CN479 .hIS"" ""'™™">' "*""^
,. 3 1924 029 769 316
olin
'/M
Cornell University
Library
The original of this book is in
the Cornell University Library.
There are no known copyright restrictions in
the United States on the use of the text.
http://www.archive.org/details/cu31924029769316
EARLY
ETRUSCAN INSCRIPTIONS
FABRETTI 2343-2346
GEORGE HEMPL
THIS PAPER IS REPRINTED FROM
THE MATZKE MEMORIAL VOLUME
PUBLISHED BY THE UNIVERSITY
Ube XHnlversitg Press
STANFORD UNIVERSITY
CALIFORNIA
191 1
/.
f{\A
IN MEMORY OF JOHN ERNST MATZKE
EARLY ETRUSCAN INSCRIPTIONS
FABRETTI 2342-2346
How I, a Germanic scholar, came to be interested in Venetic and Etrus-
can, I have told in my report on the results of my Italic studies. This
report has been delayed, chiefly by the difficulties inherent in such an under-
taking, but it will now be published in a very short time. The present
paper is an abstract from it. A few weeks before his death. Professor
Matzke urged me to hasten the publication of my report. He said that my
silence was being misinterpreted, and that I owed it not only to myself but
also to my friends to publish something at once — if only a fragment. I
was touched by what he said and the way in which he said it. It was
almost exactly what another friend. Otto Jespersen, had written me from
Copenhagen not long before, and what still others, as if by concert, now
began to urge upon me. I saw the force of their arguments and decided to
drop everything else and complete my report. And now that Fate has sud-
denly cut short the life of one of them, I can find no more appropriate
tribute to lay on his grave than the fragment he so recently urged me to
publish. If it seems abrupt or takes things for granted that I have given in
the report but could not well incorporate here, scholars will, I trust, con-
sider the circumstances of publication. There are, however, a few matters
of a general character that must be briefly touched upon before I proceed
with the interpretation of the inscriptions.
Etruscan is a sister of Latin. In their earlier stages the two languages
can hardly be distinguished. But Etruscan matured far more rapidly and
had already passed into old age when Latin was only attaining her majority.
To put it otherwise, Etruscan is an Italic dialect that became a modem
Romanic language in ancient times. Many of the forms, phonologfical
changes, and other linguistic phenomena that we are wont to regard as char-
acteristic of modern languages, are to be found in Etruscan long before the
days of Julius Caesar. The development of the language, from the early
stage when it was almost identical with Latin, down to the time when it was
so different that the Romans regarded it as a totally alien tongue, can be
traced step by step on the chiseled monuments that stand in our museums.
In attempting to read Etruscan, scholars have, however, largely confined
themselves to a study of late inscriptions, and have permitted the relatively
O EARLY ETRUSCAN INSCRIPTIONS
modem forms that they there found to blind them to the original character
of the language — much as Old-English scholars once did with West Saxon.
The situation in which the philological world at present finds itself with ref-
erence to Etruscan is as regrettable as it is extraordinary. The kinship of
Etruscan and Latin lies open for all men to observe, and yet this fact is
denied by practically all Indo-European philologists. Moreover, the study
of this important Italic language has been permitted to fall into the hands of
a school, which has malie of this denial a cardinal dogma and has prescribed
for its"tt& disciples a rule of conduct, or only- legitimate method of study.
Here, carefully screened from the light of Indo-European philology, "Etrus-
cologists ' ' spin airy webs, while without, in the busy world of learning,
scholars concern themselves with other things. It would be impossible for
me here to discuss the situation in detail. I must, in this paper, restrict
myself to the presentation of three very early inscriptions, and I shall be
satisfied if I succeed in making it clear that early Etruscan can scarcely be
distinguished from Old Latin.
