a-ma-ko-to me-no /ham-agorto menos/ 'in the
month of the Assembly'? See Taillardat REGr. 97 (1984): 365-373.
oCOMP 6fl'lyep�c; 'gathered together' (ll.), ve<peA'lyepeTa 'cloud-gatherer' (ll.).
o DER � ayopu S.v.; ayopoc; 'gathering' (E.). Often ayup- (cf. Schwyzer: 351): ayuplC;
'gathering, mass' (ll.), with 1tav�yuplC; 'gathering of all' (Arc. 1tavuyopmc;,
1tavayopla); ayupT'lC; 'beggar', denominative ayupTu(W 'to beg', ayupTeuw (Str.) ,
whence ayupTeLa, adjective ayupLlK6C; (Str., PIu.). Also ayupT�p 'beggar', ayupTpla
'beggar-woman' (A. Ag. 1273). Also ayupfl6c; 'gathering' and ayupfla 'anything
collected'. The formations in ayep- kept the connection with the verb: ayepmc;
'gathering, inspection of the army' (Hdt.), ayepfl6c; 'gathering of funds, troops, etc.'
(inscr., Arist.), ayepfloaUv'l (Opp.), ayepTac; 'collector' (IG 14, 423: I 35 [Taurom.]).
ayap- is found in ayapplC; 'meeting' (IG 14, 759: 12 [Naples]); also ayopplC;' ayopu,
a9pOlmc; 'gathering' (H.) which may be Aeolic, see Chantraine 1933: 280.
oETYM No direct cognates, but the reconstruction *h,ger- is unproblematic. See
further � yepyepa . 1tOAAU 'lots, often' (H.) and Ta � yupyapa 'heaps, lots'.
�yepe90VTat and -TO have a present suffIx -9- (cf. Schwyzer: 703); �yepe90VTat (r
231) and �yepe9w9at (K 127, Aristarchus) have an unexpected long vowel; the forms
were built on frequent �yepe90vTo.
aytATJ [f.] 'herd, troop' (ll.). <!!! IE *h2eg- 'drive'�
oETYM From � ayw, with a suffIx *-1-. Comparison with Lat. agilis, Skt. ajirci- 'mobile,
quick' and Lat. agolum 'staff of a shepherd' makes little sense; the formations are
probably independent.
aytp6a = axep80c;.
ay£ppaKa�o� [m.] . O"La<puA� 'bunch of grapes' (H.). <!!! PG(v)�



oVAR ayyepuKoflov, O"La<puA�v 'id.'; aypuKa�oc;· aTa<puA� 'id.' (both H.).
oETYM Latte rejects two of these forms (how to decide which?), followed by Fur.: 221.
I think this is indefensible. The word is Pre-Greek in any case, like so many words
concerning wine (e.g. apaaxu8ec;). Note the element -a�-. The -e- is a prop vowel
(see Fur.: 378ff.); variation a/o and �/fl are also well-known in Pre-Greek (which
means all the glosses are real); the yy may represent a prenasalized consonant, one of
the clearest characteristics of Pre-Greek words.
aytpwxoc; [adj.] 'magnanimous, proud' (ll.). <!!! ?�
oDER ayepwxla 'magnanimity' (LXX, Plb.). Uncertain is Dor. yepwxla (Ar. Lys. 980).
oETYM No etymology.
aytTpla [f.] . flala. TapavTlvol 'midwife (Tarent.) (H.); avayeTpla· � TalC; TlKTOUaatC;
U1t'lpe-rouaa yuv� 1tapa TapavTlvOlC; 'woman who watches over the midwives in
Tarent.' (H.). <!!! ?�
oETYM McKenzie Class. Quart. 15 (1921): 48 assumes that it stands for earlier
*aypeTpla, from � aypew with a suffIx -Tpla and subsequent dissimilation. This
seems quite improbable semantically.
aYTJ = aya-.
ay�vwp [adj.] epithet of 9ufl6C; of uncertain mg.; perhaps 'proud' (ll.). <!!! IE *h2eg­
oVAR PN Ay�vwp.
oDER aY'lvopl'l (Hom.), whence aY'lvopew (Nonnos).
oETYM First part probably related to ayw (Risch IF 59 (1949): 39f), rather than to aya­
(Sommer 1948: 169f.); cf. also Kuiper MKNA W 14: 5 (1951): 5 : 207. On the evolution
of the meaning, see DELG.
ay�paTov [n.] a plant, 'Origanum onites' (Dsc.). <!!! ?�
oETYM Assuming an original meaning 'unaging', it can be connected to Y'lpuaKw,
y�pac;. Semantic parallels are given by Stromberg 1940: 103; these are uncompelling.
ay�paTo� [m.] stone used to polish women's shoes (Gal.). <!!! ?�
oETYM A connection with the word for 'age' does not seem appropriate. Szemerenyi
Gnomon 43 (1971): 641-75 proposes ay-�paToc; 'very lovely', which is not much
aY�Twp = �yeoflat.
aYlo� [adj.] 'holy' (Hdt.). <!!! IE *Hieh,g-�
oDER Yod-present a(oflat [v.] 'to honor' (ll.), from *aYloflat. Late nouns aYl6T'lC; and
aylwaUv'l (LXX), verbs ayl(w 'to consecrate, dedicate' (Pi., S.) and aYlu(w 'id.'
(LXX), whence aYlafl6c; 'offering to the dead' (D. S.), aYlaO"fl6c; 'consecration' (LXX,
NT), aylaafla 'id., sanctuary' (LXX); aylO"L�pLOV 'holy vessel' (Inscr. Perg. 255, 9),
aylaO"L�plOV 'sanctuary' (LXX) and aylO"LUC; 'ceremony' (Call.). aYlaTeuw [v.] 'be
holy, be pure; consecrate' (Pl., E.) also presupposes a nominal form in -aT- (aYlaT6c;


only in Et. Cud. S.V. uYL<JTela); thence UYl<JT£Ufla 'sanctuary' (Procop.) and uYL<JTela
'ritual, service' (Isoc.) .
ETYM The connection with Skt. yajati 'honor with sacrifice and prayer' is
semantically unobjectionable and formally explained by Lubotsky's rule (Lubotsky
MSS (1981): 133-8), i.e. in *Hieh,g-, the glottal element of the pre-glottalized *g was
lost before a consonant. This means that it is unnecessary to assume a-vocalism for
this root in PIE, as is done by many scholars (e.g. LlV2 s.v. *Hjag-). The primary
noun ayLO<; seems to have a suffIx *-iHo-. A different formation is found in � uyv6<;.

aYKuAo<; [adj.] 'curved, bent' (11.). <!l IE *h2enk- 'bend'�
.COMP ayKuAofl�Tll<; 'who thinks crooked thoughts' (11.), from an old verb *flll-flL (cf.
Skt. mati), see Ruijgh Lingua 25 (1970): 306.
DER Denominatives aYKuAAw [v.] 'to bend backwards' (Aret.) and aYKuA6w [v.] 'to
bend' (Ar.); from the latter aYKuAwfla 'loop' (Gal.), -W<JL<; medical term, name of
various lame or stiff conditions (Gal.); aYKuAll 'strap, loop, hook, hinge of a door,
etc.' (B., Hp., S., E.). Thence aYKuAllT6<; 'provided with aYKUAaL' (A.), aYKuAl<; [f.]
'hook' (Opp.), to which aYKuAuSWT6<; 'provided with a loop' (Hp. apud Gal.),
aYKuALOv 'loop, etc.' (medic.).
Many derivatives built on the root *h2enk- with a different suffIx:
a) With *-1-: aYKuAll [f.] 'curved arm, armful' (Archil.), mostly plur.; diminutive
aYKaAl<;, mostly plur. -loe<; (11.), which is more frequent in the epic for metrical
reasons. aYKaAov [] 'armful, sheaf (h. Mere. 82) is not entirely certain.
Denominative aYKaAl�OflaL [v.] 'to embrace' (Semon.), action nouns aYKuAL<Jfla
(Tim. Pers.), aYKaAL<Jfl6<; (pap.). A derivation in *-1- without suffIxal vowel is found
in aYKAOv, <JKOAL6v 'curved, bent, crooked' (H.), provided that the gloss is correct.
b) With *-n-: aYKwv, -wvo<; [m.] 'elbow' (11.), aYKU<JL (Opp.), said of many
protruding objects; cf. aYKu<; below. Late diminutives aYKwvLOv, -l<JKO<;, -l<JKLOV;
denominative aYKwvl�W [v.] 'to lean on the elbows' (Corn. Adesp., gloss.), whence
aYKwvL<Jfl6<; (Eust.). Further aYKOLVaL [] 'arms' (11.), enlargement with -la. An old
e-grade may be found in £Jt-llYKevlee<; 'long planks on a ship' (see Bechtel I914).
c) With *-s-: ayKo<; [n.] 'mountain glen' (11.), formally comparable with Skt. ankas­
[n.] 'curve'.
d) With *-tro-: aYKL<JTpOV 'hook' (Od.; from *aYKl�w), the formation of which is
unclear (see Chantraine 1933: 333f.); thence ayKl<JTpLOV, ayKL<JTpeUW, ayKL<JTp£UTLK6<;
and aYKL<JTpela (only marginal attestations).
Unclear is the formation of ayKu<;· aYKuAa<; 'armful, sheaf (H., see Bechtel 1914).
Adverbs aYKa8ev [adv.] 'in(to) the arms, on the elbows' (A.), ayKu<; 'into the arms'
(11., only antevocalic except in 'I' 711), probably the elided with a zero grade
suffIx belonging to aYKwv; thence aYKu�OflaL [v.] 'to lift up in the arms' (11.).
.ETYM All forms derive from a widespread lE root *h2enk-, found in e.g. Skt. aneati
'to bend, curve' and aeati 'id.'. Not connected with this group of words is Hitt.
bai(n)k-tta 'to bestow', med. 'to bow' (see Kloekhorst 2008). In Greek, we also find
� OYKO<; 'barb' continuing *h2onk-o-. For the stem in -u-, cf. Skt. anku-ra- 'hook'
(note the operation of Wheeler's Law in Greek). The I-suffix is found in OHG angul
'fishhook', ON 61 [f.] 'belt' (perhaps an old formation; cf. aYKuAll), 611 'germ' etc.


One generally connects it with � ayKupa 'anchor' (Ale.), but I suggest that this is a
substrate word.
aYKUATJ VAR ayKwv, ayKo<;, etc. => aYKuAo<;.

aYKupa [f.] 'anchor' (Ale.).
.DER Dervatives are scarce: ayKupwT6<; 'anchor-shaped' (Ph. Bel.), aYKupLOv (Ph.
Bel.), aYKupl�W [v.] 'to make sbd. stumble' (old com.).
.ETYM The suffix -upa is typically Pre-Greek, as in yecpupa, y6pyupa; it therefore
does not continue an inherited formation * -ur-ja-. Borrowed as Lat. ancora.
aYAa6e; [adj.] 'splendid, beautiful, famous' (11.); formulary epithet, e.g. of ui6<;. <!I ?�
VAR The Cretan and Cyprian gloss ayAaov· YAaqmpov 'hollow(ed), polished' are
due to misunderstanding of the Homeric usage, acc. to Leumann 1950: 272'8 .
.DER aYAa!a 'splendor, beauty' (11.), also PN; ayAal�w [v.] 'to embellish, glorify',
med. 'to take delight in'.
.ETYM Probably from *ayAaFo<;. Connected with yaA�vll ' ayuUoflaL (cf. Szemerenyi
1964: 155), or to � ayavo<;, � ayauo<;. The connection with ayuUoflaL enjoys a certain

*aYAle; => *aiyAl<;.
uYAle;, -t8oe; [f.] 'clove of garlic' (Ar.). <!I PG (v) �
.DER aYAl8Lov in: ayAloLa· <JKopooa 'garlic' (H.), with interchange 81 0 (Fur.: 194).
.ETYM Fur.: 127, 282 connects it with �yeAYL<;, -L80<;, -LOO<; as yeA-Y-: a-YA-, for which
cf. Kep-K-a: a-Kp-l<;. This seems quite possible.
aYAut:0"8m [v.] . �Au1tTw8aL 'to be hindered, damaged' (H.). <!I ?�
.ETYM von Blumenthal IF 49 (1931): 176 thinks it is Hyllaean or rather Messapian,
connecting it with Go. agls. Quite uncertain.
ayv6e; [adj.] '(ritually) pure, holy' (Od., mostly poet.). <!l IE *(H)ih,g-no- 'holy'�
.DER UYVOTll<; 'purity' (NT, etc.). Verbal derivations: 1. uyveuw [v.] 'to consider holy,
be pure, purify' (lA), whence uyvela 'purification', uyv£ufla, UyV£UT�pLO<;,
UYV£UTLKO<;; 2. ayvl�w [v.] 'to purify, consecrate' (poetic), whence uyvL<Jfla, -L<Jflo<;,
-L<JTLKO<;, etc., UYVlTll<; 'purifier' (Lyc.) after nouns in -lTll<;, cf. Redard 1949: 11.
ETYM Related to � UyLO<;; it may be the same formation as Skt. yajna- 'sacrifice'.

uyvoe; [f., m.] tree name: 'withy', 'Vitex agnus castus' (h. Merc.), = AUyO<;. The name
probably also denotes a fish and a bird, see DELG S.v. <!I EUR?, PG?�
VAR Also ayovo <;; cf. ayovov = flUP<JlVll aypla 'wild myrtle' (Fur.: 381).
.ETYM Comparable with OCS jagnfd'b 'black poplar' (Liden IF 18 (1905-1906): 506); if
this is correct, it is perhaps a European substrate word (on which phenomenon, see
Beekes 2000). On the folk-etymological connection with the notion of chastity
(UYVOTll<;), see Stromberg 1940: 154.

uyvur.u [v.] 'to break' (11.). <!l IE *ueh,g- 'break'�
VAR Fut. a�w, aor. £a�a or ��d, perf. £aya, pass. aor. UYllV or £&YllV (on verse-final
£&1'11 A 559 see Wackernagel l916: 141, Chantraine 1942: 18).



.COMP Mostly in KunlYVUf.ll, with a from -u-Fuy- (Bjorck 1950: 42, 147).
DER ay� 'fragment' (A., E.), long a- certainly in A. R. 1, 554; 4, 941; further twy� <
*FL-Fwy-� 'shelter', if originally 'breaking of the wind' (� 533), also in composition
e1tlwyul, -� 'place of shelter' (E 404), perhaps dissimilated from *bn-FLFwyul (but see
Bechtel I914). Further aYf.Lo<; 'fracture, cleft' (Hp., E.), aYf.Lu 'fragment' (late); a�o<; =
aYf.Lo<; (Crete), but hardly from the O'-aor.; appurtenance of the TN 'Oa�o<; (Hdt. 4,
154) as Fa�o<; is uncertain. Cf. also YUKTO<; ( F-} KACtO'f.lU 'fragment' (H.). On
FUYuvo- (Thespiae), see CBG 6, and Taillardat RPh. ser. 3: 40 (1966): 76. Also ayuvo<;
[adj.] 'broken' ( 231).
ETYM From *Fayvuf.ll (the F is clearly visible in Homer) < *uh,g-n(eu)-, belonging to
ToB wak- 'to go apart', caus. 'to split' and perhaps also to Hitt. l;!aV / l;!akk- 'to bite'
(cf. Kloekhorst 2008 s.v.). A palatovelar is best reconstructed based on Skt. vajra­
'thunderbolt' and its Indo-Iranian cognates. Greek -Fwy- is from *uoh,g-. Perhaps
Lat. vagina is also related; cf. MoHG Scheide 'id.' related to scheiden 'to separate'.

received a convincing explanation. At any rate, it did not serve to distinguish the
word from aylo<; (per DELG) .





ayvu<; [f.] 'weaving stones' (Piu.). � PG (s) �
.VAR -UeE<; [plo] .
ETYM See Chantraine 1933: 366. Probably a substrate word; suffrxes of the type -ueare typical of Pre-Greek.

ayopa [f.] 'gathering, assembly, market, trade, traffrc' (Hom.).
.DER ayopT]T�<; 'speaker' (epic), which perhaps rather derives from ayopaof.lUl
(Fraenkel 1910: 25f.). Denominative verbs: 1. ayopaof.lUl 'to speak (in public, in the
assembly) (epic Ion., poet.), only in isolated forms; ayopT]Tu<; 'eloquence' (epic),
ayopuTpO<; 'speaker' (inscr. Delphi, cf. Bechtel l921, 2: 151); 2. ayopEuw 'id.' (Hom.),
as a simplex rare in Attic (Wackernagel I916: 220ff., Fournier 1946: 41ff.), whence
ayopEUT�<; 'speaker', -T�PlOV 'podium', -m<; 'speech' (all rare and late); 3. ayopa(w 'to
be on the market, do shoppings' (lA); ilience ayopum<; 'purchase' (Plo), Boeot.
ayopaam<; (Holt 1941: 49f.), ayopaalu 'id.', ayopaaf.lo<; 'id.' (LXX),, usually
plur. ayopaO'f.luTu 'purchased wares' (D.); agent noun ayopaaT�<; 'purchaser' (X.),
fem. ayopamplu (pap.), ayopaaTlKo<; 'pertaining to trade' (Plo).
.ETYM Verbal noun related to � ayElpw.
ayo<; [n.] 'pollution, guilt; expiation' (Hdt., A., Th.); the word denotes the notion of
'sacredness' in ayw· TEf.lEVT] 'consecrated piece of land' (H.), for which Bechtel l921,
1: 115 suggests Lesbian origin, and also in S. Ant. 775, A. Ch. 155, etc. � IE? *(H)ieh,g­
.COMP ev-uy�<; 'under a curse or pollution' (Hdt., S.), whence evuyl(w [v.] 'to
sacrifice to the dead', evuYlO'flO<;, -lO'f.lU. Rare and late evaYlo<; [adj.] (after ayLO<;),
evuylKo<;. The opposite EU-Uy�<; 'immaculate' (Parm., S.) is found as EuhuYT]<; (lG
l2(9), 56 [Styra val); the simplex ay�<; (of the sun, Emp. 47) is clearly secondary.
.ETYM Formerly connected with Skt. agas- [n.] 'fault, sin', but the long vowel of
Sanskrit cannot be accounted for. DELG s.v. explains it as a psilotic form of *ayo<;
belonging with � ayLO<; (cf. Chantraine and Masson 1954: 85-107), which certainly fits
the attested meaning 'sacredness' well. As Chantraine remarks, all forms can easily
be derived from the root *hag- 'sacred', except for ayo<;, the psilosis of which has not


aYO<JT()<; [m.] in Homer only in the formula £AE yuluv ayomtp, which is usually taken
to mean 'with the hand (bent like a claw) (A 425). Taken by Hell. imitative poets (A.
R., Theoc.) as 'arm, elbow'. � ?�
-ETYM Solmsen 1909: Iff. proposed an original *ayop-O'TO<; "collector", derived from
� ayElpw 'to gather' wiili a suffrx * -st- found in semantically close TtUAUO'T� 'flat hand,
breadth of four fingers', Skt. hasta- 'hand', MoHG Faust 'fist', OCS gr'bstb 'handful'
etc. Not really convincing.
aypa [f.] 'hunting, way of catching; prey' (Od.). � ?, PG? (v) �
-COMP Instruments: Ttupaypu 'fire tongs' (11.), KpEaypu 'meat tongs' (Ar.); as medical
terms 600vTaypu 'tooth tongs'; diseases: Tt08aypu 'podagra', XElpaypu 'gout in the
hand'. Compounds in -aypETo<;: TtUAlvaypETo<; 'to be taken back' (epic since 11.),
uUTaypETo<; 'self-chosen' (Od.). �oayplu 'what was taken from a cow (= shield)',
av8payplov 'spoils of a slain enemy'. The interpretation of these compounds is
debated, see DELG.
-DER aypEu<; 'hunter' (Pi., A., etc.), more common is aypEuw [v.] 'to hunt' (Hdt., S.,
E., X.), whence aypEUT�<; 'hunter' (Sol., S. [lyr.l), aypEUT�p 'id.' (Theoc., Call.), 'catch, hunting net' (Sol., A., E., X.,); on the mg. of aypETT]<; see Redard
1949: 23658 • Further aypwO'O'w [v.] 'to catch' (Od.), cf. Schwyzer: 733; aypEw [v.] 'to
take, seize' (11., Sapph., ArchiL), in Hom. only ipv. aypEl, -TE (see Wackernagel l916:
166f.), Aeol. KUTaypEvToV [ipv.] ; aor. ptc. aYPEeEVTU, -TE<;, verbal adj. aypETul (Cos).
Agent noun aypEf.lwv (also -f.lwv) 'hunting spear, hunter', etc. (A., H., BM), whence
aypEf.lloV 'catch' (AP) .
-ETYM The relation between aypu and aypEw is unclear. Schwyzer: 727' pleads
against aypEw as a denominative from aypu. McKenzie Class. Quart. 15 (1921): 46f.
and 125 separated the two words: aypu and aypEu<; would belong to aypo<; 'field',
whereas aypEw would derive from the compounds in -aypETo<;, which themselves
belong to � ayElpw 'to gather'.
The existence of compounds like uUTaypETo<; beside uueulpETo<; could indicate that
UipEW and aypEw were associated, and this may explain formal variants like
-ulypETo<;. Connection with Indo-Iranian (Skt. ghase-ajra-, Av. vilhrkqm azrodaiolm,
both hapaxes of which the mg. is uncertain) and Celtic words (OW hair, MW aer
'battle' < *agra, OIr. ar [n.] 'defeat' < *agro-, Gaul.'EN Veragri) is rejected by DELG,
where it is remarked that none of these words bear the concrete meaning of
'catching' that is attested in Greek.
Fur.: (see index) thinks that aypEw is a substrate word because of the prenasalized
forms (Thess. UyypE-), the form with Ul for u (PN 'E�ulypETo<; on coins from Asia
Minor, on which see Vendryes 1938: 331-334; this form can hardly be reliable), the
variant eypEw, and the metathesized form UPYElTE. See � (WypEW.
aypaKa�o<; => ayEppaKu�o<;.
aypei<pva [f.] 'harrow' (AP 6, 297). � PG? (V) �


ayxoup0<; 1

o DER aYP[cpT] [f.] 'harrow' (Hdn., H.) .
ETYM One compares yplcpaoElat· YPUCP£lV. AUKWV£<;. ot O£ �U£lV KaL afluoo£lv 'to
write (Lacon.), others: to plane and scratch' (H.). The a- would remain unexplained.
But note that the attestations are very late. Pre-Greek origin with a prothetic vowel is
possible. The form in -va also suggests Pre-Greek, cf. Fur. 13265•

oETYM Formerly supposed to be the fem. of aypwoTT]<; 'countryman', from � aypo<;
(Bechtel 1914 S.v., Stromberg 1940: 117). However, Meier-Briigger KZ 103 (1990): 33f.
convincingly explained the word as *h2egro-h,d-ti- 'Feld-Futter', comparing � V�OTl<;
for the formation.