The inscriptions will be found in Fabretti's Corpus, Tab. XLii, 2343-
2346. All I know about their origin is what Fabretti quotes* on page cciv,
namely, that they are from various friezes or paintings found in ancient
Etruscan tombs near Corneto-Tarquinii. So far as I can learn, their gen-
uineness has never been questioned, nor their Etruscan character. A study
of the forms of the letters shows that we have three, not four, inscriptions ;
for 2345 and 2346 are parts of one and the same. The first two give the
names etc. of the dead ; the last (2346-2345) pertains to the formalities of
cremation. I can not detect any material difference in age, either epi-
graphic or Unguistic. Perhaps the forms of the letter for a prove 2344 to be
the oldest, and 2346-2345 to be the youngest. According to Etruscan cus-
tom, the writing runs from right to left, but I have reproduced it in our
way. I have also substituted for the Etruscan characters the conventional
transcription, in Italics, and have added the phonetic transcription in Roman.
In the grammatical discussion, a prefixed asterisk (*) marks a recon-
structed word or form ; a prefixed period (.) indicates phonetic spelling.
An apostrophe (') before an .1 .m .n or .r signifies that the consonant is
syllabic. The letters .y and .w represent the voiced fricative consonants
heard in English ye and woo. The macron (~) is added to a long vowel,
rather than placed over it. The pitch accents of Greek and Sanskrit
are indicated by the usual signs. The sign for stress, in accordance with
the best modern usage, precedes the stressed syllable ; but, to avoid misun-
derstanding, I use the grave ('), in order that it may point toward the syl-
*"essais des diff (rentes frises ou peintures qui se trouvent dans les souterrains
des anciens Etrusques pres de Corneto (Piranesi De Rom. magnHicentia ct archi-
tectura, Romae 1761)."
INTRODUCTORY REMARKS 7
lable to be stressed, not away from it. I have indicated the stress only when
it has shifted from the initial syllable. The phonetic characters employed
are the usual ones, the vowels having their German values, and the consonants
their English. The turned e (.a) stands for an obscure vowel, for example,
that beginning the word ago and that ending the name Anna. Small capital
.K and .G represent the velar stops in cook (.kuk) &wA good {.gw^, as dis-
tinguished from the more palatal sounds in kick (.kik) and give (.giv). A
curl over a velar consonant indicates labiaUzation. Thus we hear quick
sometimes as .icik, sometimes as .Kwik. Compare also alive (.a^laiv),
devout (.diVaut), eatable (.i^tab'l), Wednesday (.wenzdi), purely
(.pyu~rli), etc. s' is a conventional way of representing a distinct j-sound , / /
usually due to contraction ; it has nothing to do with accent. /\ Tjie, (2 k (or c).
Thus, while k stands for .K or .G , x *^°o ™ay stand for .K or .g . Simi-
larly in English, while k stands for .K (for example, in kook), so too, c
may stand for .K (as in cool), while it may also stand for .s (for example, in
«^etc.). Moreover, in Etruscan a u may stand for .u or .o , or for an
obscure vowel (like that in Latin optumus or optimus), while v too may
stand for .u or .o , or for a similar consonant (.w .f etc.).
I should state that in Etruscan, as in primitive Italic generally, the
stress originally fell on the first syllable. From this it shifted in Latin to a
long penult or to the syllable preceding a short penult. In Etruscan a dif-
ferent principle prevailed. The stress regularly stayed on the first syllable if
that contained a long vowel or a diphthong. If it contained a short vowel,
the stress shifted to the nearest long vowel or diphthong ; see 'Kertnupi,
^ericaru, stani, etc. , below. This shift is often betrayed by the reduction of
the vowel of the initial syllable (for example, na(f)ods' = . n3^po~ts' <
.nepo~ts, Latin nepo~s, ' grandson, ' Lemnos Stone ; also .a(n)na^ke~, vari-
ously spelled annice (F. iii, 391), eneke (F. 2614), uneke (F. i, 234),
unuke (Gamurrini, 607 and 608) 'granted, gave,' Latin annuit 'assented,
granted,' see also raned below) ; or by its total disappearance (s&e fravvmv t^fx to
below; 6ne -am- > -ma- are common, cf. Greek Her-ac-les > Latin
Her-cu-les. From the names VecaQ and 'Kermupi, it is clear that we have to
do with members of a Greek family. There evidently were not a few
Greeks among the Etruscans, even in the earliest times (observe the V-, not
H-, of the name Vecad), who acquired wealth and standing. A notable
case is that recorded on the grande sepolcro (Gamurrino, 799) of Laris
Pulena, who is stated to be ' the great-grand-son of Laris Pule the Greek' —
prumts Pules Larisal Creices. Compare also the story of Demaratus, the
reputed Greek progenitor of the Tarquin family.