uYPTJv6v [n.] . <evoufla> OlKTUO£lO£<; 0 1t£PlT[El£VTat ot �aKx£uovT£<; L'HOVU041.
EpaTooEltvT]<; O£ mho KaA£l [yp�vuv] � y�vov 'garment like a net which those
possessed by Dionysus put on. Eratosthenes calls it a y.' (H.). -<! ?�
oVAR aYPT]va· OIKTua KaL eVOufla 'nets and clothing' (H.); cf. aYPT]vov 1tOlKIAov
EP£OUV OtKTuO£lO£<; KaL eVOufla O£ 1tOlOV (EM 14, 2).
oETYM Does the word have a prothetic vowel? Cf. also yp�vT]' avElT] GUflfllKTU 'mixed
flowers' (H.), cf. Stromberg 1944: IS. A derivation from aypa (DELG) is quite
uncertain. Nilsson 1941(1): 204 says that the net on the Omphalos was called
aYPT]vov; this statement is ascribed to Hesychius and Pollux (4, 116), but neither
author says so: it was only a guess by PW S.v. ("wohl auch"). In fact, Hesychius states
that it was called � yuyyaflov.

aYPTJOKETat 1tlKpalv£Tat 'is made bitter, tastes bitter' (H.). -<! ?�
oETYM Latte suggested that it stands for aYP[oK£Tat and derives from ayplo<;,
comparing aAEl[oKw I aAEl�oKw to � aAEla[vw. Semantically not convincing.

liypl1t1toe;; [m.] Laconian name for the wild olive (Zen.). -<! PG (v) �
oVAR ayplcpo<;· ytvo<; Tl aypla<; EAe la<; 'species of wild olive' (H.).
o ETYM Fur.: IS8 notes that these words have the variation 1t/cp, characteristic of Pre­
Greek words.
uyp6e;; [m.] 'field' (Il.). -<! IE *h2eg-ro- 'field'�
oVAR Mye. a-ko-ro lagros/; PN a-ko-ro-qo-ro IAgrokWolos/.
o COMP aypolKo<; 'who lives in the country' (aypo-FOlK-), � aypwoTl<;; also aypu1tvo<;
"who sleeps outside", which developed into 'sleepless, awake' (lA); cf. ayp-auAo<; (Il.)
'who has his bed/lair in the field'.
oDER Thence ayplo<; 'wild', with derivations: ayploTT]<; [f.] 'wildness' (PI., D., X.),
ayplooflat, ayplow, aypla[vw [v.] 'to become (make) wild'. aypoTT]<; [m.]
'countryman, rustic' (1t 218, E.), aypoT�p [m.] (E.) 'id.', also aypwTT]<; (E.) and
aypwoTT]<; (S., E.), of unclear formation (see Bechtel 1914 S.v. aypwoTl<;, but also
Meier-Briigger KZ 103 (1990 ) above).
aypOlwTT]<; (Il.) for aypwTT]<; would have arisen at verse end (Risch 1937: 32). On
ayptTT]<; see � aypa. Comp. aypoT£pO<; 'wild', cf. optaT£po<;. 'living in the mountains
(as opposed to the fields)'.
oETYM Old lE word, originally designating the uncultivated field: cf. Skt. ajra-, Lat.
ager, Go. akrs and Arm. art. Derivation from *h2eg- 'drive' is probable.
liypv1tvoe;; => aypo<;.
liypwoTle;;, -u50e;;, -EWe;; [f.] 'dog's-tooth grass' (Od.). -<! GR�


liyVla [f.] 'street, road' (Il.). -<! PG (s) �
ovAR Plur. aYUlaL
oDER AyUl£U<; [m.] 'guardian of the streets', epithet of Apollo (com., E.), whence the
month name Ayu[T]o<; (Argos); AyUlUTT]<; [m.] 'id.' (A.), also 'inhabitant of an a.'
(Pharsalos), cf. ayul�Tat· KWfl�Tat 'village dwellers' (H.); fem. aYUlaTl<; (Pi., E. [lyr.]).
oETYM Mainly a poetic word. Generally considered to be a perf. ptc. of � ayw 'to
drive' without reduplication, but this makes little sense as the formation is without a
parallel (save archaic iouia); more probably a substrate word in -Ula, for which cf.
� KWOUla, see Szemerenyi 1964: 203ff. and Beekes 1998: 2sf.
liYXL [adv., prep.] 'near' (Il.). -<! IE *h2em/- 'tie, betroth'�
oCOMP ayxt-flaxo<; (Il.) probably after TT]At-flaxo<; (only as a PN), see Triimpy 19S0:
oDER Further adverbial forms ayxo-Ell, -El£v; ayxou. Comp. 6.ooov, aooOTtpw, superl.
aYXloTa, -ov, also aomoTa, Elean amOTa (see Peters 1980a: 288). From the superl.
aYXloTivo<; 'near each other' (Il.), see Chantraine 1933: 204. On LOCf. aYXlOT£OUV =
ayXlaT�8av see Fraenkel Glotta 20 (1932): 84f. More forms in DELG.
oETYM Considered to be the locative of a root noun related to � ayxw (Schwyzer:
622), or a direct derivation from � ayxw after 1ttpl, aVTl. West Glotta 77 (1999): 118f.
suggests reconsidering the reading aYXT]aTiVOl, -at for aYXlaTivOl, -at, which is a v.l.
at all Homeric places, and which he interprets as liYXl + to-Tl- 'a sitting close
together'. More probable is the suggestion by Watkins (apud West Le.) that the
compound contains the root �o- 'to sit'.
UYXlAW'" [f.] 'swelling which obstructs the lacrymal duct' (Gal. 19, 438). -<! PG (V) �
oVAR Synonym aiylAw", (Cels.).
oETYM Galen analyzes it as � ayx( and � w"'. Stromberg 1944: 9Sf. follows this, and
explains the -A- from the synonym aiyIAw",. Frisk suggests that the first member is
from � ayxw instead. Not very convincing. The synonym points to a Pre-Greek
origin, due to variation a/at and y/X and prenasalization. Influence of aYXll ayxw on
aiY[Aw", is improbable. Note that at before Ne is not tolerated in Greek; perhaps the
first i derives from a palatalized Ig' I. The analysis in terms of Pre-Greek is *a(n)g'-il­

(jp ..
uYX6vTJ [f.] flavOpayopa 'mandrake' (Ps.-Dsc.). -<! ?�
oETYM Unknown.

liyxovpoe;; 1 [m.] 'gold' (AP, PIu.), presumably the name of the son of Midas (PIu. 2,
30M.). -<! ?�
oETYM Fur.: 391 compares TUYXOUp0<; yap 6 Xpuoo<;, � AE�l<; n£pmK� 'T. means gold
in Persian' (sch. Theoe., p : 3S1 W.) and Tuyxapa<; 'gold' (Cosmas ad OGI 199). If the


ayxoupo<; 2

word is Pre-Greek, I propose a sequence -arw- giving either -ap- or -oup-; see
� ayxoupo<; 2 and, most notably, � apaaxuoe<;.
ayxoupo<; 2 [m.] 'dawn' (Call.). <!I ?�
VAR ayxoupo<;· 6pepo<;. KlmpLOl. � <pwa<popo<; Kat oi auv aunp 'dawn (Cypr.);
bringer of light, and what comes with him' (H.). Variant ayxaupo<; 'near the
morning' (AP 4, m) ; 'time near dawn' (Call.), see LST Supp.
.ETYM Unknown. The interchange ou/au is reliable, as both occur in Callimachus (so
if au is explained from aupLov, we cannot understand -oup-). Perhaps PG -arW­
yielded variants in -aup- (with anticipation of the labial element) and oup- (with
additional coloring of the vowel); cf. � apaaxuoe<;.

ayxouaa [f.] a plant, 'Anchusa tinctoria' (Thphr., Dsc.). <!I PG(v)�
.VAR Also eyxouaa (Ar.).
·COMP Also KaTuyxouaa (Ps.-Dse.); cf. 'l'euouyxouaa (Plin.).
.ETYM The variant eyxouaa excludes derivation from � ayxw (which is defended by
Stromberg 1940: 64). A typical substrate word; see Fur.: 346 and 19755 on the suffIx
-ouaa. Not related to � ayxuvw'l'.
ayxpav [adj.] . fluwJta. AOKpO[ 'short-sighted (Locrian)' (H.). <!I PG(v)�
.ETYM Bechtel 1921 compares aKapov· TU<pAOV 'blind' (H.). These forms show typical
variations of Pre-Greek: K/X and prenasalization (see Fur.: 127).
ayxw [v.] 'to squeeze, strangle' (ll.). <!l IE *h2emft - 'narrow, strangle'�
DER ayxovTj 'hanging, strangling', for the suffIx cf. � JtepovTj, etc.; Lat. LW angina
(Leumann Sprache 1 (1949): 205, but see E-M s.v.). Thence ayxovLo<; 'fit for hanging'
(E., Nonn.), ayxovuw 'to strangle' (Man.). ayKT�p, -�po<; [m.] 'tool for sewing up
wounds' (Cels. Med., PIu.).
.ETYM An exact correspondence for the thematic present is found in Lat. ango 'to
bind together, strangle'. Hitt. bamanV, bame/ink- has a nasal present *h2m- (e) n -ft -.
The widespread u-stem adjective *h2emft-u- in Skt. ayt1hu- 'narrow', Go. aggwus,
Arm. anju-k, OCS PZb-k'b is not found in Greek. See � ayxL.

ayw [v.] 'to drive, lead, bring, carry; to draw, ete.' (ll.). <!l IE *h2eg- 'drive, lead'�
·VAR Aor. �yayov, perf. �xa (Att.); Dor. ay�yoxa, whence ay�oxa, ayewxa.
.DIAL Myc. a-ke /agei/.
.COMP With aJt-, eia-, £�-, KaT-, ete.; aTpaTTjy0<; 'leader (of the army), commander'.
On -ayeTTj<; in compounds (apXTjyeTTj<;, ete.) see Fraenkel l91O: 59ff., Sommer 1950:
DER ayo<; 'leader' (poet. since ll.), which is formally identical with Skt. aja- 'driver',
but perhaps a parallel formation; ay� 'transport' (Chios), 'winding' (Arat.); aywv,
-wvo<; [m.] 'gathering, rally (to see games)' (ll.), whence aywvLO<;, aywv[a, aywvLuw,
aywvLuTTj<;; verb aywv[�OflaL 'to contend for a prize, etc.', whence aywVLaL<;,
aywvLafla, aywVLaT�<;, aywvLaTLKo<;, ete. Further aKTwp, -opo<; 'leader' (A.), also a PN
(ll.), but Lat. actor may be an independent formation; aYfla· KAeflfla 'theft' (H.).
Reduplicated nouns: aywyo<; [m.] 'leader' (lA), aywy� 'carrying away' (lA),
formation unclear, whence aywyeu<;, aywYLflo<;, aywyLOv, aywyalo<;, aywyLKu.

wyavov 'spoke' seems unrelated (in spite of Frisk 1938: 17f.). Unclear is the formation
of aYlveflevaL, aYLvew 'to lead, carry' (ll.), and the relation with Dor. Aetol. ayvew
Cf. further on � ayeATj, � a�Lo<;, � a�wv, � oyflo<;; also � aypo<;. Not here � ayULa.
oETYM Old thematic present, also found in Skt. ajati, Av. azaiti, Arm. acem, Lat. ago,
0Ir. -aig, ON aka and ToAB ak-, all 'to drive, lead' vel sim. Originally the verb was
present (see LIV2 s.v. *h2eg-) . � oYfl0<; 'furrow, ete.' derives from *h2og-mo-, neatly
corresponding to Skt. ajma- 'trajectory', and proves that the regular reflex of *h2o in
Greek was 0, not a (as was contended by Ruijgh).
a�aY!16<; [m.] . KVTjaflo<; 'itch' (H.). <!I PG(v)�
.VAR Also in S. Tr. 770 acc. to Phot. (codd. 68aYflo<;).
.DER Cf. aoaKTw· KV�eOflaL 'to itch', aoa��aaL· KV�aaL 'to scrape, scratch', aoaXq.·
KVq., KV�e£L Ke<paA�v. 'l'TjAa<pq. 'scratches the head, gropes' (H.).
.ETYM The old explanation of aoaYflo<; as assimilated from 68a- is due to the desire
to reduce everything to Indo-European roots; see Van Beek fthc.b. There is no
compelling reason to connect a word for 'scratch' with 'tooth'; the vocalic variation
points to PG origin. See � 68u�.
Maq<;, -t<; => oa�vaL.
a�aA6<; [m.] . aa�oAo<; 'soot' (H.). <!I LW Maced.�
.ETYM Macedonian for a'i8aAo<;; see von Blumenthal l930: 5. See � ao�, � aOPaLa.
aM!1a<;, -avTo<; [m.] a strong metal, 'steel' (Hes.); 'diamond' (Thphr.). <!I ?�
.VAR PN Aoufla<; (Horn.).
.DER aoafluvTLvO<; (Pi., A.).
.ETYM Both the appellative and the PN are often derived from � ouflvTjflL as
'indomitable'; for the formation, cf. � aKufla<; and Chantraine 1933: 269. But
semantically, the etymology is rather strange. It is rather a loanword that was
adapted by folk etymology. Ace. to Troxler 1964: 19-21 and Barb 1969: 66-82, it is a
loan from Semitic, comparing Akk. adamu. Lubotsky 1998: 4143 refers to NPhr.
(a)TWfla 'stone'.
aMpE�a . e[p�VTj 'peace' (H.). <!I ?�
.ETYM von Blumenthal l930: 24 compares aTapa�[a and thinks the word is of Illyrian
origin. Very uncertain.
aMpKTJ [f.] 'salt deposit on the herbage of marshes' (Dse., Gal.). <!I LW Celt.�
.VAR Also -Tj<; [m.] , aoapKo<; [m.] .
.ETYM Like Lat. adarca (Plin.), a loanword from Gaulish: cf. Ir. adarc 'horn', which is
from Basque adar 'horn', with a Celt. suffix -k- (Pokorny Zeitschr. j celt. Phi!. 14
(1923): 273; Pokorny Zeitschr. j celt. Phi!. 16 (1927): 1l2).
aMauov => a�w l.
aMl�, -lX0<; [f.] a measure offour choinikes (Ar. fr. 709). <!I ?�




.ETYM Unknown. The suffix is also found in XOtVl� (which has -lK-). Geminated 80 is
rare in lA, and names for measures are often borrowed. Cf. Kao8txo<.; s.v. � Kaoo<.;,
which is combined by Fur.: 13059, who assumes a substrate word with the alternation
K/zero. Cf. also Szemerenyi 1969b: 248.
Meahwhate [3sg.opt.aor.] unknown (inscr. Elis [approx. 350"]). <! ?�
.ETYM DELG (see s.v. for more details) thinks of OEATO<'; 'writing tablet', referring to
Buck 1955: 263·
MeAcpeO<.; [m.] 'brother' (11.). <! GR�
.VAR Att. aoEACPO<'; (shortened form); aOEAcpE� 'sister' (Pi. -Ea, Att. -cp�). Cret.
.DER a8£AcploEO<';, -OE'1 (Att. -oou<,;, -o�) 'nephew', 'niece'. aOEAlcp�p· aOEAcpEo<,;,
AaKwvE<.; 'brother (Lacon.) (H.) is due to contamination with cppaT'1p. Diminutive
aoEACP(OLOV (Ar.), aoEACPlKO<'; 'brotherly, etc.' (Arist.), aOEAcpoT'1<'; 'brotherhood'
(LXX), aoEACP( W [v.] 'to accept as a brother' (Hecat.), aoEACPl�l<'; (Hp.).
.ETYM Derived from a word for 'womb' with copulative a < *s1]1-, meaning 'from the
same womb'; cf. a8£Acpo(· 01 eK T�<'; U1h�<.; OEACPUO<'; yEyovOTE<';. OEACPU<'; yap � fl�Tpa
(H.). However, the -E- cannot be from -EF- because of Cret. -LO<';; acc. to Wackernagel
1916: 52f., it must derive from * -eio-, which forms material adjectives. We may also
assume a ntr. *OEACPO<'; as a base noun, thus *ha-gWelpheh-o-. Att. aoEACPO<'; is from
contracted forms like a8£Acpou < -EaU. The inherited word for 'brother', cppaT'1P, was
primarily used in a religious or political sense (e.g. cppaTpa, cppmp(a), and could
perhaps also be used for other members of the extended family, like nephews. Greek
probably introduced the expression *cppaT'1p aOEAcpEo<,; 'brother proper', cf. Skt.
sagarbhya- and, within Greek, OflOYU<HpLO<.;. It has been suggested that the word
derived from pre-Greek matrilinear societies (Kretschmer Glotta 2 (1910): 201ff.), but
the word may instead point to a society with concubines (naAAaKa(); see Gonda
Mnem. 15 (1962): 390-2. See � OEAcpU<,;.
MevKI\<.;, -e<.; [adj.] unknown, said of OAE8po<.;, nOTflo<,;, Cp�fll<'; (Od.). <! IE? *deuk­
.DIAL Perhaps Myc. de-u-ka-ri-jo lDeukalion/.
.ETYM Like IIoAv-oEuK'1<';, it presupposes a noun *OEUKO<'; [n.], for which no cognates
can be suggested. Not to Lat. duco 'to lead', etc., as per Lagercrantz KZ 35 (1899): 276.
Cf. OEUKEl· CPPOVT( El 'consider, ponder' (H.), evouKEw<.; 'careful'; aoWK�<'; would then
mean 'careless, thoughtless', which fits the meaning very well. In a sch. on A. R. 1,
1027, OEUKO<.; is glossed as yAEUKO<.;, which seems most improbable. Is it a mistake for
r�EYKOL? The name �wKaA(wv may derive from *AwKaA(wv; see Bechtel 1914 s.v.
� aowK�<';.
Mi1 · oupavo<.;. MaKE86vE<.; 'heaven (Maced.)' (H.). <! LW Mac.�
ETYM Identical with � ai8�p, which features the typical
Lautverschiebung (0 for Gr. 8, etc.).