*In Etruscan we fortunately are not, as in Latin, dependent upon the evidence
furnished by metrical texts and upon the conflicting interpretations that modern
scholars have put upon ancient metrical usage. The loss of -o~ after a short
syllable and its preservation after a long syllable are facts that can not be argued
away. At another time I shall show that the Old-English loss of a final short
-u after a long syllable and its retention after a short syllable, instead of offering
(as argued by Sweet and Sonnenschein, Classical Philology, January, 1911, page 3)
evidence against the doctrine of iambic shortening, offers evidence for it.
FABRETTI 2343 9
^ericaru (.perrKa~ro) 'very dear,* Latin percaTrus. In all but the
very oldest Etruscan (for example, apastvs ed palamneus tupanktvs, F.
2341), final -s disappeared, as in Old Latin, after a short vowel, provided
no vowel followed. At an early date, the j-less forms prevailed in Etruscan,
the j-forms in Latin. Observe puru, s6vnimv, Titv, xsimeu, icamus,
fravvmv h&loyn ; aho eluri nnder 9 ri<^tent, page 15. Later the -o became
-3, written -e, see Titv, page 11. peri- is, of course, older than the per- of
Latin per-ca~rus.
:n is an abbreviation for natu (.na~to) 'son,' Latin (g)na~tus. (For
n- < on-, see stani below.) Compare the use oi f in Latin inscriptions as
an abbreviation ior fi'lius.
stani (.stan^ni") 'of alloy,' genitive of the word seen in Latin stannum
'alloy.' Italian stagno and the other Romanic forms point to a Latin *stag-
num 'alloy.' Compare Latin stagnoTre 'plate with alloy.' As medial -Gn-
becomes -nn- in Etruscan but not in Latin, it is evident that the Romans
adopted the Etruscan form with its -nn-. This suggests Etruscan influence
on Rome in the matter of the working of the metals, which is just what we
have reason to expect. But differentiation from stagnum. ' water that has
flooded land,' 'standing water,' 'swamp,' doubtless contributed towards
fixing the Etruscan form in Roman speech. Historically, stagnum ' a piece
of shallow standing water, a fen ' is one and the same word as stagnum/
j/fl:«««M 'a metal wash, the alloy used in plating.' English wask too has
both meanings. Initial on- became (n)n- in both Etruscan and Latin, see
n[atu'} above.
puru (.pu~ro), Latin /«~;^j 'pure,' 'free of,' compare Horace's sceleris
purus. For the loss of -s see ^ericaru above.
hem, is from .hemo < .hemo~. Old Latin ^hemo", hemo~nem, Latin
homo ' man.' For the loss of -o~, see VecaO, page 8.
Note on Janus
The ^«« 'Janus' that is often cited (Korte, Etrusker, in the Pauly-
Wissowa Real-encyclopadie, 767), as a sort of mate to Uni 'Juno,' is a
ghost-word. The idea is based upon the erroneous supposition that the
Bronze Liver has Ani One, in section 6. Close scrutiny reveals the fact that
the text reads Tins One, corresponding to the Tins One-0 %uflOas in the
inner sections 18-20. It is remarkable, however, how near Janus was all the
time. For OuflOa, or OuplOa (F. 1054, CIE. 445) is .dupTta" < .dupiPta"
< .du^pla"ti < .du'pla~tio < .dupla"tio" ' doubleness. ' (For the shorten-
ing and loss of -o~ after -i , see VecaO, page 8. ) That is, we have before us
no other than Janus Bifrons, who so often appears in the form of a lamp-
statuette, called alpan (.aPpa^n < .lam'pa", Greek Xafiirr)). This Etruscan
lO EARLY ETRUSCAN INSCRIPTIONS
word for 'lamp' has been deified' and been used to add formidableness to
the list of Etruscan divinities whose names are supposed to be incapable of
reconciliation with Indo-European. The Tins 6ne (e)6 %ufl6as on the
Bronze Liver is Jovis templum et Jarni~ ' The section ( ' temple ' in the
technical sense) of Jupiter and Janus.' One is .tne~ < .ta^me" < Greek
TO/itj; 'section' (compare the tern- in Latin templum). The Etruscan -e~ <
Greek -»/ betrays non-Doric Greek influence on Etruscan divination. While
on some of the lamp-statuettes the old Italic god is called %upl9a {®upldas'
alpan 'lamp-statuette of Tkupltha,' F. 1054, CIE.445), on others he is given
a name corresponding to \^2.^n Janus Clusius. Thus on F. 1051, CIE. 437,
we find Cuts' ans'i alpan, and on F. 1052, CIE. 438, S' elans' I alpan. Cut-
s' an-s'i (.K3rs'a~ns'i) and S'elan-s'l (.sVla~ns"l) are genitives of one and the
same name, that is 'Janus CoUus,' or ' Turn-neck Janus, ' compare our slang
term rubber neck 'person who has the habit of looking back to watch
others, while he tries to keep on walking ahead.' This collus is from
.Kolsos, .Kelsos, -j/icel 'turn, twist,' cf. Latin collum, collus, etc. The
Etruscan Cuts'- is for Kols- ; Set- is from .sKel < Kels-. The initial i- or y-
before the long a~ of .ia~no- regularly assimilated to the preceding conso-
nant (see xsimeu below), whence the syllable -a~n- . But we do not find
Colsos or Collus in Latin ; for through popular etymology and association
with clu~do, clu~sus, the name assumed the form Clwsius. The use of an
effigy of Janus to hold a light is, of course, due to his being the old Italic
god of light.
FABRETTI 2344
Maru : sdvnimv-m Titv: "xsimeu letive : lapzi smalvi : ®anri6er
(.marVo" stu~nimo-m tito Ksimeo le~ti~ve~ i^a~ptsi~ smalVi" tanritei")
'Maru the sturdy and Titv the distinguished, dead {or who died) of the
lapygian disease at Tarentum.'
It will be observed that the only punctuation consists of double puncts
to mark proper names. Such punctuation stands after the two personal
names and before the two geographical names.
Maru (.marVo") is to Latin Maro as Marro~nius is to Maro'nius
(Schulze, Zur Geschichte lateinischer Eigennamen, page 189). Maro
would appear in Etruscan as *Mar (see Vecad above), while Marius
would be *Mariu, later Marie (F. 654, CIE. 2451), still later *Mari.
sdvnimv (.stu-nimo) 'very sturdy' or 'very steadfast,' a superlative in
-imo-s, cf. Sanskrit sthu~rds or sthu-lds 'thick, large, strong,' and Sanskrit
stM~na~, Greek cttvKo^ 'post, column,' originally 'phallus.' The change of
medial -1- to -n- is common to a large part of the Etruscan territory. Still,
the Etruscan adjective may have the -n- of Sanskrit stM'na- rather than
the -1- of Greek o-tOXo?. For the loss of -s, see tjtericaru, page 9.
FABRETTI 2344 II
-m 'and,' here = .m, as on the Lemnos Stone and often elsewhere.