lu5'1Kon<.; [] in Kaflanf' ao'1KOTE<'; �8£ Kat U7tV4J (K 98), mg. uncertain.
<! ?�

.ETYM Connected with � ao'1v or � �8U<.;, and aaoE1v (cf. � aaoa). For discussion, see
a611!1ovew [v.] 'to be restless, scared' (Hp., PI., X.). <! ?�
.VAR With short a- ( 16).
.DER ao'1flov(a (Epicur., PIu.), ao'1flOo1JV'1 (Democr., X.) .
ETYM An Ionic word. AlIen Class. Rev. 20 (1906): 5 connected it with oa�vm;
Debrunner 1937: 266 assumed a contraction from *aoa'1flovEw; semantically not
compelling. Not related to �8U<.;, nor to a'10�<'; (Leumann 1950: 30982).

a6l\v, -evo<.; [f.] 'gland' (Hp.). <! IE?, PG?�
.vAR Later [m.] .
ETYM One compares Lat. inguen, -inis [n.] 'groin, swelling on the groin' and Mole.
ekkr [m.] 'growth, tumor' < PGm. *enkwa- < *engWo-. However, since PIE had no
words beginning with a vowel, this would be reconstructed as *h,engW-, but a zero
grade *h,ngW-e- would give Gr. *ev8£- acc. to Rix' Law (*HRC- > Gr. alE/oRC). This
implies that the Greek word cannot be cognate with the Germanic one (the latter can
be cognate with the Latin word, of course, as *h,engW-, see Schrijver 1991: 58). Greek
ao�v thus remains isolated. Fur.: 172"8 suggests a substrate origin (words in -'1v). For
a recent challenge of the validity of Rix's Law, see Nikolaev 2005, and on the present
word, p. 5054•

MllV [adv.] 'to one's fill' (11.). <! IE *seh2- 'satiate'�
VAR In epic with psilosis. The a- may undergo metrical lengthening.
.DER aoo<.; [m., n.] 'satiety' (11.), a.8tvo<.; 'crowded, thronging, vehement, etc' (mainly
epic), � a.o po<.;. From *a.o'1-: aoato<.; 'leading to satiety, unpleasant' (Sophr., H.). Cf.
further the glosses � aaoa and aaoE1v (H.), and � aOflwA� and � aa'1.
.ETYM Old accusative of a noun supposed in the first member of a.o'1-cpayo<.; 'glutton'
(but what kind of compound is it?). Often connected with the root *seh2- 'to satisfy',
which is found in several Greek verbal forms: uflEvm (11.), aor. � aam, uaaa8m (epic)
'to become satiated' and � aaTO<';. a.o'1v was analyzed by Frisk as containing a stem
a.o- and connected with Arm. at-ok' 'full, ripe' (cf. � a.o po<.;); but, as Clackson 1994:
170f. remarks, a.o'1v itself may simply contain a suffix -0'1v. Still, this does not explain
the other derivations with -0-. Other languages have an enlargement in -t(i)-: Lat.
satis 'enough', Go. saps 'satiated', ga-sopjan, Lith. sotis 'satiaty' etc. (see � aam).

MiavTov [n.] name of a plant, 'Adiantum' (Thphr.). <! IE?, GR�
.VAR ao(avTo<.; [m.] .
.ETYM Explained as 'what cannot be irrigated' (� ola(vw); see Stri:imberg 1940: 74f.
u6lKll [f.] 'nettle', = aKaA�cp'1 (Ps.-Dsc. 4, 93). <! PG?�
.ETYM Cf. � £A(K'1 'willow'. The connection with OHG nazza, nezzila etc. as *7;ld-ika
(cf. Frisk) is most improbable; it is rather a substrate word.
MlVO<.; => /io'1v, a.opo<.;.
Ml<.; [?] . W<'; f\n(wv, a8pOQ l, � e<1xapa 'in masses, hearth' (H.). <! LW Mac.�

oETYM In the meaning aepOOl, probably a mistake for aAL<;; in the sense eoxapa,
Macedonian (= Lat. aedes), acc. to von Blumenthal IF 49 (1931): 179.
MIlWAt] [f.] . a1topla, oALywpla, ayvOla, �auXla 'difficulty, contempt or negligence,
ignorance, rest or quiet' (H.); cf. Hdn. Gr.1, 324, etc. -<! ?�
oVAR aOllwAla· � ayvOla 'ignorance' (Suid., Call. fr. 717 Pf.), aoIlOAl'l (EM). Further
aOllwA£lv = ayvoElv (H., EM 155, 33).
oDER aOllwAw, aK'l0LW 'am careless' (Suid.), aollwA£lv, ayvoElv � ayvwllov£lv �
aK'l0L<lV 'to be ignorant, act or treat unfairly, be careless' (EM). aOIlWAel' XWPl<;
06Aou 'without resource' (Suid.).
oETYM Frisk derived it from ao- (in � a0'lv) with "suffIxal" -llwA- (Frisk Eranos 41
(1943): 52), which is highly improbable. Acc. to Fur.: 2638, the word can hardly be
inherited. On the interchange 0ll/ Oil, see Schwyzer: 208 (unclear).
MIlWV£1:; [pl.] a sea-fish (Opp. Hal. 3, 371). -<! ?�
oVAR Also aollwE<;.
oETYM Unknown; see Thompson 1947.
a()VQV [adj.] . o.yvov. Kp�TE<; 'holy (Cret.) (H.). -<! ?�
oETYM Is it a hypercorrect form, caused by the development OV > yv? Or just a form
invented to explain ApLaov'l ? See Bechtel 1921(2): 777.
a()oX£crX'l1:; [m.] 'idle talker' (Ar.). -<! ?�
oVAR On a see Bjorck 1950: 142, 41.
oDER Later aooAwx0<;; thence aooAwXla, -tw, -LKO<;.
oETYM Perhaps from *aaoo-AtoX'l<; to aaOelv· 0XA£lv (H.), with AtoX'l as a second
member in the sense of 'conversation', and a first member *a-oFao- (see � o.voavw,
� �ou<;); cf. Schulze 1892: 452f. Quite uncertain. See DELG for more details.
Mpu[u · aiepla 'clear weather' (H.). -<! LW Mac.�
oETYM Probably Macedonian; see von Blumenthal 1930: 5.
a()pucpu�UI:; => aTpacpa�u<;.
a()ouO'lucrucrOat [v.aor.] 'to accept the membership of, cpuA�<; KaL oqllou KaL cppaTpla<;
(IG 2\ 553: 15, LSJ Supp.). -<! GR�
oDER aOOUaLov, apWTOV, oUllcpwVOV 'pleased, harmonious, agreed' (H.), aoouaLaoa­
IlEVOL' 0lloAoY'l0aIlEvOl 'who agreed' (H.).
oETYM Probably connected with aoo<; 'decree' < 'decision'.
Mpol:; [adj.] 'thick, strong, full-grown, ripe' (Hdt.). -<! IE *seh2- 'satiate'�
oDER aoPOT'l<; 'strength' (Hell.); on Horn. a(v)opoT�-ra [acc.] see � av�p. aopuvw [v.]
'to make ripe', med. 'to ripen', whence aoPUVaL<;; sometimes also aoptw, aopoOllaL.
Plant name aopw0'l<; (Stromberg 1940: 82).
oETYM From root 0.0- in � a0'lv, with a suffix -po-.
Mpuu . 1tAola 1l0vo�uAa, KU1tpLOl 'ships made out of one piece of wood (Cypr.)'.
AtyoVTaL oE KaL ol ev n:p apoTp4> oTuAOl 'the poles in the plough'. LLKEAoL OE aopua
AtyOUaL Ta Il�Aa 'apples (Sicilian)" 1tapa oE AHLKOl<; aKpoopua 'fruits grown on

aelPW 1


upper branches of trees (Att.) (H.). Also aopua· ol oTuAOl apoTpou, OL' tiJv °
la-ro�oEu<; o.PIlO�£laL 'by which the plough beam is fixed' (H.). -<! IE *dru- 'tree', PG�
oETYM In the first gloss, it seems to continue *a-opua '(consisting) of one Single tree',
a compound from 0.- < *SI]1- and � opu<;. This can also be assumed for the second,
agricultiral meaning. For the meaning 'one single' of 0.-, cf. IlWVUXE<; S.v. � ovu�. But
in the third meaning, 'apples', it is probably folk-etymological, as it is a variant of
� Ilaopuu, which is non-IE. Compounds with a second member -opu- are rare and
doubtful; see DELG S.v. opu<;.
A()WVLI:;, -L()OI:; [m.] theonym. -<! LW?�
oVAR Also AOWV, -0<;.
oETYM Supposed to be a loan from Semitic (Hebr. adon 'Lord'). But no cult
connected with this name is known in the Semitic world, nor a myth parallel to that
in Greece. See Burkert 1985: 176f., arguing against Kretschmer, e.g. Kretschmer
Glotta 7 (1916): 29ff. and Kretschmer Glotta 10 (1920): 235f. Fur.: 32821 points to the
proper name AOWV, an Armenian general and a Phrygian flute-player.
a£OXol:; [m.] 'burden, contest, prize of a contest' (ll.). On the mg. Triimpy 1950: 150f.
-<! IE? *h2uedh- 'contest'?�
oVAR Also -ov [n.] (ll.).
o DIAL Arc. [a] FEeAa (IG 5(2), 75), Att. contr. &eAo<;, -ov.
oDER ateALov 'id.' (epic), ateALO<; 'of the contest' (Thgn., Call.) , aeALO<; 'unhappy'
(Att.), aeALoT'l<;; a(E)eAtw, -EUW [v.] 'to contend for', a(E)eA'lTqp, -Tq<;, a(t)eA'llla,
-aL<;, -LLKO<;.
oETYM The original meaning probably was 'contest for a prize'. Not related to Skt.
vayati 'be tired' (as per Triimpy 1950: 150-151), which is from *h,ueh2-. The word
looks Indo-European; the form suggests a root reconstruction *h2uedh-.
ad()w [v.] 'to sing (the praises of)' (ll.). -<! IE? *h2ueid- 'sing'�
oDIAL Att. Mw.
oDER aOloq, <V0q 'song', whence aOlOLllo<;, <V0LKO<;. Agent noun aOl06<;, <v06<; 'singer'.
Thence, or from aOloq, the verb aOlOLaw (epic) = aElow; derived from <V0q: 'fhO£lov a
building in Athens for musical contests. Further aELolla, qOllu [n.] 'song' (lA),
<;tollaLLov (Pl. Corn.); <;tollo<; [m.] 'id.' (Pl. Corn.).
oETYM The ablaut suggests PIE *h2ueid-, but no cognate outside Greek is known.
Older speculations, now dated, are found in Frisk and DELG. Haroarson 1993b: 163
assumed the reconstruction *h2ue-(h2)ud-, with loss of the laryngeal (which seems
diffIcult) and dissimilation as in £EL1tOV < *h,e-ue-ukw-. See � auoq, � a'l0wv.
adpw 1 [v.] 'to raise' (ll.). -<! IE? *h2uer- 'raise'?�
oVAR awpTo 'hung' (ll.), which Tichy 1983: 364f. explains as an injunctive *aopTo <
PIE *h21d[-to .
DIAL Att. a'(pw, for which Frisk suggested a recent formation to the Att. fut. a pw <
*aEpw. Improbable is a root *S[-, as per Heubeck Orbis 13 (1964): 264-7. Taillardat
RPh. 57 (1983): 21-25 convincingly assumes a zero grade verb *awr-je/o- > a'(pw.


uelpw 2


.COMP uepal-ltooe<; 'lifting their feet' (11.); fleT�Opo<; '(sth.) in the air', Att. fleTEWpO<;,
Aeol. lteOCtopo<; (these are not from � u�p).
DER apat<; [f.] 'raising' (Arist.); � UpT�p .
ETYM Not from a.�p 'air' (which has long *a-). No cognates are known, but the form
requires the reconstruction *h2uer-. See also � UelpW 2; an important question is
whether these are originally the same verb. See � UpT�p, � alwpa; cf. also � apfla 2,
� �epEeOVTat.

.VAR a'LAlOl· auyyafl�pOl 'the husbands of two sisters' (H.); elAlove<; in Pollux 3, 32 (ot
O£ uoeA<pa<; y�flavTe<; 6floyafl�pOl � auyyafl�pOl � flCinov aUYK'lOeaTat Kat ltapa
TOL<; ltOl'lTaL<; elAlove<;), with metrical lengthening of *eAlove<; or *eAlove<;.
.ETYM a'LA lOl may be an itacistic notation for *eALOl (*EAlOl). The u- in aEAlOl is taken
as a 'copulative' a-. Cognate with ON svilar [] 'brothers-in-law, whose wives are
sisters', PIE *sue-lo-, a derivative in -1- of the reflexive pronoun *sue.




U£lpW 2 [v.] 'to bind together, join' (11.); mostly with auv-. Cf. �uvalPeTat· auva1tTeTat
'is attached' (H.). <{ IE? *h2uer-? 'bind'�
VAR Note awpTo 'was hanging'; ltap'lEpe'l 'was made to hang beside' IT 34l.
.DIAL Perhaps Myc. o-pa-wo-ta lop-awortal 'pieces of armor'?
.COMP Nominal stem -aop- 'band, link' in *TeTp(a)-aopo<;, TeTpaopo<;, contracted
TETPWpO<; 'which yokes four together' (Od.), whence TeTpaopla 'four-horse chariot'
(Pi.); from auvaelpw also auvaopo<;, auv�opo<; 'coupled together, spouse', Att.
auvwpl<;, -100<; [f.] 'two-horse team', to which belongs auvwplKeUeTat 'drives with a
team of horses' (Ar. Nub. 15), but *auVWPlKO<; is unattested; from auvwPl<; also
aUVWplaaT�<; 'driver of a auvwpl<;' (Luc.), which presupposes a verb *auvwpla�£lv.
Contrasted with auv�opo<; is ltap�opo<;, ltapaopo<; (11.) ,(horse) joined beside', also
'outstretched' and 'reckless' (see Leumann 1950: 222ff.); a verb ltapaelpw seems to
have existed beside auvaelpw, but it is attested only in a special use ltap'lEpe'l ot
Kap'l (IT 341) 'the head hung on one side', cf. Leumann loco cit. Isolated Ult�Opo<; 'far
away'. elt�opo<; 'suspended over', KaT�OpO<; 'hanging down', and KaTwpl<; 'band
hanging down' rather belong with fleT�Opo<;, fleTEWpO<;, see � UelpW l.
DER Action noun UOpT� *'attaching', 'sth. attached, bag for tying' (Men.), medic.
term referring to the bronchi and the hose-like aorta (Hp., Arist.). Agent or
instrument noun UOpT�P, -�po<; [m.] *'attacher', 'sword-belt' (Od.), with unclear 0vocalism; also Ct6PT'l<; (pap., H.) and uopTeu<; (H.). Denominative or deverbative ptc.
UOpT'leel<; 'hanged' (AP). aopTpa [] 'pulmonary lobe' (Hp.), after nouns in
-TPOV (Chantraine 1933: 331f.). See further � aop, -opo<; [n.] 'sword' (probably
unrelated); related is � upTaw 'to hang'.
.ETYM Solmsen 1901: 289ff. separated it from � uelpw 1 'to raise', but DELG tends to
consider the second as a specialization of the first (see extensive discussion in
DELG). The present entry is mostly connected with a root *uerH- 'to bind, hang on,
strick' in Balto-Slavic, e.g. Lith. virve 'string', OCS obora < *ob-vora 'string, twine',
LitlI. verti 'to pierce, string', etc., but the exact demarcation from the root *Huer- 'to
open' is unclear; perhaps it was originally one root. For Greek, the closest connection
is with Alb. vjerr 'to hang, suspend'. See � UpT�p, � upTaw, � aop.


UEK�AlO" [adj.] 'unbearable' (only L 77, epya). <{ GR�
.ETYM The form was based on the root *uek- 'to wish' in � eKwv, � EK'lAO<;; see
Philipp in LfgrE s.v.
CtEAlOl [] . ot uOeA<pa<; yuvaLKa<; eaX'lKoTe<; 'who have sisters as wives', i.e.
'brothers-in-law' (H.). <{ IE *sue-lo-�

aEUu [f.] 'storm wind' (11.). <{ IE? *h2uel-�
.VAR UEn'l (IT 374).
.DIAL Aeol. avena.
.DER Aenw, -ou<; [f.] name of a Harpy (Hes.); uenmo<;, also uena<; 'fast like a
storm' (S.), uen�£l<; (Nonn.), uenwo'l<; (sch. 11.). Bird name ueno<; (H.) and
aenov· Taxu 'quick' (EM). Note uen�<; (of KovlaaAo<; r 13), perhaps from � uon�<;,
but rather not after aena. Verb UEneTat· meL (EM).
.ETYM Cf. euena. Direct derivation fron the root of � a'lfll, *h2ueh,-, is impossible
because of related W awel [f.] 'wind', which requires the reconstruction *h2eu-el-.
The Greek form can also be derived from *uFA-la. The gloss UelA'l· ltVO� 'breath,
etc.' (H.) does not fit in. See � aETflov.
aE!!!!U [n.] 'bow' (Call.). <{ ?�
.ETYM Assuming an original meaning 'bowstring' (cf. � v£upa), explained as an
artificial form for aflfla 'knot, cord' (to � a1tTw); this is doubtful.
aET!!OV => UTflo<;.
at�w => av�w, au�avw.
aE1tTO" [adj.] of uncertain mg., cf. ae1tTov· laxupov, uOlK'lTov (aelKTOV) 'strong,
uninhabited' (H.). <{ ?�
.VAR Sometimes we find � aa1tTo<; or aeA1tTo<; (A. Supp. 908, Ag. 141, etc.).
.ETYM Unknown. See Wackernagel Stud. ital. fil. class. 5 (1897): 27ff.
CtEPO'l' [m.] Boeotian name for the bird flEpO'l' (sch. Ar. Av. 1354). <{ PG? (S, V)�
.VAR Also AEpOlte<;· eevo<;, TpOl��va KaTOlKouvTe<;. Kal ev MaKeoovlq yEVO<; Tl. Kal
opvea Tlva 'people inhabiting Troezen; lineage in Macedonia; kind of bird' (H.); the
gloss uepolto<;· KOXAla<; 'snail' (H.) is corrupted acc. to DELG following Latte.
Variants dpo,\! = flEpO'\!, also Boeotian (Arist.); PN'HEpolto<;.
.ETYM See Beekes Glotta 73 (1995-1996): 12-34; it is clearly a substrate element.
Chantraine thinks that the a- is long because of Ant. Lib. 18, 3 �EpOltO<;. Fur.: 243,
246, 352 assumes the interchange fll F and prothesis u-I e-I �-: flEpOlt- I UFEPOlt- I
dpolt- < *eFepolt- I �FEpOlt- (but this remains uncertain; perhaps d-, �- are
secondary lengthenings). Pre-Greek origin of the bird name and the proper name is
probable (cf. the suffix -Olt-). See � flEpO'\!, � lt'lVEAO,\!.
aEOU [v.aor.] 'to spend (the night)' (Od.), always with vUKTa(<;). <{ IE *h2ues- 'live, dwell,
spend the night'�
VAR Secondary pres. aEaKW (Hdn., H.).