In F. 2345 below, it is syllabic ( = .'m) ; but we also find -em (^les'c-em,
Magliano) and -um {axr-um F. 2598) = .am. This is the same word
as Old-Latin em, im 'then,' seen also in Latin (znterjim, iormed like inter-
dum. Compare Umbrian enom 'then' = Palignian inom 'and' = Latin
enim 'for.' Also Umbrian pun-um, 'quando-que' (Buck's Oscan and
Umbrian Grammar, §201, 5). For the relation of idea between 'and' and
'then,' compare the like use of "then — — , then , then ," and
"and , and , and ," in narration. The Etruscan word is
almost always enclitic ; in F. 2345 (page 16), it introduces a new clause.
When so used, it seems to have retained the old sense 'then.'
Titv (.tito) < .titos, Latin Titus. This .tito later became .tita, written
Tite, see j>ericaru, page 9.
Xsimeu (.Ksimeo), Latin eximius 'distinguished.' An initial short
vowel was often lost in forms having the stress on a succeeding long
vowel, compare (»V«~, later ca~, page 7. Before a long vowel, i was
consonantal (that is, .y) and assimilated to a preceding consonant, doub-
ling it if immediately preceded by a short vowel. Thus .3ksimya""s
> .Ksim^ma~s. (This gemination of the preceding consonant after a
short vowel is proved by Latin spellings like Spurinna, Porsenna, etc.;
see also raneO, page 17.) Observe the masculine family names TarxnaTs
and Tare nor < * Tarcni'a's and * Tarcnia" ( originally respectively
genitive and ablative of the -ia feminine name of the family, which was
added to the first name just as the tribal name in -ia was in Latin),
also the derived feminine Tarx^nai ; and compare these with the original
feminine nominative ^TarXfnid. As e- is lost in ^simeu too, it is clear
that the comparatively few forms of the adjective that did not end in a
long vowel adopted the shortened form of those that did ; but it is not cer-
tain whether they also shifted their stress to the final syllable or simply had
it on the first syllable of the shortened form. The medial i or e certainly did
not disappear before the short -o. Moreover, it is clear that the initial stress
of the nominative, accusative, and vocative (the most commonly used cases
of a name) held its own in masculine names like Tavarsio (.taursio, Greek
Tai5pe(cr)to9, Latin Taurius) on the Lemnos Stone, and the later Askamie
(.asKamia < .asKanios, Greek 'Aff/cawo?, 'L.^'iva. Ascanius)., F. 2614 quat.,
and the still younger Fulni (.folni < .folnia < .folnio(s) < .folwios. Lat-
inized as Folnius, but in real Latin, Fulvius) F. 251 and 11. p. 28 Tab.,
CIE. 428. xsifneu and sdvnimv may be common adjectives or cognomens,
I am not sure which.
letive (.le~ti"we~), the nominative plural masculine of .le~ti~wo, Latin
*le~ti~vus 'killed,' 'dead,' is to le~tum 'death,' as furti~vus 'stolen'
is to furtum 'theft,' and as capti'vus 'captured' is to captus 'capture,' etc.
12 EARLY ETRUSCAN INSCRIPTIONS
The ending -e~ (compare Old-Latin ploirume~ for the later filu~rimi~) is
the intermediate stage between older (-oi » -ei and later -i". Compare
(F. 314, Tab. XXV, CIE. e^2)fvimv — pace (.fu"imo paK^Ke", for kk < Kt,
cf. Qw^tiQ, page 16), Latin /«m«j pacti~ ; with fvimv, compare fravvmv,
page 16.
lapei {.'i^arpXsr), genitive of ^lapzu (.i^a"ptso) 'lapygian,' 'pertaining
to the 'Ia7ru7e5' (the natives of 'lairvyia, whence Latin lapygia, the terri-
tory about Tarentum, and thus northwest of Greece), of which 'Iu'ttv^ 'the
northwest wind' is really the singular. For the origin of the name, see
page 13.
smalvi (.smarwi"), the genitive singular of *-smalvu .gherohe^n , genitive Sh'r'ghenos.
According as the stress permitted, the first syllable was .oher- or .Sh'r- ,
and this variety was regularly reflected in Greek by forms with rep- and
forms with irap- . Thus, while the nominative was *Tep6rjv, there were
oblique forms like the genitive irapOevos. A competition set in, whereby
one of the rival forms (.ter- or .par-) won the day, or a compromise form
(.tar-) resulted. In Greek we find the old genitive irapdeva employed, but
regarded as the nominative of an o-stem. This passed into North-Eastern
Etruscan ; but with it the old nominative *Tep6ijv also came in, becoming
*The identity of tlie initial letter is partly concealed by an accidental mark,
slighter and different from the strokes of the letters.
FABRETTI 2344 1 3
by metathesis tre'te'it. Thus at the beginning of the very old Novilara
inscription we find : —
Partenus polem is'airon tet
'A maid presented this pedestal,'
and at the close, in more elaborate form : —
Treten telet aunem polem tis'u s'otris' eus
'A maid presented this (same) pedestal in accordance with the
testament of her benefactor.'
tet (.de~d) is from tetet (deMe~d), while telet (/deled) is from /deded, see
under raneO, page 17. polem < .podem = 7ro'8a. is'airon < .isto- +
.aiso-m, cf. Latin iste and Oscan eisu-d . aunein = Latin eundem, with reg-
ular change of medial -nd- to -nn- . For other details see my report.
The compromise form .tar- is found in Tdpas, -avro^, and in Latin
Tarentum. and Etruscan ®anri6er. All the forms show metathesis of
one kind or another. The Greek and the Latin forms betray the influ-
ence of nt- and nto-st&ms, the name thus falling in with such names as
Uzentum, Hydruntum, Sipontum, etc. (Kretschmer, Einleitung, page 260).
The Etruscan form retains the dental stop in its original position (.tarthen-,
®anri6er), but it suffers a peculiar metathesis of the surrounding sonorants :
thus the r and the n exchange places, but under the influence of the original
form, the r is restored next to the n that had taken its place. The form
%anrider is a locative without suffix (Brugmann, 11. §256 and 257), the
word being a consonant stem in Etruscan, as originally in Greek.
This makes it clear that the name of the city is no other than tre''te~n,
iOr 7rap6evo-a~l3K- , page 18.
But in .tripoMe~nt, the -d- began the stressed syllable and was, therefore,
virtually initial. Etruscan eluri corresponds to the Old-Latin variant edoris.
Latin edo'ris would be .eMo~ri in Etruscan and would have been written
*eturi; cf. /^/«/ under raned, page 17.
-X« (-Ka), Latin -que 'and.' The Capuan Tablet still has xue (.2.3),
but Etruscan .k early became .K,* hence we usually find -ce (F. 2598), -xe
(F. 2327 ter b and very frequently elsewhere), or -x« (as here), all = .K3,
or we find -c (see under Ori^tent above), with the obscure vowel lost. But
unstressed -ga at times became -xla, thus we find -de, -xleand -xla '-que,'
which persisted after . k had become . k in other words : Lautnes'-cle caresri
Aules' 'of Lautne and dearest Aule' (F. 1915, CIE. 4116). With the
*Where .kw is found in later texts, it is foreign (Cvinie < Latin Quintus) or of
secondary development {-xva < -trains) in the Mummy Ms., 8-3 etc.: celi huBis!
zadrumis' fler-xva NeBunsl 'walk thirty-four times past the statue of Neptune').
celi is .Ke~Ie- < .Ke-de", ' walk, stride.' From Latin cerde we should expect eel,
with loss of the short -e, certainly in so late a text, celi ( < .Ke~de~) makes more
probable Brugmann's derivation from ce -f- ^sed (see Walde). It would seem actu-
ally to be from ce-sedeo ' sit here,' 'come on over here and sit with me,' 'come along,
don't stand there,' 'come on,' ' march,' etc. As in Umbrian etc., the .e~ in Etruscan
was very close and was often written i, especially when unstressed.
l6 EARLY ETRUSCAN INSCRIPTIONS
more usual Etruscan -xa, -xe, etc., compare Venetic -ka (No. 22), -ke (No.