.ETYM Related to Skt. vasati 'to dwell, live, spend the night', Hitt. buis-zi 'to live', ToB
was- [verb] 'to stay, to wait', Go. wisan 'to be', Arm. gom 'I am' (but rejected by
Kortlandt AAL 19 (1998): 19f.) etc. There is also an old present iuvw < *h2i-h2eus-.
Not related to aaTu, eaTlu.
cu::al<ppwv => auw.
a£aKw => awu.
U�£TOV . amaTov. LlKeAol 'not trustworthy (Sicilian) (H.). � ?�
.DER a�£Tow in a�£TWeeWVTl (Delphi, SGD! 2034, 17), but the mg. 'if they are
persuaded' poses difficulties, as it is the opposite of the gloss; see DELG.
ETYM Unexplained.

a�TJxq" [adj.] 'anuuaTo<;, auvex�<;, incessant', of noise, pain (ll.). � GR�
DER In H. also a�uxe<; and a�exe<;· aOLUAel1tTOV 'incessant'; a�'lxe<;· aOLexe<;
'unceasing' (Suid.) .
ETYM Probably for *a�uex�<;, which can be read in all places in Homer, from * a�olu­
ex�<; (cf. auvex�<; 'continuous') (Schulze 1892: 471, Bechtel 1914 s.v.). But the
contraction * ue > 'l is irregular in Ionic. DELG therefore suggests the influence of
�X� and compounds like oua'lX�<; but also remarks that * ue > 'l may be found in
Thessalian, Lesbian and Arcado-Cyprian. For the meaning, Bechtel assumes 'der
ohne Einhalt etwas tut', but it simply means 'without interrruption'; olexw means 'to
stand apart, be separate'.


u�ov [adj.] . fleAuv, U'l''lAOV 'black or dark, high' (H.). � ?�
.ETYM von Blumenthal 1930: 33 corrects fleAuv to fleyuv. However, his explanation
from *ag-jon (to ayuv and fleyu<;) is quite uncertain.
a�0f.laL => CiYLO<;.
u�w 1 [v.] 'to dry' (ll.). � IE *h2ed- 'dry'�
.VAR Mostly intr. a�OflaL 'to parch'.
COMP aoouuov· �'lpov. AUKwve<; 'dry (Lacon.)' (H.) would be a compound of a�u
(see below) and uDo<; 'dry' (Fraenkel Gnomon 21 (1949): 39, Fraenkel Glotta 32 (1953):
22); improbable; Latte corrects the gloss to *aoouvov. Cf. a�uUTo<;· nUAaLOT'l<; KUt
KOVl<; 'antiquity, dust' (H.), which remains unclear.
.DER HelL a�u 'dryness, heat', in aUKo<; . . . nenuAuwevov a�n (X 184) often
interpreted as 'shield defiled with mold', which seems unnecessary, cf. Ci�u· Cia�oAo<;
KOVl<; 'soot, dust or ashes', nUAaLOT'l<;· Konpo<; £v aYYEl4J unoflElvuau 'filth having
remained behind in a receptacle' (H.).
Adjective a�uAeo<; 'barren, arid' (n.), cf. iaxuAeo<;, UUaTUAeo<;. Verbs a�uvoflaL (h.
Ven.), a�ulvw (Nic.) 'to parch', both deverbative.
.ETYM Acc. to Frisk, the nearest cognates are found in Slavic: OPoL ozd 'dried malt',
Cz., SIn. ozditi 'to dry malt' < *h2esd-. Germanic has a group of words with velars in
place of dentals: Go. azgo, OHG asca 'ashes'. However, it is more probable that Gr.
a�w is comparable to Hitt. bat-i 'to dry up' < *h2od-ei, and should be reconstructed as
*h2ed-ie/o- (cf. Kloekhorst 2008). What is perhaps originally an extension of the



same root, if *h2eh,s- < pre-PIE *h2eds-, is found in Lat. area 'to be dry', ara 'altar'
(OLat. asa), Hitt. bassa- 'hearth' < *h2eh,s-h2. Skt. asa- [m.] 'ashes, dust' probably
continues *h2eh,s-o-. Cf. � uDo<;, � auaTuAeo<;.
a�w 2 [v.] 'to cry, groan'. => 0. 5.
aTJ!5wv, -ovo" [f.] 'nightingale' (Od.). � PG?(S)�
.VAR Also allow, -ou<; [f.] (S.). Also [m.] , but rare.
.ETYM From * aF'lowv; cf. a�'loovu· a'loovu (H.). Connection with � aElow and
� uuo� (which is almost universally accepted) is difficult, since a lengthened grade
*h2ued- is improbable. The word could therefore well be Pre-Greek; for the suffix, cf.
bird and animals names like � xeAiowv 'swallow', � TeVep'lOwv 'wasp' .
UTJf.lL [v.] 'to blow' (ll.). � IE *h2ueh,- 'blow'�
VAR Forms in Schwyzer: 680 .
.DER a�T'l [f.], a�T'l<; [m.] 'wind'; rare are the verbal nouns a'lflu, a'lat<;. A zero grade
ae- « *h2uh,- in aeTflov· TO nveuflu (H.), whence � aTfl0<;. Cf. furilier � aeAAu; Cio<;·
nveuflu � a'lflu (cod. '(uflu) 'breath, wind' (H.) is an innovation. Unrelated is � a�p,
which has long a-.
.ETYM An old verbal root *h2ueh,- is also reflected in Skt. vati, Go. waian, OHG waen
and OCS ISg. ve-jr, all 'to blow'. The word for 'wind' is (a thematization of) the
participle of this root: *h2ueh,-nt-o- > Skt. vata- [m.] , Av. viita-, both of which must
often be read with three syllables, representing Proto-Indo-Iranian *vaHata-. The
same word is found in Lat. ventus, Go. winds, ToA want, ToB yente; Hitt. buyant- is
from unthematicized *h2uh,-ent-. A form with suffix *-t- has been assumed in in
� a�aupo<; 'airy, quick (as wind?)' (poet.), for which cf. Skt. vatula- 'windy', but this
gives formal problems (*tu > au is not regular). See � CieAAu, � uupu, � a�aupo<;,
� aTfl°<;·

lU1P [f.] 'mist, haze, clouds' (n.), later 'air, atmosphere'. � IE *h2eus-er-�

.VAR Gen. �epo<;. The nom. Q�P (also Att.; thence gen. Qepo<;) arose by dissimilation;
is it an Atticism in Homer? Later Ion. ��p .
.DIAL AeoL uU'lP, Dor. a��p (= uu�p) (H.).
.DER Derivatives: �epoel<;, �epoelo�<; 'dim, cloudy', also � uupu.
.ETYM a�p is not cognate with a'lfll. Meillet BSL 26 (1925): 7ff. assumed an original
meaning 'suspension' and derived the word from aelpw 'raise'. However, this leaves
the length of the initial vowel unexplained. Kiparsky Lang. 43 (1967): 619, 626 derives
the word from *auser, arguing that related � uupu < *h2eus-r-h2 still means 'morning
mist' in e 469. See � �eplo<;, � � pl.

aqo"uAo" [adj.] in the hapax a�auAu epyu (E 876) 'criminal acts'. � ?, PG? (V) �
.VAR ui�auAOV· avoflov, KUKonOlOV 'lawless, doing ill' (H.).
.ETYM It has been proposed that the word is a modification of U'LauAO<; 'unseemly,
evil' (U'LauAu pe�£lv, E 403, etc.) after a'lfll or a�aupo<;. Different explanations are
found in Bechtel 1914 and Brugmann Sachs. Ber. 53 (1901): 94. Fraenkel Glotta 34
(1955): 307ff proposed * u(F)laauAu, connecting it with lao<; (highly doubtful). Fur.:


253 points to the variant ui- in the gloss, which may imply a substrate origin
(comparing aTjToc;, which may have a variant u'iTjTOC;).
a�<Jl)po" [adj.] probably 'light, agile', said of ants (A.). � PG?(s)�
.VAR Cf. a�oupov· TO Ae1HOV, TO il£Tewpov Kui KOU<pOV 1tUpa TO aepl mJPw8at £1ti
opvewv 'delicate, elevated, light, after being dragged through the air by birds' (Suda).
.ETYM Connection with � aTjill is improbable; rather a substrate word (where the
suffix -up- is not infrequent).
«TjTO" [adj.] in 8apooc; aTjTov (<D 395), mg. uncertain. � PG?(v)�
VAR Cf. also 8apooc; acnov (Q. S. 1, 217). Also aTjTol' UKOPWTOl, a1tATjOTOl
'insatiable, greedy'; a�Touc;· il£yaAuc;. AioxuAoC; A8ailuvTl 'great (Aesch.) (H.),
aTjToc;, 0 aKUTa1tUUOTOC; 'incessant' (Hdn. Gr. 1, 220) .
ETYM The first explanation connects the word with ail£vat, � &Oat 'to satiate', but
this would mean that it differs from aUToc;, aTOC; by its long vowel, which is
improbable. Perhaps it is the same word as � U'iTjTOC; (in 1teAWp u'iTjTOV 2: 410, said of
Hephaistos). If so, the variation al at might point to a substrate word; metrical
lengthening is improbable, and a < at impossible. See Fur.: 253, though his
connection with � a�auAoc; remains uncertain. Palmer 1963: 339 connects the epithet
of Hephaistos with Myc. a-ja-me-no as 'artist'; this is improbable. Not related to
aTjill. See Sabbadini Riv. studi class. 15 (1967): 78-84.


ae&pTj [f.] 'gruel, porridge' (Ar.). � ?�
.vAR Also a8�pTj, -a [f.] (Hell.); influence of � a8�p?
DER a8upw8TjC; (Ruf. Med.) and a8�pwila 'kind of ulcer' (Gal.) .
ETYM Unexplained; an Egyptian word according to Pliny (N. H. 22, 121). The final -Tj
in Attic, which is confirmed by Moeris, would lead us to suppose a pre-form
*a8UpF<l. Connection with a8�p is neither formally nor semantically plausible. Not
related to Lat. ador (Hamp TPS (1968): 106), as this belongs to Go. atisk and Iran.
adu 'grain' (Szemerenyi 1969a: 968f.).


a8£A�- - a8eAym.
a8t:\.ynv [v.] aileAY£IV 'to milk' (H., EM), (E�)u8eAy£Tat (Hp.), explained as 1tUple-rat,
81£KAU£Tat 'lets go, releases' by Galen. Also a8eAYTjTat· 8TjAa(£Tat � 8A(�TjTat 'is
suckled, compressed' (H.), cf. a8eAYTjTat· BUKX£1oC; <PTjat 8TjAa(£Tat � EmomlTat, Kui
EK8Ai�TjTat WC; Kul N(Kuv8poC; 'is suckled, drawn after one; is squeezed (out),
crowded' (Erotian. 20, 1; see Hp. De med. off. 11). � PG(v)�
.VAR Cf. a8eA�£Tat· 81Tj8£LTat 'is strained through' (AB); a8£A�ci· £AK£1 'draws' (H.)
and a8£A�a(£lv' 8lTj8civ 'to strain through' (H.); further a8eA8£Tat· 81Tj8£LTat 'id.'
(Diocl. Com. fr. 7 Kock, An. Bekk. 350) .
ETYM These verbs, meaning 'to press, draw away, filter', have no etymology. The
variation �I 81 I' should not be explained from an IE labiovelar (as per Solmsen 1909:
9'), as contamination leading to three different forms is improbable. Since Pre-Greek
also had labiovelars (cf. �UatA£UC;, Myc. qa-si-re-u), substrate origin is most probably
the source of the alternations. See Beekes Glotta 73 (1995-1996): 12f., and cf. ye<pupu
beside M<pupu, see Fur.: 388, 390.



a8£:\.8- - a8eAym.
a8£p(�w [v.] 'to disparage, neglect' (ll.). Originally always with negation. � ?�
.DER a8eplmoc;· a<ppovTlmOC; (Zonar., A. Fr. 128), cod. -ITOV.
.ETYM Unknown. The glosses a8£pec;· avoTjTov, avoatov, aKpl�ec; 'stupid, not in
order, precise' (H.) and a8£p�c;· <'> a(8Tjp0C; aT£lp�c; amv 8£p((n, etc. 'indestructible
iron when it is heated' (H.) seem unrelated in view of their meaning. An older
proposal derives it from *a8£poc;, connecting it with Skt. adhara- '(be)low' (see
Bechtel 1914). Another proposal (DELG) derives it from � a8�p, as flocci facio.
Neither is convincing.
A8qVT) [f.] th� goddess (n.), a common Greek goddess dating from Minoan times,
protecting the palace, and depicted with a snake. � PG�
.DIAL Myc. a-ta-na-po-ti-ni-ja [] ? (see Aura Jorro 1985-1993: 112). Dor. (etc.)
DER The town A8�vat (Dor. A8Civat) contains the same onomastic element; for the
suffIx cf. MUK�VUI. Thence A8Tjvaloc; 'Athenian' (n.), whence fem. A8Tjvu(u, -Tj,
which is also used as the name of the goddess (88 times im Hom.). This form gave
*A8Tjvau (Aeol. ' A8uvuu with short vu), which was contracted to Att. A8TjvCi.
.ETYM Like the goddess itself, the name is pre-Greek. Note the suffix * -an-. See

a8qp, -tpo" [m.] 'awn', plur. 'chaff, barb of a weapon, spine or prickle of a fish', also
'edge of a weapon' (Hes.). � PG(v)�
.VAR With a nasal av8epl�, -IKOC;, etc.; with suffix -£wv: av8£p£wv, -WVOC; (see below) .
.COMP a8TjpTjAOlYOC; 'winnowing-fan', from 'consumer of chaff (Od. A 128 = 8 275),
but the two Tj's are surprising.
DER a8£p(vTj f., -1voC; m. 'kind of smelt, Atherina hepsetus' (Arist.), cf. Chantraine
1933: 204, Thompson 1947 S.v. Also a8£pTjic;, -(80c; [f.] 'prickly' (Nic.), o.8£pw8TjC;
Variants with nasal: av8epl�, -IKOC; [m.] = a8�p, also 'ear' (ll.), av8eplKoc; [m.] 'stalk
of an asphodel, asphodel plant' (Hp.), av8£pIKW8TjC; (Thphr.).
With the suffIx of
place -£wv: av8£p£wv, -WVOC; [m.] 'chin' (ll.).
.ETYM No etymology. IE ablaut *h2endh- : *h2ndh- is impossible, as both forms would
give Gr. av8- acc. to Rix's Law. The nasalized forms could be due to folk etymology,
but rather point to Pre-Greek prenasalization. Fur.: 296 further adduces o.V8£p(OKOC;
= av8£pIK- with the interchange oKI K; perhaps also � av8puoKOV I E- 'chervil'. Not
related to Lat. ador, both because of the meaning, and because this is kindred with
Iran. adu, Go. atisk; see Szemerenyi 1969a: 958f. The word has nothing to do with
names of the wasp or forest bee (� o.v8p�vTj, � av8pTj8wv), nor with � av8pw1toc;.

a8paytvT) [f.] a plant, 'Clematis vitalba' (Thphr.). � PG(v)�
.ETYM Frisk compares � a8puc; 'chariot', which he thinks would fit a climbing plant
well, but I don't see what a car has to do with a plant. See Stromberg 1940: 108. Fur.:
288 compares av8paxvTj and concludes to a substrate origin (prenasalization), which
is probable anyhow.




«6pa" [m., f.] . upfla. 'POOlOl 'chariot (Rhodian) (H.). -<! ?�
.ETYM Unexplained. Formerly compared with Skt. vandhura- [m.] 'wagon-seat',
taken as a wicker basket tied upon the wagon, and connected with MoHG winden,
etc. (Pok.: 1148) as IE *uendh- : undh- >(» Gr. *Fa8-. As the formation of the Sanskrit
word is unclear, and the root is hardly attested outside Germanic, the connection
must be false. Banateanu REIE 3 (1943): 149 calls the word Anatolian. Connection
with Kuvva8pov is improbable.
a6pEw [v.] 'to gaze at, observe' (11.). -<! ?�
.DER a8p�flaTa· owpa nEflnoflEva napa TWV auyyEVWV TaT<; yaflouflEvm<; nap8Evm<;
napa AW�[Ol<; 'gifts having been sent by kinsfolk to maidens being given in marriage
(Lesbian) (H.) (Snell Glotta 37 (1959): 282-287, cf. Renehan Glotta 49 (1971): 66).
ETYM No etymology. One compares £v8puv· <jlUAaO"<J£lV 'to guard' (H.) and 8p�aKw·
vow 'to perceive' (H.), 8pfjaKEuw 'to observe' etc., but this leaves the a- unexplained.
It cannot be the zero grade of £v-, since a zero grade *h,n- would still yield £v-. The
derivation by Hoffmann 1921: 78f. from a noun *a8po<; 'directed at a goal', from IE
*dher- 'to hold' and copulative 0.- (cf. � a8poo<;), is most improbable. On the use of
a8pEw, see Prevot RPh. 61 (1935): 246f.

a6poo" [adj.] 'in crowds, gathered together' (11.). -<! ?�
.DIAL Att. a8poo<; (spiritus asper perhaps restored after una<;, ufla).
.DER a8po[(w (a-) 'to gather together' (Archil.).
.ETYM Compared with Skt. sadhry-aiic- 'united' (Brugmann 1894: 14ff.), which was
formerly analyzed as containing the root *dher- 'to hold', but this is no longer
maintained by EWAia. It does not belong to � a8pEw, nor to � 8povo<;. It is probable
that a- is from *S111 -, but the further analysis is uncertain. Risch 1937: 179 compares
aAAo-8poo<; 'speaking a foreign language', in which case it would mean 'calling
a6upw [v.] 'to play, sport' (11.). -<! ?�
VAR Only present.
.DER Ci8upfla 'plaything, toy' (11.), plur. also 'adornments', diminutive a8upflunov.
Deverbative a8upEuw8m· na[(Elv, fllyVU£lV, aKlpTuv 'to play, intermingle, leap' (H.).
ETYM From *a8up-yw. Compared with Lith. padurmai 'impetuous', Ru. dur'
'foolishness', from PIE *dhuer(H)- 'to whirl, rush'. However, a- cannot be the zero
grade of *h,en- 'in' (which would not really make sense semantically anyway), and an
initial laryngeal gives the improbable root structure *h2dhuerH-.


a'l exclamation of surprise, pain, or sorrow. -<! ONOM�
VAR Also aiaT.
ETYM Elementary formation, found in many languages.


ai 'if. - EL
ata 1 [f.] 'earth' (11.). -<! ?�
.ETYM By some considered identical with � ala 2 (Brugmann IF 15 (1903-1904): 94ff.,
Brugmann IF 29 (1911-1912): 206ff.), in which case it would originally mean 'mother'.


The relation with yaTa and flaTa is uncertain; cf. Giintert 1914: l26f., Brandenstein
1954b: 80. Van Windekens assumed *as-ya from *h2s- 'dry' (but this root perhaps did
not exist; see � «(w). Woodhouse KZ 107 (1994): 99f.) assumes *sausja, but cf. � 0.130<;.
ata 2 [f.] 'mother, grandmother'? (11., poet.). -<! ?�
.VAR Cf. ala· uno Kupfjva[wv Tfj8[<; Kat flaTa Kat a8£A<jl� Kp�Tfj<;. Kat <jlUTOV Tl. £Tl O£
o Kupno<; aUT<f' oflwvuflo<; (EM 27, 24).
.ETYM Compared with Lat. avia 'grandmother'; uncertain. Elementary word? See
� ala 1.
aiuvq" [adj.] 'horrible' (Archil.); the mg. 'eternal' (A.) through association with � aIEL

-<! ?�
.VAR Ion. alfjv�<; .
.ETYM There is a speculative hypothesis by Wackernagel 1897: 7, who assumed the
reconstruction *amF-uv�<; 'with terrifying face' (whence Lat. saevus, the word for
'face' seen in � anfjv�<;, etc.). See Degani Helikon 2 (1962): 37-56.

ai�oi exclamation of disgust (Ar.). -<! ONOM�
VAR Also al�Ol�oT, oflaughter.
.ETYM Onomatopoeic, elementary formation. See Schwyzer: 600.

aiyavEIl [f.] 'hunting spear, javelin' (11.). -<! ?�
.ETYM For the suffIx, cf. the names of trees and animal skins in -Efj, -Ea: flfjAEfj,
1tTEAEfj, KUVEfj ete. (see Chantraine 1933: 91f.). If named after the material, one
compares the word for 'oak' in PGm. *aik-, which is also supposed in � aiy[Aw'i' and
Lat. aesculus. Triimpy 1950: 52, 57 explains that the alyavEfj was thrown by a strap.
Laser Gymnasium 60 (1953): 115-l21 connected it with PIE *h2eig- 'to stir, set in
movement', in Skt. ejati, to which alYE<;· KUflaTa (see � a'i�) is compared. This is
semantically improbable, and if Skt. ej- belongs to ilig- (see Mayrhofer EWAia 1:
264), then the root contained a labiovelar, which would make the connection
impossible. To � aiXfl� ace. to Bechtel 1914·
a'(ynpo" [f.] 'black poplar' (11.). -<! PG(v) �
.VAR a'lyEpo<; (Com. Adesp. 1276, Kock) .
DER alyElpwv 'poplar grove', aiy£lplvo<;, aly£lpLTfj<; 'of the poplar' (all Hell. and late) .
.ETYM The connection with � aiy[Aw'i', � alyavtfj is uncertain. Sommer IF 55 (1937):
260 pointed to numerous non-IE words like a'iyl80<; and names in Aiy- (A'lylva,
Alya[, etc.), suggesting a Pre-Greek origin. This would be confirmed by the form
with -E-.

aiYla;\o" [adj.] 'sea-shore, beach'; also TN, e.g. the coast of Achaea (11.). -<! ?�
.DIAL Myc. aJ-ki-a2-ri-jo probably laigihalio-I, see Aura Torro 1985-1993: 134·
.DER alYlUA£lO<;, aiYlaA£1.J<;, alYlaAlKo<;, -A[Tfj<;, -AWOfj<; all are late derivations. Also
AiYlaAu<; name of the inhabitants of the coast of Achaea Hdt.) .
ETYM The Myc. form seems to confirm that the second element is derived from
� UA<;. For the first member, cf. alYE<;· Ta KUflaTa. LiWPlET<; 'waves (Dor.) (H.) and
Artem. 2, 12 Kat yap Ta f.J:EyUAa KUflaTa alya<; £V Tn auvfj8£l<;t MyoflEV 'we usually call



large waves aiya<;'. This in turn is comparable with Skt. ej- 'to storm', but see
� aiyavE'l ' It is highly improbable that aly£<; in the sense of KUf1aLa is a metaphorical
use of a'(� 'goat' (as per Heubeck IF 68 (1963): 13-21). The word is Pre-Greek acc. to
Chantraine 1933: 248, which cannot be excluded, though Chantraine now calls it 'all
too easy' in DELG s.v.
a'(Yl8o<; [m.] 'titmouse (Parus)' (Ar.). <!! PG(S,v)�
.VAR a'(YlvSO<; (Dionys. Av. 1, 12) ; a'(YlvSo<; f1lKp6<; parra modica' (gloss.), not
mentioned by Frisk or DELG; aiyloSo<; (Arist. v.l.), for *a'(YlaSo<;?
.DER aiy(SaAAo<;, -SdAO<; (Ar.).
ETYM A typical substrate word, discernible from its prenasalization and the suffIx
-aA(A)- with variation AI AA. See Fur.: 267, 288, 2548, 304, 387, with further literature.

aiYlAlljl [adj.] 'sheer, steep' (ll.), also as a TN. <!! ?�
VAR aiYIAl,/,' U'/''lA� TtETpa Kat Tt6Al<; Kat hEa UTtO 80UplWV 'high rock, citadel,
willow (Thourian) (H.) .
ETYM The Ancients explained it as 'abandoned even by goats', which is clearly a folk
etymology. In modern times, it is connected with Lith. lipti 'to clamber' (see Solmsen
1901: 731) as 'what can be climbed only by goats', which is also highly doubtful. The
meaning 'clamber' for the root *leip- is secondary to 'stick, cleave', and it is far from
certain that Greek underwent the same development as Lithuanian. The formation
in aiyl- is unexplained as well (see � a'(�). Cf. also � aAl,/, . TtETpa 'rock' (H.), the
explanation of which from 'what cannot be climbed' is doubtful too; the gloss AI,/,'
TtETpa a<p' �<; uowp aTCt�£l 'rock from which water drips' may be due to later
interpretation (Solmsen, cf. Persson 1912(1): 1521); Marzullo 1969: 101f thinks it is a
mistake for a[iyl]Al'/'.


aiYlAw,/" -WTto<; [m.] 'kind of oak' (Thphr.), also 'oat-grass' (Thphr.); 'ulcer in the eye,
lacrymal fistula', for which see � aYXIAW'/'. On the mg. see Stromberg 1940: 87. <!! ?�
.ETYM As the name of a kind of oak, aiylAw,/, has been connected with the Gm. word
for 'oak', PGm. *aik- (Pok.: 13) . Kretschmer Glotta 3 (1910-1912): 335 connected -AW,/,
with AWTt'l 'cork' (cf. AW'/" XAaf1u<; 'cloak, mantle' H.), adduced from Pliny (H. N. 16,

6, 13): aegilops jert pannos arentes . . . non in eortiee modo, verum et e ramis
dependentes. Within Greek, it is doubtful whether one can connect it with � aiyavE'l
and � a'(Y£lp0<;. Stromberg 1940: 137 derives aiylAw,/, in the second meaning from
a'(YlAo<; 'havergrass' (Theoc.), which is certainly correct.

aiYl<; [f.] 'goatskin' (E. Cye., Hdt. 4, 189), a mantle protecting Zeus and Athena (ll.);
later also 'storm wind' (A.); further 'heart-wood of the Corsican pine or the silver fir
in Arcadia' (Thphr). Also 'speck in the eye' (Hp.), on which see � ayAI'l. <!! GR�
.DER aiylox0 <; epithet of Zeus (ll.), cf. yatCtFOXo<;. With the last mg. £Tt-atYI�w 'rush
upon' (from a storm wind) (Horn) .
ETYM Probably a goatskin in origin (thus Hdt. 4, 189); formation like V£�PI<; 'fawn­
skin' (see � v£�p6<;), etc. In Homer, gods and men are frightened when it is shaken.
For the meaning 'storm wind', see Heubeck IF 68 (1963): 13-21. Kretschmer Glotta 27

(1939) : 28, connected it with a'(YA'l and aiy£<;' Ta KUf1aTa 'waves' (H.); one might
think of Skt. ejati 'to move, set in motion'.
*aiYAl<;, -lSO<; 'speck in the eye' (Gall.). <!! ?�
.ETYM The form is reconstructed in order to account for aiy(<; and aYAI'l(<;); cf. also
aiYlaAlo£<; (H.). See RPh. 73 (1999) 81f. s.v. aiYI<; (derived from a'(yA'l). Or should we
reconstruct *aYAl<;?
aIYATJ 1 [f.] 'light (of sun or moon), gleam, radiance' (ll.). <!! ?�
.DER aiyA�£l<; 'gleaming' (ll.), aiYACtTa<;, -�T'l<; epithet of Apollo (inscr. Anaphe,
Thera; A. R.); aiYACt�W 'to shine, gleam' (Man.) .
.ETYM The connection with Skt. ejati 'to move, tremble' (cf. � aiyavE'l) is rightly
rejected by DELG. The epithets ATt6AAWV � Aay£AaLa<; (Anaphe) and ATt6AAWV
AiYACtLa<; (Anaphe, Thera) are often compared. In view of the variation aiyA- I
aay(£)A-, the epithets must be Pre-Greek. The noun may be of Pre-Greek origin too
(it has no etymology), but it does not necessarily have the same origin as the
epithets; it is methodically incorrect to clarify the formation of the appellative by that
of a name.
aIYATJ 2 [f.] 'ring' (deduced from glosses). <!! ?�
.ETYM Lewy KZ 59 (1932) derived it from a'(yAa<;· af1<plOEa<; Kat ,/,EAla 'iron rings,
anklets'. Ta Tt£pt T�V UVlV TOU ap6TpOU 'things around the plowshare' (H.), from
aiy<l>Ala· oaKTuAlola 'small rings' (H.), and other words attested in lexicons. He
explained it as a loan from Hebr. 'agIl '(ear-)ring', which remains hypothetical. As
Frisk remarks, metonymic use of � a'(yA'l 1 'gleam, splendor' is well possible.
aiyvmo<; [m.] 'vulture' (ll.). <!! PG(v)�
.VAR aiYITto,/,' 0.£<6<; UTtO MaK£oovwv 'eagle (Maced.) (EM 28, 19) .
.ETYM The comparison with Skt. rji-pya-, epithet of the bird syena- 'eagle, falcon',
Av. JrJzi-jiia- is formally difficult, since we expect *apC-; influence by a'(� and yu,/,
'vulture' has been assumed, but this seems unlikely. Fur.: 364 compares the gloss
aiYITto,/" which is evidently a variant of the same word. Variation between i and u is
well-attested in substrate words (cf. � f1Ctpal1tTto<;1 f1CtpOUTtTto<;), and -oTt-is a Pre­
Greek suffix. � yu,/, itself is no doubt a substrate word as well; it may be a variant of
*(a)g'up-, which also gives aiyuTt-, with prothetic vowel and palatalized Ig'l. No
conclusion is reached by Meier-Brugger KZ 108 (1995): 50-55.
aiywAlo<; [m.] a kind of owl, 'Stix flammea' (Arist.). <!! ?�
VAR Also aiywAlo<;.
.ETYM The reading ahwAlo<; (Arist. HA 563a 31) is wrong, as evidenced by forms
from modern southern Italy (agoleo etc.); see Rohlfs ByzZ 37 (1937): 55. Etymology

a"tSTJAO<; [adj.] mg. not quite clear: 'hated; annihilating, destructive;
. invisible, unseen'
(ll.). See the discussion in DELG. <!! GR�
.ETYM It is difficult to decide what the primary meaning is, but it is rather clear that
the word contains a privative a with a form of iodv. Ivanov 1999: 283-292 compares


Ru. nevidal' for the semantics and assumes the development 'invisible' > 'strange,
dangerous' .
1\:tS'1C;, -ao [m.] Hades (11.). -<!l IE *I:J-uid- 'unseen, invisible'�
.VAR With metathesis of quantity gen. -£w; also gen. 'AI06C;, dat. -L Att. Ato'1C;, -Ou;
Ato'1C; is found in later Ionic poetry (Semon., Herodas), the Doric form Atoac;, gen.
-a in tragedy. An inscription from Thessaly (SEG 16, 380) gives ApOav.
.DER A'LOwv£uC; (11.), see Risch 1937: 158.
ETYM Thieme proposed (Thieme 1952: 35-55) that the word derives from *syt1 uid­
(Skt. sam vid-) as 'das Sichzusammenfinden [gathering of the family in the
underworld] . To my mind, this is not correct, as then Alo- should denote the
Underworld, not the God of the Underworld. In Homer, it rather seems to denote
the God, e.g. in formulaic 06floV 1\IOOC; daw. The aspiration in Attic is secondary
and arose by contraction from 6 AtollC; (Kamerbeek apud Ruijgh Lingua 25 (1970):
The other explanation, as *I:J-uid-, 'the Unseen', seems the correct one. In Beekes
1998: 17-19, I pointed out that the replacement of a root noun, first in the nominative
(here as the final element of a compound), is parallel to the case of <puy� : <puya8e.
The initial A- is sometimes lengthened for metrical reasons: it is not lengthened
when it is not necessary or impossible, like in 1\106aOe.

a'lSo!1at [v.] 'to hold back, be ashamed; to honor, respect' (11.). -<!l IE? *h2eis-d- 'honor'�
VAR a'L8eo, ai06fl£voC;, a'Lono, see Chantraine 1942: 310f.; more frequent is
denominative aiOEOflat (from *aides-je/o-).
.DER aiowc; [f.] 'shame, reverence' (11.), whence 1. aiOoloc; < *-os-jo- 'inspiring aiOwc;'
(11., epic poet.) substantivized ntr. TO aiOolov, usually plur. Ta aiOola 'private parts'
(11.), whence aioolwOllC; and aiOo'LKOC;; 2. compound UV-atO�C; 'shameless' (11.),
uvalO£la, etc.; 3. aiOEOflat < *aides-je/o- 'to hold back, revere', as a legal term also 'to
be reconciled' (Hom., lA), fut. ai<'l£aoflat. Thence a'LO£atC; 'mercy, pardon' (D.,
Arist.), aiOwToc; 'honorable' (PIu.), aiOwLlKoC; (sch.); aiO�flwv 'modest' (X., Arist.),
ai0'1flovIKOC; and -floaUv'1 (late and rare). aiO£atflOC; (post-class. prose) 'object of
aiOwc;', also aiO�atfl0C; (Orph.), whence aiOWlfloT'1C; (pap.). 4. aiooaUvll
ai0'1floaUv'1 (AB, Phot.) .
ETYM It is formally uncertain that a PIE root *h2eisd-, from which Go. aistan 'to
hold back, respect' and Skt. 1# < PIIr. *Hizd- 'to praise, honor' derive, would give
Gr. aiO-: we would expect *h2eisd- to appear as Gr. ai�- (cf. r�w < *si-sd-). Of course,
the connection is semantically very tempting. Without final * -d, the root *h2eis- is
found in MoHG Ehre 'honor', ON eir, etc., and in Osc. aisusis [] 'sacrificiis',
and perhaps originally in the verb *h2is-ske/o- 'to demand' > Skt. icchati 'search for',
OCS iskati 'to search', etc.



(ifSUAOC; [adj.] 8paaUC; 'bold, rash' (H., EM). -<!l ?�
ETYM A mistake for Ct"tOllAoC; (E 897)? Leumann Glotta 32 (1953): 2184 differs. Fur.:
262f. compares a'LauAoc;.



a'lSwaaa [f.] (cod. aiowaaa} T�C; aUA�C; Ta mXla 'the walls of the court or hall' (H.).
-<!l PG(V)�
.ETYM von Blumenthal 1930: 5f. suggested that the word is Illyrian for areouaa. Latte
thinks it is a corruption (but on three points?). Fur.: 197 considers it a a substrate
word, taking � areouaa as a variant form, and comparing K�Awaaa / KllAouaa, a
mountain in Sicyon.
aid [adv.] 'always' (11.) -<!l IE *h2ei-u- 'time of living, well-being'�
.VAR aiw (A. Ch. 350, AB 363), see below.
.DIAL aiF£l (Cypr., Locr., Phoc.), ud (Att.), ai£c; (Dor.); ai'1 (Tarent.) .
DER UtOIOC; 'eternal' (Hes., lA), whence u'LOtoTllC; 'eternity' (Arist., Hell.).
.ETYM From *aiwes-i, the old locative of an s-stem, which is found without ending in
Dor. aiEC;, and also in the accus. aiw < *aiwos-yt1. Beside the s-stem, Greek had an n­
stem in � aiwv, from which aiEv derives. The Tarentine form would be an old
instrumental *h2eiu-eh" but this is unclear. On the accentuation of the Greek forms
see Hamp Glotta 67 (1989): 41.
The s- and n-stems derive from the old PIE u-stem *h2oi-u- > Av. iiiiu- 'life, time of
life', gen. *h2i-eu-s > OAv. yaos, thematicized *h2ei-u-o- > Lat. aevum, which in Greek
may be found in Aeol. alt(v) , <'h(v) < *aiw-i(n). As Weiss MSS 55 (1994): 151f.
suggested, the zero grade *h2iu- is perhaps found as the first member of Cypr. u-wa­
i-se 'forever' < *h2iu(y)-h2ei-s- (but see � u) and of � uyl�C;. Both the s-stem and the n­
stem extensions seem to be found in Sanskrit: ayu-n-i [loc.], and ayu-$- [n.] .

aiEAovpoC; [m., f.] probably '(wild) cat' (Hdt., Ar.); domestic cats were not found in
the Greek world. -<!l ?�
VAR Also a'(Aou pOC; (Arist.).
.ETYM The explanation as a compound of aioAoc; « *ai£Aoc;) and oupa 'with moving
tail' (cf. EM 34, 8: a'(AoupoC; TIapa TO aiOAA£lv Kat UVaY£lV T�V oupav Kat KIVdv) still
seems possible, although the -£- is problematic. The connection with Lat. vlverra
'ferret' and Lith. vaiveris 'male of the polecat / pitchew', starting from *FatFEpOUpOC;
(Ehrlich 1912: l28ff.), should be abandoned. Note that the word may well have been
adapted by folk etymology, and that from ai£- we expect an Attic development to
u£-, not the loss of -£-.

aiiv [adv.] 'always' (epic poet.). - aiwv.
aiEC; - aieL
ai£-roc; [m.] 'eagle' (11.), also metaph. 'tympanum, pediment'. -<!l IE *h2eu-i- 'bird'�
.VAR Att. anOC;; ai�noc;' unoc;, II£pyalOl (H.), with � = F; aillToC; (Arat. 522), which
is artificial (DELG).
.DER ueno£uc; [m.] 'young eagle' (Ael., Aesop.), unlTllC; (Al80c;) (Ael.), unwo'1C;
(Philostr.), aino£lC; (Opp.); aienaloc; 'ptng. to the tympanum' (inscr.); uETwfla
'tympanum' (Hp., Att. inscr.), UETwatC; 'arched roof of a x£Awvll' (Ath. Mech.) .
.ETYM From * aiFnoc; < *awjet6-, cf. Lat. avis. For the suffix -no-, cf. vI<pnoc;,
TIu pnoc;. The artificial form ai'1Toc; does not allow us to conclude that it is a substrate
word, as per Fur.: 1154• Not a Semitic word (Astour JAOS 86 (1966): 278B).

ai�T)6c; [adj.] unknown; 'strong'? (ll.). � ?�
.VAR Also ai��·lOe;; ai��ELe; (Theopomp. Col.), ui�aEv· Einpa<pEe; �MOTT]fla 'well-fed
offspring' (H.).
.ETYM Danielsson 1892 gives no definitive answer. The gloss suggested a connection
with ad and ��v, which may be folk-etymological (DELG). Fur.: 234, who connects it
with A'(oT]1toe;, is incorrect.
aiT)Toc; => UT]TOe;.
ai9uAT) [f.] 'soot' (Hp.).
.VAR a'l8aAoe; [m.] (Hp., E.), also as an adj. = ai8aAOELe; (Nic. Th. 659).
DER ai8aAOELe; (n., poet.) 'smoky, dark brown', also of the sheet of lightning (E. Ph.
183 [lyr.]), perhaps 'fiery, burning'; ai8aAEOe; 'id.' (A. R., Nic.); ai8uAlwV, -lWVOe;
(Theoc. 7, 138), epithet of the TETnyEe;, probably a color term with verse-final
metrical lengthening); ai8aAwOT]e; 'id.' (Arist., Gal.). Unclear ai8aAloae;· Ta £v nil
OlT4> YLvoflEva, � TOUe; £v T4J MaTL a-ruAaYflOUe; TOU £Aalou 'what is in the food, or
drops of olive oil in ilie water' (H.). Denominative verb ai8aAOw 'to make sooty',
med. -OOflaL 'to soot up' (E., Lyc.); ai8aAWOELe; 'clouds of sooty smoke' (Max. Tyr. 41,
4), perhaps directly from a'l8aAOe;.
.ETYM See .. a'l8w.

ai81lP, -tpoc; [f., m.] 'clear sky, heaven' (ll.). � IE *h2eidh- 'kindle, ignite'�
.COMP On ai8�p and a'l8pT] as a second member (e.g. in lJ1tal8pLOe;, u1taL8poe;) see
Sommer 1948: 151f.
DER a'l8pT], -a. 'id.' (ll.); ai8plT], -la 'clear sky, nice weather'; a,(8pLOe;, -ov [adj.] 'of the
sky, bright' (lA); the ntr. a'(8pLov, diminutive ai8plOLoV was used in imperial times as
a folk-etymological adaptation of Lat. atrium.
aI8poe; 'fresh, cold air' (� 318 a'l8p4> KaL KafluT4> OEoflT]flEVOV). Cf. ai8pci· XELflU�EL 'to
expose to or pass the winter' (H.), ai8pLvoV· 1tPW·lVOV 'early' (H.); improbable
Bouquiaus-Simon Ant. class. 31 (1962): 25ff.
With full-grade of the suffix ai8EpLOe; 'in the air, heavenly' (trag.), beside which rare
and late ai8EpWOT]e;, ai8EpLWOT]e;, ai8EplTT]e;, ai8EpOoflaL.
i8apoe; 'cheerful, bright' (Ale.) may contain an old ablauting form of the root. The
verb i8alVELV (A. D., H.) could point to an rln-stem.
.ETYM Generally derived from .. «'(8w; perhaps the formation was influenced by a�p.