9), -xe (No. 291). The genitive ending -es (in northern spelling, -es),
seen in Pules and Creices, page 8, and in Lautnes and Auks', is .e~s < .eis,
Oscan -eis, Umbrian -es j -er (Buck, §171), which arose in the 2'-stems.
/«■ (.pi~), Latin /«" or pi~, 'the devout ones,' here probably the hired
mourners.
m- G'm) 'then,' see -m, page 11.
BuxtiS (duKtit, Latin ductita~, imperative of ductito 'lead, lead along,
lead forth.' In Etruscan this verb is of the third conjugation, with final
short -e regularly lost. The -Kt- is a sign of great age, for .Kt early became
.k(k) in Etruscan, cf. pace under letive, page 11.
nvhunt (.no~o"nt) 'messenger.' The word originally meant 'new-
comer,' being due to the conglomeration of *neu(i)os ^uentos, nov(i)us
ventus. The Etruscan word is the same as Latin nWntius ' messenger ' ;
but the peculiar phonology led to changes that were not identical in the two
languages. In considering these, it should not be forgotten that the Italic
stress rested on the initial syllable. In Latin we find the development :
*neuios uentos, which by metathesis of -ios and -os and regular change of
-sw- to -w- (Sommers, p. 231 3) became *neuouentios > (Sommers, p. 97)
*neuentios > (Sommers p. 74, 2) nouentios > (Sommers, §86) *nountios >
(Sommers, p. 175) nu~ntius. Etruscan, on the other hand, developed as
follows (in Latin spelling) *neuos uentos > *neuouentos > *nououentos
> *nouountos > nororntos, the vocative of which (with regular loss of final
-e) is our .no~o~nt, speld nvhunt.
aisaru (.aissro") < .aisaro~m, genitive plural, 'of the gods.' The
' messenger of the gods ' was Mercury, who also conducted the souls of the
dead to the lower world. The invocation is, therefore, to him. aisar
{aesar Etrusca lingua ' deus' Suetonius) is a variant of aisos (aicrol
' 6eol' vTTo Tvpprjv&v, Hesychius), see Buck, page 12 etc. Contrast the
retention of -u, which represents -o~ < -o~m, with the complete loss of
short -om in fust below.
ustvn (.usHo~n) < .us^to~r'm, Latin usio~rem, accusative of ustor
'cremator.'
fravvmv (.fra^wamo), Latin *fera~viinus 'tulimus.' This form, by
its loss of -e-, betrays the shift of stress spoken of above ; compare also One,
p. 10. 'With, fravvmv, compare fvimv under letive, page 11.
pici (.piki) < .piKim, Latin picem, accusative of pix, Lithuanian /?>^m'
'pitch.' For the loss of -m, see ia above.
fust (.bust) < .bustom, Latin bustum 'place of cremation.' I reg-
ularly represent the Etruscan letter 8 by the conventional transcription f
but it is obvious that it here still has its original function of representing the
labial stop .b . Pauli was mistaken in deriving the Etruscan 8 from the old
FABRETTI 2345 & 2346 IJ
three-barred h. It is nothing but a cursive form of B , and got the value .f
as old .bh became the labial fricative. In the inscription Gamurrini 804, the
first two /'s are written 8 , the following three have the form B . For the
loss of -om, see ia, page 15.