Ai8iom:c; [] ethnonym, 'Ethiopians'(?). � PG(S)�
.DIAL Myc. PN a-i-ti-jo-qo IAithiokWsl (or 1-0-/); for the u-stem in Me-to-qe-u, Wo­
no-qe-we (/WoinokWewei/) cf. Ai8Lo1t�Ee; (Hom.).
ETYM Since antiquity explained as '(people) with burnt faces'. In Beekes Glotta 73
(1995-1996): 12-34, I objected that ai8- always means 'burning' in the sense of
'brilliant, emitting light' (cf. a18o'\l) , and never 'burnt'. Also, the -L- is unexplained,
and -01t- is a typical substrate suffix (as opposed to 'face' = -W1t-). Therefore, the
word must be compared with ethnonyms like �pU01tEe;, �OA01tEe; and is of Pre-Greek

a'lOouaa [f.] 'portico' (ll.); also a plant, cf. .. a'(8w. � PG(v)�

.VAR a'l8ouooa (Hdn. Gr. 2, 919).
.ETYM Generally explained as a ptc. of .. a'(8w meaning 'glowing, place where the sun
burns' or 'place where fire can be kindled', which is hardly convincing. A better
explanation is that of Fur.: 19754, who deems it a substrate word, as a technical term
of building, because of the form with -00-. The form .. aiowooa confirms this; it can
hardly be a mistake for a'l8ouoa (which would imply three mistakes), and it has -00-.
Of course, a folk-etymological connection with .. a'l8w is likely.
a'lOw [v.] 'to kindle', intr. med. 'burn (with light)' (ll.). � IE *h2eidh- 'kindle'�
vAR Only present.
.COMP Cf. .. Ai8lom:e;. On KaK-L8�e; see on KEYKEL s.v. .. KUyKavOe; .
.DER a18oe; [m.] 'burning heat' (E.) = Skt. edha- [m.] 'firewood', OHG eit [m.] , OE ad
'blaze, pyre'; ai8oe; 'sparkling, glowing', also 'dark-colored'; also a'l8wv, -wvoe; (ll.)
and aI80'\l (on the mgs. see Beekes Glotta 73 (1995-1996): 15-17).
a18oe; [n.] 'fire' (A. R.); ai8�ELe; 'sooty' (Nic.), also ai8�e; 'burning' (Cratin. 88), if this
does not stand for ai8�e; from ai8�ELe;; u'l8Lvoe; (H., EM).
Perhaps related is ai8oALKEe; 'pustule, pimple' (Hp., Gal.), cf. 1Wfl<POAU� 'bubble' for
the formation.
ai8uoow 'to stir violently' (Sapph., Pi.), also prefixed wiili av-, OL-, KaT-, 1tap-, is
usually considered cognate, but the development of meaning is strange. The verbal
noun a'l8uYfla 'spark, glow' (Plb.) did not undergo the metaphorical development of
ai8uoow, as opposed to ai8uKT�p 'stirring violently' (Opp.). But note that these are
late derivations. Also Ka-raT8u� 0fl�poe;· 6 KaTaL8uoowv 'which is floating down' (H.);
diff. Pisani Paideia 15 (1960): 245f.
a'l8ma [f.] name of a bird (see Thompson 1895 s.v.), also epithet of Athena, see Kock
Arch. f Religionswiss. 18 (1915): 127ff. but also Kretschmer Glotta 9 (1918): 229f.,
mostly explained as a color term, but rather a substrate word (Szemerenyi 1964: 207,
Beekes 1998: 25 on the suffix -ma.). On a'l8ouoa 'hemlock, Conium maculatum' (Ps.­
Dsc.) see CEG 4 (from 'black'). Cf. also .. ai8�p, .. ai8uAT], .. a'l8ouoa.
.ETYM Old PIE verbal root, of which the zero grade *h2idh- probably appears in
i8apoe;, i8alvw. Sanskrit has the root form idh-, with a nasal present i-n-ddhe 'to
kindle'. Thematic a18oe; can be of PIE date, cf. Skt. edha- [m.] 'firewood', OHG eit
[m.], OE ad 'glow, funeral pyre'. a18oe; [n.] 'fire' (A. R.) and Skt. edhas- [n.]
'firewood' are independent formations, since the Greek word is late. Remarkable
forms in other languages include Av. aesma- [m.] 'firewood', Lith. iesme 'id.', Lat.
aedes 'dwelling place, temple', aestas 'summer', aestus 'heat', and several Germanic
forms, e.g. OHG eit (see above) and ON eisa [f.] 'burning coals' .

aiKu�£l [v.] KaAci 'calls' (H.). � ?�
.ETYM Pisani IF 58 (1942): 243 compared it to Osc. aikdafed, which he interpreted as
'proclamavit'; this is highly dubious. The connection with Latv. aicinat 'to call' is
doubted by Pok.: 15. Is it a mistake for .. aiKuAAw?

aiKuAAw [v.] 'to flatter, fondle' (trag.), especially
said of animals. � ?�
VAR Only present.

.DER alKaAO<;' K6Aa� 'flatterer' (H.); alKCtAll' aTtCtTll 'deceit' (Zonar.).
.ETYM Looks like a denominative of the forms given in the glosses (unless these are
based on the verb). Etymology unknown.
CilKq<; [adj.] 'improper, unseemly'. <!I GR�
.VAR Att. alKq<; < * a- pK-q <;, next to a£LKq<; (Ion. poet.).
DER a£LK£ll1, alK£la, alKla 'unseemly treatment, disrespect'; a£LKl�w, aiKl�w, -OflaL 'to
maltreat', whence a'(KLafla (trag., Lys.), alKLafl6<; (D., LXX). a£LK£ALO<;, alK£ALo<;
(Horn., poet.) contain an enlargement of synonymous a£LKq<;, alKq<;.
ETYM Privative verbal adjective to £oLKa, duo E"iKTov < *we-w(o)ik-. In a£LKq<;, -£L- is
probably secondary after £LKCt�w, £LKWV, ete. See � £LKCt�W, � £OLKa.


aIKAol ' aL yWVlaL TOU �£AOU<; 'the angles of the missile' (H.). => alXflq.
aIKAov [n.] 'evening meal at Sparta' (Epich.). <!I PG(v)�
.VAR Ci"iKAov; <auv>aLYAla = auvaLKAla; AUKaLxAla<;· 6 AUK6�pWTO<; 'eaten by wolves'
(-�pOTO<; codd.).
.DER avalKA£La· Ci8£LTtVa 'suppedess' (H.). Also a[Kvov, O£lTtVOV 'meal' (H., Suid.);
£TtCt'(KAa (Pl.).
.ETYM Fur.: 139 points to clKAOV' O£lTtVOV and £LKA£l' O£LTtV£l (H.), and compares
iKv£lav· Tpo<p£la 'nourishment' and iKV£lO<;' TpO<p£U<;. 'P6OLOL 'feeder (Rhod.) (H.);
he correctly concludes that it is a substrate word. This may be reconstructed as
*(a)wiklln-: for the prothetic vowel, cf. aep oTt- / fl£POTt-; the interchange between A
and v may be secondary. A comparison with aiKCt�£L' KaA£l 'calls' (H.) or � alKCtAAW
is completely uncertain.
at\.lvo<; [m.] 'song of mourning' (trag.), sometimes as an adj. 'plaintive' (E. Hel. 171).
<!I ?�
VAR a'LALVa [adv.] (Call.).
.ETYM Etymology unknown. Boisacq's postulation of a Phrygian origin lacks support
(cf. � £A£y0<;) ' It appears to be derived from the interjection � a'( and � AlVO<;.

a'(AlOl => a£ALOL.
alfla [n.] 'blood' (ll.). <!I ?�
.COMP aLflaKouplaL 'offerings of blood' to the dead (Pi.); aLflCtAw\j! 'mass of blood'
(Hp.), see CEG 6 .
DER aLflCt<; 'flow of blood' (S.); diminutive aLflcmov, also name of a dish (Arr., M.
Ant., inscr. Cos, Miletus, ete.), aLflaTla 'Spartan blood soop' (Poll.). Derived
adjectives: aLflaT6£ L<; 'bloody' (epic, poet.); aLflaTll p 6<; (poet.), aLflllp6<; (Man.);
aLflaTwoll<; (Hp., Th., Arist., Hell.), aLflwOll<; (Luc.), see on � aLflwO£w ; aLflaTLK6<;
(Arist.), aLflCtTLvo<; (Arist.); aLflaA£o<; (AP, Nonnos); aLflwv (E.), aLflwv LO <; 'red as
blood' (Ath.); aLflaTlTll <; 'like blood' (Hp., Thphr.); aLflaTwTt6<; (E.), aLflwTt6<; (Ph.). In
a similar meaning compounds like £VaL fl0<;, u<PaLflo<;. Denominative verbs: l.
aLflCtaaw, -CtTTW 'to make or be bloody' (A.), whence late nouns aLflaYfl6<;, aLfla�L<;,
and adjectives aLflaKT6<;, aLflaKTLK6<;; 2. aLflaT6w (lA), aLflCtTw<1L<; (Gal.); 3. aLflaTl�w
(A., Arist.).


.ETYM alfla replaces the old lE word for blood £ap < *h,esh2-r. It has no established
etymology. The connection with OHG seim 'virgin honey', W hufen from the
uncertain PIE root *sei- 'to drip' (Pok. 889) is accepted by Weiss HSPh. 98 (1998): 3161, but cannot explain the Greek vocalism. Ace. to Sommer 1905: 29ff., it is related to
Skt. i$- 'refreshment'. See Szemerenyi Gnomon 43 (1971): 651; cf. also � alOVCtW,
� ixwp .
aiflamCt [f.] 'wall around a terrain', of stone (thus Hdt. 2, 138) or thorns (Od.). <!I ?�
.VAR Cf. aLflol' OpufloL AiaxuAo<; AiTvalaL<; 'copses, thickets (A. Aitnaiai) ' (H.) .
.ETYM Comparable with Lat. saepes 'hedge, fence', which has p for m. Fur.: 223 finds
the variation in other non-lE loans, e.g. II£v£aTaL / M£V£aTaL (Schwyzer: 333),
y£<pupa / Arm. kamurj. On the accentuation, see Scheller 1951: 87f.; on the meaning,
Picard Rev. arch. (1946): 68f.
aifluAO<; [adj.] conventionally translated as 'flattering', mostly said of words (Hes.);
ace. to Guntert 1921: 103, it means 'wily'; Weiss HSPh. 98 (1998): 31-61 points to the
semantics of spellbinding, e.g. in aLfluA lOL<1L A6YOL<1L // 8 £AY£L. <!I ?�
VAR Also aLfluALO<; (Od.).
.ETYM The suffix is also found in (1TW flUAO<; 'talkative'. A direct connection with
OHG seim 'virgin honey' is impossible (see on � alfla), as this etymon originally
denoted a 'thick liquid'. Acc. to Guntert (ibid.), aLfluAo<; derives from � aLflwv in the
same way as ayKwv relates to aYKuAo<;. This idea is integrated by Weiss HSPh. 98
(1998): 31-61 in his account of � Ifl£ PO <; and � aLflwv.

aLflwSiw [v.] 'to be set on edge', of the teeth, as caused by sour stuff (Hp.). <!I ?�
.DER aLflwola (Hp., Arist., Dsc.); whence aLflwO LCtw 'to have aiflwola' (Hp., Arist.),
whence aiflw OLaafl6<; (H.). aLflwoll <; is a back-formation in the sense of 'having
aLflwola' (Gal.). Further aiflwo ll <; 'bloody', to � alfla.
.ETYM The second element can hardly be separated from 60wv 'tooth' (Szemerenyi
1964: 81). It is difficult to assume that the first part is from � alfla, as we would expect
to see a trace of the suffIx -aT-. Solmsen 1909: 25ff. connected the first member with
Gm. *sai-ra- in Go. sair, OHG ser 'pain', ON sar 'wound', positing *aL-flo<; for Greek.
alflwv, -OVO<; [adj.] only in LKaflCtvOpLOV aLflova 8qPll <; (E 49), mg. sometimes glossed
as 'skillful', but rather 'eager' (see below). <!l IE *seh2i-mon- 'bond'�
.ETYM The word is found in the Thessalian names '!TtTtalflwv, ALflovo<; (see Bechtel
1921, 1: 203). Weiss HSPh. 98 (1998): 31-61 assumes a pre-form *seh2i-mon-, from the
root *Sh2i- 'to bind' which he also assumes to be present in � lfl£ P O<; 'longing, love'
and in � aLfluAo<;. The fact that the meaning 'eager' fits so well in this passage
confirms Weiss's proposal.
alvo<; [m.] 'meaningful words, praise' (ll.), also 'decision' (inscr.). <!I ?�
.VAR a'(vll (Hdt.). Cf. the primary formation in avalvo flaL 'to deny, refuse' (ll.), from
* ava-alvoflaL (cf. ava-v£uw), ace. to Bechtel 1914.
.DER aiv£w [v.] 'to approve, praise', also 'to decide' (ll., mainly epic Ion. poet.), fut.
-q aw, secondarily -taw, ete. (see Wackernagel 1916: 180f.); Att. has £TtaLV£W, Aeol.
(Hes.) a'(vll flL. From aiv£w: a'(vwL<; 'praise' (LXX, NT), a'(vll <1L<; (Ph.). Rare is



aivi�0!lm [V.] 'to praise' (Hom.); usually aiviaao!lm (-n-) (lA) 'to speak in riddles',
from 'to speak words full of content, i.e. difficult to understand'; thence with a'(vlY!la
'dark saying, riddle' (Pi.); thence aivlwaTwollC;, aivlwanaT�C;, aivlwaTiac;,
aivlwanKoc;; also aivlW0C; 'id.' (Att.); a'(v l�lC; 'id.' (Plot.). aivlKT�p 'who speaks in
riddles' (S.), aivlKT�C; (Timo), aivlKTllPiwC; (A.).
.ETYM Etymology unknown. Compared by Pok. 11 with Germanic words for 'oath'
(Go. aips, OHG eid), which is mostly rejected.

.COMP ainoAoc; 'goatherd' < * aiy -noAoc;, see � n£Aw and � �OUKOAOC; (cf. Meier­
Briigger 1992a: 92) ; thence ainoA£w [v.] 'to herd goats' (A., Lys., Theoc.), only
present; ainoAla [] 'herds of goats' (ll.), also -lOV [sg.]; ainoAlKoc; (Theoc.).
Further aiyi-�oToC; 'browsed by goats' (Od.); unclear the gloss ainoAoc;· Ko.nllAoC;
napa KunpiOlC; 'peddlar (Cypr.)' (H.), see Leumann 1950: 271ff.; to be rejected is
Latte's correction ai- ::: aE t.
.DER See � aiyic; 'goatskin'. Diminutive of a'(�: aiyiolov (Pherecr., Antiph.). On alYEC;'
Ta KU!laLa 'waves'. �wpldC; (H.), see on � aiylaA6c;. Connection with toponyms like
Aiyai, AiyruoC;, A'lYlva, etc., is at best folk-etymological.
.ETYM The compounds in -l- are unexplained (Heubeck IF 69 (1964) : 13-21 is
unclear); the type ainoAoc; is old in any case. a'(� is cognate with Arm. aye 'goat' (i­
stem), which is reconstructed as *h2eig-ih2 (Clackson 1994: 88-90 after Meillet). The
zero grade is often supposed in Av. lzaena- 'of leather', but we do not know if the
word referred to the skin of a goat. If the connection is correct, the word could be IE.
However, Skt. aja- 'goat' looks similar, but is formally deviant. This could suggest
that Avestan, Greek and Armenian borrowed the word from a common source,
perhaps Anatolian. On the distribution, see Mallory & Adams 1997 s.v. Connection
with *h2eig- as 'jumper' (Thieme 1953: 571) is rejected by Mayrhofer EWAia 1: 264,
since Skt. ejati did not have a palatovelar, nor does it mean 'to jump'. Cf. � a'(YlAoc;,
� aiyiAw'\!, � oi�a.


aivoc; [adj.] 'terrible' (ll.). � ?�

.VAR On enmv� see Leumann 1950: 258f.
.COMP Common as the first member of poetic compounds, but no derivatives.
.ETYM No etymology. Connected with Skt. enas 'crime' by Pok. 10. On the
expression aivo0Ev aivwc;, see Leumann l.c. See also LfgrE.
a'{vu!1at [v.] 'to take, seize' (ll.). � IE *h2ei- 'give'�

VAR Only present.
DIAL Myc. PN a3-nu-me-no IAinumenosl .
COMP Often with e�-, also in the verbal noun E�-mLOC; 'selected' (ll.).
.ETYM The verbal noun *ahoc; is at the base of � aiL£w. The same root is found in
ToB ai-, ToA e- 'to give (act.), take (med.)" and it was previously thought to be
reflected in Hitt. pai) 'to give', which was analyzed as *pe-ai-. Recently, however,
Kloekhorst 2008 has given a convincing alternative: an i-present to the root *h1p- 'to
seize'. Thus, the present root is eliminated as an example for PIE *a. In nominal
form the root is found in Oscan aeteis [] 'part (of a possession)'. YAv.
aetahmiiiius does not contain a noun aeta- 'punishment', but rather the pronoun
aeta- 'that'; cf. Fischer and Ritter MSS 52 (1991) : 9-13. See � araa, � aiTia, � 8[mTa.



a'{vw [v.] 'winnow' (Pherecr., Hp.), but see the glosses. � ?�

VAR Aor. �vm; pres. also QV£W (Ar. Fr. 694 (uncertain), Ath.), acpuv£w (Ar. Eq. 394
v.l.), acpllva· EKo,\!a 'struck', acp�vm' TO TaC; emla!l£vac; Kpl0ac; XEpal Tpl'\!m 'rubbing
by hand of the winnowed barley-corns' (H.); further aLVWV miaawv 'winnowing',
�vac;' KO,\!ac; 'having struck' and yo.vm ( Fuvm)- nEplmiam 'strip off the husk or
skin' (cod. -nTuam); see Solmsen 1901: 280 .
DER Bechtel KZ 46 (1914) : 374 compares the name of a phratry Faviom (Argos).
.ETYM Comparable with Lat. vannus 'winnowing-basket', OHG winton 'to fan', Go.
dis-winpjan 'AlK!lUV, to winnow'. The Germanic words seem to derive from the word
for 'wind' (cf. Lat. ventilare 'to fan'), but aLVW has no trace of the -t-. Derivation of
the Greek word from *h2ueh1- seems to be excluded by yo.vm, which has no vowel
before the F. QV£W has been explained from *a-Fav-£w (Solmsen 1901: 272) , which
beside yo.vm would imply a non-IE word. Note that the exact meaning of tlIe word is



a'(�, aiyoc; [f.] 'goat', rarely msc. (ll.). Also a water bird (Janzen 1937: 17, a meteor

(Arist.) and a star (Aratos). � IE? *h2eig- 'goat'�
.DIAL Myc. a3-ki-pa-ta laigi-pa(s)tasl (?) 'goatherd'; a3-ki-po-de, interpretation


aioAoC; [adj.] 'agile, glittering, variegated' (ll.). � ?�

.DIAL Myc. a3-wo-ro IAiwolosl name of a cow.
.COMP As a second member, e.g. Kopu0aioAoc; 'with glittering (colorful) helmet'.
.DER Denominaitve aioUw [v.] 'to move quickly to and fro' (u 27), (med.) 'to change
color' (Hes. Se. 399) , 'to make colorful' (Nic. Th. 155) ; aioA£w ::: nOlKiAAw (PI. Cra.
409a) , aioAllC1lC; 'rapid movement' (sch. Pi. P. 4, 412) ; aioAi�w 'to trick with words' (S .
fr. 912) , aioAla!la 'varied tones' (S. Iehn . 319) ; aioACto!lm 'to be restless' (Hp. Mu!. 2,
174b, uncertain). aioAiac; [m.] fish name (cf. Stromberg 1943: 23, Thompson 1947
s.v.), aiOAElOC; (EM), aioAioac;· nOlKiAoUC;, LaXdC; 'variegated, quick' (H.). PN A'(OAOC;,
EN AioAElC;.
ETYM Etymology unknown. Benveniste BSL 38 (1937) : 107 connected aiwv, Skt. dyu­
'vital force' (formally improbable); Risch Mus. Helv. 29 (1972) : 97 argued that the
original meaning was a color. On � ai£AoupOC;, see s.v. For the type *Cai-CoR-, cf.
� aiwpa, � aiovo.w.

aiovaw [v.] 'to moisten, bathe (a wound)' (Hp.). � ?�

.DER Verbal nouns aiovllC1lC; and aiovll!la.
.ETYM Etymology unknown.



aimJC; [adj.] 'steep, sheer' (ll., mostly epic and poet.) � PG(v)�

VAR A different stem in aino. (aim't p£E0pa El 369) and ain�v (noAlv . . . ain�v y 130,
etc.), maybe a metrical device.

alpa 1


.DER ai1t�El<; (Horn.), is an enlargement of aimJ<;, see Schwyzer: 527. Further alrro<;
[n.] 'steepness, precipice' (E.), whence aim:tVo<; < *airrw-vo<; 'steep'.
.ETYM � ahjla probably belongs here, too. Furthermore, Fur.: 158 connects it with
e�al<pvTj<; and � e�arrlvTj<;, as well as � a<pvw and � a<pap, which is highly convincing.
Variations shown here include rrl<p, labial I '\' (cf. � OE<PW I 8£,\,w and �ITTaKo<; I
'\'lTTaKo<;), and anticipation of a palatalized consonant *ap'- as ai-.

aIO"a [f.] 'share, destiny, decree' (ll., epic and lyr.). <! IE *h2ei- 'give, take'�
.DIAL Mye. aJ-sa laisa/; also found in Arc.-Cypr.
.DER a'(0Lo<; 'auspicious, opportune'; also with ev-, e�-, KaT-, rrap-, whence ai0LooflaL
[v.] 'to take as a good omen' (PIu., App.); a'(0Lfl0<; 'destined, fitting' (Horn.),
eval0Lflo<;, uval0Lflo<; (Emp.).
Thence prefixed denominative UV-aL0LflOW [v.] 'to consume (the apportioned share)
(Ion.), whence UVaL0LflwflaTa 'expenses' (Hdt.); KaTaL0Lflow [v.] 'to consume
entirely' (corn.). Adjectival abstract ai01fllaL rrAouTou 'the due apportionment of
wealth' (A. Eu. 996). See on � ai0Lflvaw, � ai0Uflvaw, � ai0Uflv�TTj<;.
Some PNs: A'(awv, ALala<;, etc.
ETYM alaa is from the root seen in a'(vuflaL, derived with -la from a form in -t­
found in Osc. aeteis [] 'part', Gr. *aho<; (see � aiTEw), a'iTlo<;. An ablauting
root shape could be found in �'(a0aa6aL · KATjpoua6al. AEa�lol (H.) .


aIpa 1 [f.] 'sledgehammer' (Call. fr. 115, 12), a<pupa, c#vTj 'hammer, axe-head' (H.),
a<pupa (Et. Gen.). <! ?�
ETYM Unknown. Acc. to Schwyzer: 474, it is from � a'(pw (improbable).


alpa 2 [f.] 'rye-grass, darnel, Lolium temulentum' (Thphr.). <! ?�
VAR Often plur. alPaL.
DER a'(plvo<; 'of rye-grass' (Dsc.), aipwoTj<; 'mixed with rye-grass' (Thphr.).
Denominative e�-aLPOOflaL [v.] 'to change into rye-grass' (Thphr.).
.ETYM Specht KZ 66 (1939): 12 connected it with Skt. eraka- [f.] a kind of grass,
assuming that both languages borrowed the word from an Oriental language; this is
rejected by Thieme 1953: 586. Berger WZKSS 3 (1959): 48 thinks that the Sanskrit
word is of Austro-Asiatic origin. On eraka-, see now Klaus MSS 57 (1997): 49-64; see
� aipomvov.


aipEw [v.] 'to take, grasp, seize', med. 'to take for oneself, choose' (ll.). <! ?�
.VAR Except for late forms like uv-npTjaa (Q. S.), � EAElV is used as a suppletive
.DIAL Cret. a[Atw is a contamination of a[pEw and EAElv; Pamphyl. uyAta6w from
uypEw and EAElv; for other such forms see Vendryes 1938: 331ff.
DER atpWl<; 'capture, choice, party, philosophical school (whence heresy) (lA),
a[pE0Lflo<; 'pregnable' (X.); a[p£To<; 'what can be taken or chosen' (lA), a[p£TlKo<; 'to
be chosen, causing schisms' (late); a[pET�<; 'who chooses' (Vett. Val.), Ka6aLpETTj<;
'destroyer' (Th.), a[p£Tl<; [f.] 'who chooses' (LXX), back-formation to a[p£Tl(w [v.] 'to
elect' (HelL), which is a denominative to a[p£To<;. From a[p£Tl(w also a[p£TlaT�<;
'adherent' (Plb., D. L.).
.ETYM No etymology.

aip6mvov [n.] 'sieve' (Ar. fr. 480). <! GR?�
.VAR Cf. aipomvov· aKoTElVOV, Kat KoaKlVov ev tP rrupot a�60VTaL 'dark, also a sieve
through which wheat is sieved'; An. Bk. 359, 24 continues with ll1tEp TOU Ta<; alpa<;
OLEA6Elv 'because it passes through the rye-grass'; also aipomvov· TO UpaLOV
KoaKlvov· rrapa TO Ta<; alpa<; rrolElv urrElVaL Kat Xwpl(ElV � OLa TO a'(pElV TOV rrlvov a
eaT! TOV purrov 'a porous sieve, after its making the rye-grass go apart; or after its
removing the rrlvov, i.e. the dirt' (EM 38, 42), the first part of which is clearly folk­
.ETYM DELG thinks that it consists of a'(pw 'to remove' and rrlvo<; 'filth', like the
latter part of the final gloss; for the type of compound, see Schwyzer: 442.
Alternatively, it may be a Pre-Greek word reshaped by folk etymology.
a'ipw => UelpW 1.



a'iO"aKoc; [?] 6 T�<; oa<pvTj<; KAaOO<;, av KaTExovTE<; uflvOUV TOU<; 6£Ou<; 'the branch of
the sweet bay; while grasping these, the gods were praised' (H.). Cf. PIu. Mor. 615b.
Ace. to EM 38, 49 it indicates the bird epI6aKo<;. <! PG (s,o) �
.ETYM Etymology unknown. The word is Pre-Greek (or Anatolian), ace. to Nehring
Glotta 14 (1925): 183 and Krause KZ 67 (1942): 2144. Note the initial ai-, intervocalic
-a-, and the suffix -aK-. See � aiaaAwv.

aiCJ(lAWV [m.] kind of falcon (Arist.); see Thompson 1895. <! PG (V) �
.VAR Cf. aiaapwv· £100<; [EpaKo<; 'id.' (H.).
.ETYM Fur.: 387 gives it as Pre-Greek form with the variation pi A, of which he has
more than 30 examples.
A'i0TJrroc; [m.] HN in Mysia (ll.); also PN (ll.). <! PG�
.ETYM Fur.: 234 compares ai(Tjo<;. No doubt a Pre-Greek name.
al0"8avoflat [v.] 'to perceive, apprehend, note' (lA). <! IE *h2eu-is- 'perceive'�
VAR Incidentally pres. a,(a60flaL (Th.); aor. aia6Ea6aL, fut. aia6�aw6aL (lA).
.DER a'(a6Tj0L<; 'perception, knowledge' (Hp., Pl., etc.), both the act and the object of
perception (cf. E. fA 1243, Arist.); aia6TjalTj (Aret.) a'(a6Tj0L<;; aia6TjTo<; 'perceptible'
and aia6TjTlKo<; 'able to perceive', both mainly philosophical terms; aia6TjT�pLOV 'one
of the senses' (Arist.), aia6TjT�<; [m.] 'who perceives' (Pl.).
.ETYM Interpreted as PGr. *awis-tI'-, and connected with � utw 'to perceive, hear' <
*awis-je/o-. A similar pre-form is found in Lat. audio 'to hear' < *h2eui-dhh,-ie/o- (see
De Vaan 2008 s.v.), oboedio 'to obey'. It is probable that the Greek suffIx -6-, which
builds resultative verbal forms, is from *dhh,- 'to do, etc.' as well. Further related to
Skt. avis·, Av. auuis 'manifestly', OCS (j)ave 'evidently', which are adverbial forms in



ul0"8wv [ptc.] 'to breathe out, exhale' (ll.). <! ?�
.VAR Or rather *u·la6wv (IT 468), &."la6E (y 403), of 6uflov.
.ETYM We can connect it with a·lov (== TO UrrE1tV£OV Eust.) in erreL <plAov &."lov �TOp (0
252), but much remains uncertain. See Bechtel 1914 and DELG.




i(foow [v.] 'to move quickly, dart, rush (upon) (ll.). � ?�
.VAR Pres. �aaw (Pi.), �HW (Att.), fut. <'it�w. The 0.- is always long in Horn., except in
iJ1taT�£l (<1> 126), see Chantraine 1942: 110, and in a[�n (A. R. 3, 1302); elsewhere it is
mostly short.
.COMP As a second member in nOAu-(il�, KOpUe-C(l�; also in � TPlX(ilK£<; ?
DER o."(K� 'rush' (0 709); root noun (il� in o.V£flwV alKa<; (A. R. 4, 820), or o.lKa<;?
Both a and l are long.
ETYM Unexplained. Comparison with Skt. vevijyate 'to raise, flee, move quickly' is
formally impossible (no trace of a F in Homer, and the long L remains unexplained).
Danielsson IF 14 (1903): 386£f. reconstructs *aiF-LK- and compares � aioAo<;.


a'iouAoc.; [adj.] 'unseemly, evil' (Horn.), in a'iauAa p£�£lV, eio£vm, flue�aaaem (opposed
to a'iatfla). � PG?�
.COMP aiauAo-£py0<; (Max. Astrol.) after a'iauAa P£�£lV (Horn.).
.ETYM The overall appearance is Pre-Greek: initial ai-, intervocalic -a-, suffIx -UA-.
Cf. � o.�aUAO<;.
aiou�vaw [v.] 'to be ruler' (Horn.). � PG (v) �
.VAR aiauflv�nlP (0 347) has a variant aiau(l)�T'l P now preferred by West Clatta 77
(1999): 119f. Also PNs A'iauflvo<;, Aiau�T'l<; (Horn.).
.DIAL Meg. aiatflvaw .
DER aiauflv'lT�p (0 347 v.l.), aiauflv�T'l<; (aiatflvaTa<;) title of a high magistrate in
several towns (inscr., Arist.), in Homer e 258 a referee of games; fern. aiauflv�n<;
(Suid.); aiauflv'lT£la 'office of aiauflv�T'l<;' (Arist.), Verbal noun aiauflv'lTlJ<;
(Miletus). Further aiauflvlov �OUA£UT� p LOV in Megara (Paus.), from the verb or
from *a'(auflvo<; .
ETYM Previously derived from � a[aa, a'(atfl0<;, *a'iatflvo<; by Solmsen 1909: 36£f. and
Fraenkel 1910: 172f. However, Chantraine 1933: 216 and von Blumenthal 1930: 33
assume a Pre-Greek origin, which must be correct: it explains the interchanges fll F
(cf. Fur.: 244) and ul l. The word and its derivations are reminiscent of � Ku�£pvaw.
Further details are in the LfgrB.
Deroy Ant. class. 26 (1958): 404-410 compares Lat. aerumna 'task, distress'.



aiouqHoc.; => o.aUCP'lAO<;.
atoxoc.; [n.] 'shame, ugliness', plur. 'disgraceful deeds' (ll.). � ?�
.VAR Comp. aiaxlwv, superl. a'(aXlaTo<;.
.DER aiaxpo<; 'dishonoring, ugly', denominative aiaxuvw 'to dishonor', med. 'to be
ashamed' (ll.), back-formation aiaxuv'l 'shame' (lA). PN AiaxuAo<;, perhaps an
enlargement of an old u-stem. Further derivatives: 1. from aiaxpo<;: aiaxpOT'l<;
'ugliness' (PI. Carg. 525a, Bp. Bph. 5, 4), aiaxpoaUv'l (Tz.). 2. from aiaxuvw:
aiaxuvT�p 'violator' (A. Ch. 998), aiaxuvT'lAO<; 'timid, shy', also 'disgraceful' (PI.,
Arist.), aiaXUVT'lAla (PIu.); it has -T- from the opposite o.v-alaXUVTO<; (Ale., Att.),
whence o.vmaxuvTla, -T£W, -T'lfla; secondary aiaxuvTo<; (Ps. Phoc.). Rare
aiaxuvT'lpo<; and aiaxuvnKo<;.

.ETYM The older comparison with Go. aiwiski [n.] 'aiaxuv'l' is generally abandoned
in view of obvious formal difficulties. De Lamberterie 1990: 835-840 plaUSibly
compares � a'iooflm, positing * aid-sk- for Greek next to * aid-st- in Go. aistan.
lifTuc.; [m.] 'eromenos' (Ar.), also a fish (pap. Tebt. 701, 44). � ?�
.VAR Fern. run<; (Hdn. Gr., Alem. 34 Page). Also o.£LTav' TOV ETalpov 'companion' .
l\plaTocpav'l<; O£ TOV epwfl£vov 'eromenos' (Ar. fr 738; also Theocr. 12, 14, where it is
called Thessalian) .
.DIAL A Doric or Thessalian word.
.ETYM Uncertain. From � a[w 'to hear' ace. to Diels Herm. 31 (1896): 372 and Bechtel
1921, 1: 201; see also Arena Riv. fil. class. 96 (1968): 257f.

ahEw [v.] 'to ask, request, beg' (ll.). � GR�
.COMP Often prefixed with o.n-, e�-, nap-, ete.
.DER 1. ah'lat<; 'demand, request' (lA), aiT�atfl0<;; 2. ah'lfla 'demand, claim' (PI.,
Arist.), aiT'lflaTlKo<; and aiT'lflaTwO'l<;; 3. aiT'lT�<; 'requestor' (pap., D. C.); aiT'lTlKO<;
(Arist., D. L.), 4. aiTl�w = aiT£w (epic since Od.) .
ETYM A denominative of*aho<;; see � a'ivuflm, � a[aa and � ahlo<;.

ahlOc.; [adj.] 'guilty, responsible' (ll.). � GR�
.DER Thence (or directly from *alTO<;): aiTla [f.] 'responsibility, guilt, cause;
accusation', also 'disease'; thence denominative ainaoflm 'to accuse, charge with',
secondary aina�oflm (X., D. C.).
To ainaoflm: aiTlaat<; (Antipho, Arist.) and aiTlafla (A., Th.) 'accusation, charge';
alnaTo<; (Arist., Plot.) 'having a cause' (TO ainaTov 'effect' as opposed to TO ahlov
'cause') is rather directly from aiTla because of the meaning; from TO ainaTov, the
grammarians created � ainaTlK� nTWat<; 'accusative case', so properly 'case of what is
effectuated' (WackernageI 192o-1924(1): 19).
From aiTla (or TO ahlov): ainwo'l<; 'causal', philosophical term (Hell. and late),
likewise aiTlwfla (pap., Act. Ap.) = aiTlafla, and with the same vocalism aiTlwat<;
(Eust.) aiTlaat<;.
.ETYM ahlo<;, aiTla and aiT£w were derived from *alTO<; 'share' (see � a'ivuflm,
� aiT£w), which is semantically understandable. The suffIx - lO<; may have been added
to aiT- after the change of *ti > at.

a'{cpv'1c.; [adv.] 'suddenly' (E. lA 1581, Hp. Int. 39). � PG (V) �
.DER More common as well as more archaic is e�alcpv'l<; (Horn., Pi., trag., etc.). The
adjective aicpvl8to<; (A., Th., Arist.), on the other hand, is more common and more
archaic than e�mcpvlOlo<; (PI., GaL). Adverbial forms aicpv'l0l<;, -Mv (Hdn.).
.ETYM Related to � ahva, s.v., and also to � acpvw, � acpap, � e�an lv'l<; ete.

aix�q [f.] 'point of a spear, spear' (ll.). On its use in Homer, see Triimpy 1950: 52ff. �IE
*h2eik-(s)m- 'spear'�
.DIAL Myc. a3-ka-sa-ma laiksmans/.
.COMP aiXfl-aAwTo<; 'prisoner of war' (Pi.), whence fern. aiXflaAwTl<;, adj.
aiXflaAwnKo<;, abstract aiXflaAwala. Thence denominative verbs aiXflaAwTl�w and



uiXflUAWT£UW (Hell. and late); from uiXfluAWTL�W: uiXfluAwnaT�e; and
.DER uiXfl�£le; 'armed with a spear' (A., Opp.); uiXfl'lT�e; 'spearman, warrior' (11.),
uiXfl'lT& (E 197), fem. u'lXfl'lTle; (EM); with secondary suffIx uiXfl'lT�p (Opp., Q. S.,
Nonn.); uiXfl'lT�pLOe; 'armed with a spear, war-like' (Lye. 454 verse-fInal).
Denominative uiXflCt�W [v.] 'to throw the spear, to arm with a spear' (11.) .
ETYM The Mycenaean form proves PGr. *aiksma. The word is connected with
a.LKAOl· ui yWVLUL TOU �eAOUe; 'points of the arrow' (H.) and with Lith. iesmas, OPr.
aysmis 'spit' which may derive from *h2eik-(s)m-. The original meaning must have
been 'point'. Within Greek, we further fInd Cypr. iKfluflevoe; 'wounded' (Ruijgh 1957:
136), iKTeu· UKovnov 'javelin' (H.), and perhaps � 'lKLUp 'near'. Uncertain is the
appurtenance of'(y8te; [f.] (Sol.), 'lYO'l (Hp.) 'mortar', but cf. Fur.: 32l.

Stadtmuller Saeculum 2 (1951): 315ff. A neuter u-stem is found in Skt. ayu-, Av. aiiu
'(life)time', OAv. gen. yaos, dat. yauuoi < *h2oi-u, *h2i-eu-s, *h2i-eu-ei. Latin has
thematicized aevus < *h2ei-u-o-; Gothic has an i-stem aiwins [] . An old
derivation is Lat. iuvenis, Skt. yuvan- 'young man' from *h2iu-Hen- with the
Hoffmann suffIx ('having vital strength'). Derived from this are Lith. jaunas, OCS
jun'b 'young' and Go. jund 'youth' < *h2iu-Hn-ti-.


aI'Va [adv.] 'quickly, suddenly' (11., poet.) <!! PG�
.DER ui'V'l poe; 'quick' (11., Pi.).
.ETYM Sommer IF 11 (1900): 243 connected the word with � ui1tue; 'steep' as *ui1t-a-u;
the forms can also be understood in substrate terms (with Fur.: 158) as alternation of
a labial with 'V. Furnee further connects it with E�ULcpV'le;, � E�U1tLV'le;, � acpup, � acpvw,
CilW [v.] 'perceive, hear' (11.). <!! IE *h2euis- 'perceive'�
VAR Ipf. a'iov (see below); verbal adj . E1t-Ct'LaToe; 'perceived, detected' (Hdt.) to
E1tUtW, E1t<�.W (Att. prose), whence aor. E1tfiau (E1t�·Lau).
.DIAL Cypr. fut. awiyesomai.
.ETYM The ipf. mOV was considered by Schulze KZ 29 (1888): 251ff. to be an original
aorist, from which a present Mw was formed secondarily. Schulze found traces of an
original pres. *UeLW in a£l' UKOU£l, an£' uKouauT£ (H.), and in E1t-q.£lV (E. HF 773),
but this is diffIcult to fIt in with the etymology prevailing today: a'LOV < *awis-e/o- is
generally connected with Skt. ay!? [adv.] 'evidently, manifestly' and OCS Wave 'id.',
so the Greek verb is probably a denominative from this adverb. As Kloekhorst
recently showed, the Hitt. verb au-i / u- 'to see' (see Kloekhorst 2008 s.v.) can also be
connected with it, since in o-grade forms *h2ou-, the laryngeal would regularly be
lost. Noteworthy, though improbable, is the connection by Szemerenyi Glotta 38
(1960): 243 with the word for 'ear'. Within Greek, compare aor. naeofl'lv, pres.
� uiaeCtVOflUL, with -e- indicating the completion of a process.

uiw - uiel.
aiwv, -WVOC; [m., f.] '(life)time, long period of time, eternity' (11.). <!! IE *h2ei-u- 'time of
living, well-being'�
.VAR uiev [adv.] 'always'.
DER UiWVlOe; 'enduring, eternal' (Pl., Hell., NT), UiWVlOT'le; 'perpetuitas' (gloss.);
UiWVL�£lV 'to make or be eternal' (Dam., Phot., Suid.), uiwvlaflu 'perpetuation,
monument' (Ostr.).
ETYM From *uiFwv, an n-stem also found in the old locative � uiev 'always', which
coexisted with the s-stem in uiw, uiee;, � Ui£L 'id.'. See also � oU. On the meaning, see



uiwpa [f.] 'swing, hammock, noose, halter' (Pl.). <!! ?�
.DER uiwpew 'to raise, hang' (Pi., lA), also -eoflUL 'to hand, hover', also prefIxed with
auv-, lntep-. Thence uiwp'lme; (mainly medic.), auv- (Pl.), lntep- (Hp.); uiwP'lflu (E.
[lyr.], Lyc.).
.ETYM Previously, an intensiv� (iterative) verb *FUL-Fwp-ew was reconstructed, from
which *FULFwpu > uiwpu was a derivation. This type is not accepted anymore.
Taillardat RPh. 57 (1983): 21-25 assumes *h2uor-eje- > *uFwpew (formation as in
1tWAeW, etc. to the root of ueLpw 'to hang'); reduplication would then have resulted in
*uFuFwP-, which would have given *UFULWP- > uiwp- (like *UFULPW > u'lpw). The
reduplication with uF-uF- seems uncertain to me, just like the development to *uFuj­
and its continuation as (UF)-UL- before a vowel.
AKacSTJ!10C; [m.] name of a hero. <!! ?�
.DER AKUO�fl£lU the gymnasium in Athens where Plato taught, the Platonic school
(Ar.), h£KUO'lfl£lUe; (Att. inscr., see LSJ Supp.) .
.ET�M Generally identifIed with the fIrst element of EKCt£pyoe; (which derives from
*uek-m); and with the PN Boeot. Fh£Kuoufloe;, Thess. FK£-; in Attic this form may
have been 'EKCtO'lfloe; (D. 1., St. Byz.); but see Lejeune 1972: §2543, who objects that
the aspiration was lost.
Fur.: 309 separates it from these words and connects the Lydian TN AKUOuflLe; and
the Carian PN j\xrao'l floe;. Cf. also TupKovo'lfloe; (Cilicia)? However, the names in
Greece seem to have (had) a F-, of which there is no trace in the Anatolian names.
The meaning of -oufl- is unknown. If the word was Pre-Greek, the varying vocalism
can be better understood (assimilations are rare in Greek).
aKUlva [f.] 'spike, prick, goad' (A. R.), Also 'ten-foot rod' in Thessaly (Bechtel 1921, 1:
116, 204), cf. aKULVU 8£ Ean fleTpov O£KCt1tOUV GwaCtAWV £upeflu (sch. A. R. 3, 1323;
Call. fr. 24, 6). In Egypt a measure of 100 square ft. (Hero, pap.). <!! GR?, PG?�
.ETYM Traditionally derived from the n-stem � aKWV with the suffIx -lU. However, it
may also contain the Pre-Greek suffIx -ULVU (see Fur.: 171"7) added directly to the
stem UK-. The measure is in origin the same word; for the semantics, DELG
compares KCtAufloe;, Lat. pertica, MoFr. perche.
UKaKUAtc;, -t6oc; [f.] name of several plants (Dsc.). <!! PG(v)�
.VAR Cf. UKUKUAALe;' aveoe; vupKLaaou. Kp�T£e; 'flower of narcissus (Cret.)' (H.) .
KUKUALe;' VCtpKlaaOe; (H.), KUKKUALU = aTpuxvoV lntvwTlKOV 'sleepy nightshade,
Withania somnifera' (Dsc. 4, 72 and 122). Further KCtYKavov = KUK(K)UALU
'Mercurialis tomentosa' (Gal., Paul. Aeg.) .


ETYM Frisk assumes an Oriental origin, possibly Egyptian, but why? Fur.: 371, 277
(see also 138) compares KUKUALC; and KCtYKUVOV, variants which prove a Pre-Greek
origin. Cf. also � CtKUKlU.

CtKCtKTJT« [adj.] epithet of Hermes (11., Hes.) and Prometheus (Hes.) , of unknown mg.
<! ?�
.DER CtKUK�moc; (CalL).
.ETYM If the glosses CtKUKlac;' ouvlac; and CtKUKl£1· <1vVl£l are reliable, the word could
mean 'ouv£T6c;' (Hoffmann BB 17 (1891): 328). DELG rejects the glosses without
reason and assumes a meaning 'benevolent'. Risch 1954: 395f. thinks it was built on
aKuKoc;, CtKCtKUC; after f!TjTl£TU (which is not easy). See also Fraenkel l956b: 168, and
LfgrE .
UKUKlU [f.] name of a tree or plant, 'acacia' or 'Genista acanthoclada' (Dsc.). <! PG (v) �
.ETYM Probably a substrate word; cf. � CtKUKUAlC;. Fur.: 321 compares KCtKTOC;. There is
no reason for an Oriental origin, as DELG assumes. Kramer ZPE 97 (1993): 146
compares Coptic KUK£, K£K£, K£Ka 'dark', the color of the wood of the acacia; the Ct­
would have been taken from aKUVeOC;. This is unconvincing.
UKUAUppdTTJC; [adj.] only in E� CtKUAUppdTUO �uevpp60v 'OKwvol0 (H 422, l' 434).
<! GR�
.COMP Similar CtKuMppooc; (Orph.).
.DER The adverbial first member is only sparsely attested (Hes., Sapph.), and in
glosses like CtKuA6v· �ouxov, TIpaov, f!UAUK6v 'quiet, mild, soft' (H.), CtKUAWC; [adv.]
(Eust.) .
ETYM From CtKUAU-PP£FE-TTjC;, a compound of CtKUACt and � pEW with suffixal -TTjC;.
Meier-Briigger Glotta 73 (1995): 9-11, derives the first member from the root *kelh2-,
seen in KEAUOOC; as 'rauschend dahineilen', and interprets Ct-KUAUPP£lTTjC; as 'kein
rauschendes Fliessen habend, still fliessend'. See � CtK� 2.

UKUAqq>TJ [f.] 'stinging nettle, sea anemone' (Eup.). <! PG? (v) �
VAR Thphr. (HP 7, 7, 2) has CtKuMq>Tj.
.ETYM Unknown; cf. Thompson 1947 s.v. There is no reason to suppose the influence
of O.KUVeU, etc. (Frisk). Semitic etymology in Lewy 1895: 50. Suffixal _bh_ is quite
common in names of trees and plants. The variation v/ Tj could point to a Pre-Greek
word, although there are no clear parallels to it.

aKuvOu [f.] 'thorn, thistle', name of different thorny plants (Stromberg 1940: 17), also
'backbone, spine' of fish, snake, or man (Od.). <! PG?�
VAR Note aKUVeOC; [m.] 'acanthus' (Acanthus mollis) .
DER Many derived adjektives: CtKCtVelVOC;, CtKUVeWOTjC;, CtKUVelK6C;, CtKuveTjp6C;,
CtKUVe�ac; 'provided with thorns, etc.'. Further diminutive CtKCtVeLOV; CtKUVelUC; kind
of shark or grasshopper (cf. Stromberg 1943: 47, Stromberg 1944: 17); CtKUVelC; name
of a bird (,goldfinch' or 'linnet', cf. Thompson 1895 s.v.), also a plant name;
CtKUVeVAAlC; bird name (Thompson s.v.), CtKUVelWV 'hedgehog', CtKUVeEU a plant,



CtKUVe£WV and -ewv 'thorny break, spinetum' CtKuVeTjA� mg. unknown.
Denominative verb CtKuve60f!aL 'to be thorny' (Thphr.).
.ETYM The basic meaning is 'thorn', whence 'backbone, spine'. Usually, aKuvoc;
'pine-thistle' is considered basic, but a connection with aveOC; (as *akan-anti'o-; see
Frisk) is improbable. AnalYSis as a compound *aK-UVeU 'sharp flower' (Kretschmer
1896: 403 A. 1) is a type of etymology of the past. Belardi Rend. Aee. Line. 10 (1955):
309-331 assumes an lndo-Mediterranean substrate word, connecting Skt. kaflt(h)a-,
but such combinations with Sanskrit are mostly incorrect, and tlIe lndo­
Mediterranean hypotlIesis is quite doubtfuL Most probable is a Pre-Greek substrate
element, though in this case there is no positive indication except for the ending in
short -u (see Pre-Greek); in this respect, there is no reason to assume a secondary
Greek formation (as per DELG). Niedermann Glotta 19 (1931): 8ff. connected it with
CtKUAUVelC; = CtKUVelC; (Ar.), by metathesis of *CtKuveuAlC;.

aKuvoc; [m.] a thistle, 'Atractylis gummifera', 'dorniger Fruchtkopf (Thphr.).
<! PG? (s) �
.VAR Also aKuv, -VOC; (LXX).
.DER CtKUVlK6c;, CtKUVWOTjC;, CtKUVl�W (all Thphr.), CtKCtVlOV (H.) .
ETYM For the formation, cf. TIACtTuvOC;, pCt<puvoC;, TIlJUVOC;, etc.; the word is mostly
derived from tlIe root CtK- 'sharp', but the suffix -UVOC; rather points to a non-IE word
(words like aKwv, CtK6vTj confirm that the -u- is foreign).

UKUpqC;, -tc; [adj.] 'small, tiny' (Ar.). <! ?�
.VAR Mostly in fixed expressions, e.g. EV CtKup£1 (XP6vov), CtKUp� 'a moment', of time
(Ar.); also KUTETIWOV CtKUp�C; 1"<ji 8ta 'it was a hair-breadth escape'; OUK CtKUp� 'not
at all'. A form aKup is attributed to Antiphon (Taillardat 1962: §248).
.DER CtKUPlaLOC; id. (D.), cf. Chantraine 1933: 49.
.ETYM Traditionally derived from Kdpw, EKCtpTjV 'to cut' as 'too short to cut': TO
�puXU 0 OUO£ K£1paL 0'[6v 1'£ (H.); this is doubtfuL Perhaps ,(not even a) louse'? See
� CtKupL
UKUpt [n.] 'mite' (Arist.). <! PG?�
.ETYM Fur.: 371 connects it with KCtpVOC; = <pedp 'louse' (H.), which is quite attractive .
It is usually connected with � CtKUp�C;, s.v. DELG suggests a contamination of CtKUp�C;
with K6plC; 'bug'. I would rather think that K6plC; is cognate, as a substrate word, with
prothetic vowel and u/o interchange.
aKupvu [f.] . 8Ct<pvTj 'sweet bay' (H.). <! PG?�
.ETYM The word has been connected with � aKU<1TOC; 'maple' (and further with OHG
ahorn), but this must be explained differently. aKupvu is most probably a substrate
word (note the sequence -pv-) .
UKclpVUV => CtXUpVWc;.
aKupov => ayxpuv.
UKUP0C; [m.]? . <1Tjf!ulva TOV EYKE<PUAOV � T�V K£q>UA�V 'brain, head' (EM 45, 13). <! ?�



.ETYM Cf. eYKapo<; and 'lYKpO<;, with the same mg., which point to EV and Kap'l
'head'. It would be the only relatively certain instance of *h,nC- yielding u-, but
Nussbaum 1986: 72f. remains sceptical, as one would expect *UKpO<; instead of
UKapo<; if the form is old.
uKaoKa => UK� 2.
UKaO"-rO<; [m.] . � ocpEvoaflvo<; 'maple' (H.). <!I PG(Y)�
ETYM We may posit *aKap-aTo<; and assume that the word is cognate with Lat. aeer,
-ris 'maple', OHG ahorn (which is sometimes connected with � aKapva · oacpv'l H.,
s.v.), and Gallo-Rom. *akar(n)os 'id.' (Hubschmied Rev. eelt. 50 (1933): 263f.); see
Pok. 20. For the formation, we may compare 1tAaTavloTo<;, but the derivation from
* -id-to- (cf. Chantraine 1933: 302) may well be wrong. Since plant names are often
borrowed, and the formation is unclear, we may envisage a substrate origin. Fur.: 371
compares KaoTov' �UAOV. A8afla.v£<; 'wood' (H.), and for the meaning ocpEvoaflvov,
�UAOV (H.), O.c. 164. A further comparison with KOaTOV 'wooden parts of a wagon'
(o.c. 343) is less certain.

uKa-ro<; [f., m.] 'light vessel' (Thgn.), 'boat-shaped cup' (com.). <!I ?�
.DER Diminutives uKaTlov, which also denotes a kind of women's shoe (Ar.), and
UKaT'lvaplov (Olsson AfP 11 (1935): 219); further uKaT£lo<;, -ra UKaT£la (sc. [aTla)
'small sails' (X.); uKa-rl<; [f.] 'millipede' (Steph. Med.), see Stromberg 1944: 11.
.ETYM Probably a technical loanword. Often connected with UK- 'sharp' (see � UK�),
but without any obvious reason. Winter 1950: 12 connected it with K'lT�V'l' 1tAOIOV
flEya w<; K�-rO<; (H.), which could perhaps belong with � K�'ro<; instead.
CtKax(�w => axoflal.
CtKaXIl£vo<; [perf.ptc.] 'sharpened' (Il.). <!l IE? *h2ek- 'sharp'�
.ETYM From the root *h2ek- 'sharp'; a reduplicated formation *uK-aK-cr-flEVO<; has
been suggested, which remains speculative.
CtK£UVO<; [m.] a kind of leguminous vegetable (Pherecr.). <!I ?�
ETYM Unexplained. For the overall structure, cf. 'OKWVO<;.

CtK£U£l => UKOUW.
CtK£WV YAR Also -Eouoa, -EOVT£. => UK� 2.

CtK� 1 [f.] . UKfl� GLO�pOU 'point of an iron tool' (Suid., H.), cod. aiXfl� . <!l IE *h2ek­
'point, sharp'�
.YAR A parallel formation is UKl<;, -lOO<; [f.] 'needle, arrow, barb' (Hp.).
.COMP On -�K'l<; see � �K�.
DER From UKl<;: UKlOlOV 'small barb' (BCH 29, 572), CtKlOWO'l<; 'pointed' (Thphr.),
UKIOWTO<; 'id.' (Paul. Aeg.), also plant names like UKlOWTOV (Dsc.), passive verbal adj.
�KlowflEVO<; (lG 2, 807), also compounded in CtKlOO£lO�<; (Procl.). UKlOKAWV []
(BGU 1028, 12; 16 [lIP]), mg. uncertain, was borrowed from Lat. aeiseulum 'small
pOinted hammer of a stonemason', cf. Schubart's comment ad loCo Reduplicated
form in � UKWK� 'point (of a lance, sword, etc.) (Hom., Theoc., Opp.), cf. uywy�.


.ETYM Probably from a root noun; see Schwyzer: 465. Derived from a root UK­
'sharp', seen in several other etyma. Not related to � aKmva, � aKavo<;. See � aKwv,
� UKfl� � UKOV'l, � aKpo<;, � aKWV.
CtK� 2 [f.] 'silence, quiet' (Il.). <!I ?�
.YAR Beside the instr. UKa., UKq. (Pi.) only acc. UK�V; in Hom. adverbial in UK�V
EYEVOVTO GLW1tn, which shows that the original mg. was 'quiet, calm'; cf. UK�V �y£<;.
�auXlav �y£<; 'were bringing quiet or calm' (H.) .
.DER UKEWV, -EOV-r£, -Eouoa is a ptc.; the form in -wv became indeclinable (I::!. 422).
The optative CtKEOl<; (A. R. 1, 765) is a late creation.
CtK�VlOV' �auxov 'quiet' (EM 48, 1); aKaaKa = �ouxw<; 'quietly' (H., Crat. 126),
uKaoKq. (Pi. fr. 28), formation unexplained; uKaACt [, adv.] (Hes. fr. 218, Sappho
43 LP); uKaAav (Sappho 68, 86 LP); uKaA6v· �ouxov, 1tpq.ov, flaAaKov 'quiet, mild,
soft' (H.); this adverb also in uKaAapp£lTQO < uKaAa-p£F£- (Il.).
.ETYM It may be connected with � �Ka, assuming ablaut.
CtK�pa-ro<; [adj.] 'undamaged, intact' (Il.). <!I ?�
.YAR uKEpmo<; 'id.' (Hdt.).
.DER uK'l paGLo<; 'pure' (Od.), 'untouched' (h. Mere., AP). Similar formations are
UK�PlO<; 'undamaged (by the K�p£<;)' (Od., epic), uKEpmo<; 'unharmed, undamaged'
(lA). From CtKEpmo<;: uK£pmo-r'l<; (Plb.), UK£pmoaUv'l (Suid.), uK£pmooflm (Eust.).
.ETYM An epic and poetic word. It is unnecessary to assume a second, independent
word meaning 'pure' (Od.), as Frisk does. DELG pleads for a unified meaning
'intact, pure'. Perhaps, uK�pa-ro<; (not from K'lpalvw, A. Supp. 999) was metrically
lengthened from *uKEpaTo<;, from the stem of K£pa-l�w 'to destroy', but influence of
K�p is improbable. In some cases, the meaning may have been influenced by
� K£pavvufll 'to mix'. Lee Glotta 39 (1961): 191-205 connects it with � Kdpw, but this
leaves the formation unexplained.
CtKlSVO<; [adj.] 'weak, small' (Od.). <!I PG?�
.YAR UKlOpO<; (Cyr.).
.DER UKlOpW1ta�w· Ufl�AUW1tW 'to be dim-sighted' (H.) .
.ETYM Unexplained. The element -ov- may point to a substrate word. In view of the
variant with -p-, Fur.: 388 assumes a substrate word with v/ p, though the
interchange is rare (cf. 1tpOKVl<;). At 360, he compares oKloapov' upmov 'thin,
slender' (H.), which cannot be considered certain. Frisk also compares UKlpO<; 'weak'
(Theoc.); cf. UKlP�' uo8£v�, OUK E1tlTeTaflEVa 'weak, not stretched' (H.) and UKlPW<;'
£uAa�w<;, u-rpEfla<; 'discreet, quiet' (H.); but UKlpO<;' �oppa.<; 'the north wind' (H.)
cannot belong here. For the interchange 0/ p, Fur.: 388 gives only ol�oa, where it is
probably conditioned by the preceding � .
CtKivUKq<; [m.] 'short sword of Persians and Scythians' (Hdt.). <!I LW Pers.?, PG?�
.YAR KlvaK'l<; ( 1061); the l was long in Hor. Od. 1, 27, 5 .
ETYM Perhaps an Iranian loan: Benveniste 1940b: 202 compares kyn 'k; see further
Bailey TPS 1955: 69. However, KlvaK'l<; in Sophocles (Belardi 1969: 202) could suggest
that the word is Pre-Greek rather than Iranian. It is supposed that CtKlvaYfla =

Documentos relacionados

Documento PDF overmans
Documento PDF dr holly lucille john o dowd grow younger blood ebook
Documento PDF tourism marketing and the increasing significance of authenticity
Documento PDF paving and driveways united kingdom
Documento PDF russian military capability in a ten year perspective 2013
Documento PDF gettingyourresearchprojectstarted

Palabras claves relacionadas