raneQ (.ran ne~d < .ran^e^d < .fanie~t) is the third singular future
of Latin 'Cranio, Greek paivm ' (be)sprinkle. ' The change of -t to -d is the
same as in Oscan, Umbrian, and Latin ; compare Old-Latin sied, fhefhaked,
etc. But this -d is found only in very old Etruscan, for it early became
silent {tenue, F. 2033 bis, E a) and in time a preceding short -e disappeared
(tenu, F. 2070, III. 329). The ending of Latin tenuit is a reconstruc-
tion on the basis of the present, which has -t < -ti, cf. Etruscan s'udit
(.s'u(d)dit), Latin subdit, F. 2335. Corresponding to Latin -it « -ed) and
-i~t « -eid), Sommers, page 618, Etruscan had -ed and -e~d: the former
in tenue(d) (F. 2033, bis, E a), later tenu (F. 2070, iii. 329), and in many
other verbs, for example, (F. 2100) : eisne~v-c eprQne~v-c macstre~v-c,
Latin aestumarvit-que imperita~vit-que magistra~vit-que ; the latter in
.anna'ke" 'gave' (see page 7), and in tetet (.deMe~d, cf. dri^tent, page 15),
with which contrast telet (.deled), page 13. The form tetet is found in
F. 2753, an inscription that is usually classed as "mixed Oscan and Etrus-
can," but which is in reality pure Etruscan from beginning to end.* The
only consideration that has led scholars to suppose that this inscription is
not pure Etruscan is the fact that several words in it are obviously Indo-
European, which was not compatible with the current doctrine that Etruscan
is not Indo-European.
It will be observed that that part of our inscription that is numbered
2346 forms a sense-unit. What is numbered 2345 falls into two sense-units.
If we write the text in this way, we get three metrical lines : —
^i^ica^mos 'ya~ tripMe~nt-K3 'pi~.
"m-duKtit, ^no~o~nt 'aisar(o~) us^to~n ;
Tra~wamo 'piki, 'bust ran'ne~d.
'Let us strike up now, and let the devout ones dance the sacred dance.
Then lead forth, O Messenger of the Gods, the burner of the dead ;
We have brought the pitch, he will sprinkle the place of burning.'
Each verse is a trochaic dimeter catalectic, with the substitution of a
dactyl for the trochee in the first foot of the first dipody, less often of the
second dipody. The meter is, of course, dynamic, that is, based on stress,
not time.
*Fabretti made a strange blunder in reading the perfectly distinct per aciam
as aeraciam, and others carelessly followed him.
EARLY ETRUSCAN INSCRIPTIONS
Note on the Numerals
As there has been occasion to refer to the numerals, I may say that the
correct order is that of Campanari : —
max 6u sal hud ci s'a sem<]> cesp- muv- mealxls
1234.5678 9 10
Every form is a regular derivative from primitive Indo-European. The
puzzling zal (.tsa~l) is from older "^zar (.tsa~r), which arose by metathesis
(cf English three but third') from .tsra" < .tis(3)Va~, Avestan tis'aro~,
Sanskrit tisrds, Old-Irish teoir, etc. , ' three. ' The Etruscan numerals were
often feminine abstracts like Greek o'lvq ' the number one ' and German die
Eins. The older '^sar (.tsa~r) is reflected in zadrm (.tsa'tr'm) < *zar6m
(.tsa~rt'm) ' thirty' < (.tis(3)Va~ ' three,' + .dez'm ' ten '). As in this word
the .-ra" became .-a~r by metathesis, the .d of .deK'm came to stand next to
the .r, and thus .tsa"rdeK'm became .tsa'rds'm > .tsa"rt'm > .tsa~tr'm,
spelled zadrm, zadrum, etc. In other numerals, the final -a~ remained
in position, and thus .-a^dcK- became .-a~l3K; (see page 15) > .-(a)lK- ,
spelled -(a)lx-, etc. : celc, cialx-us', cealx-ls, etc., all forms of the word for
'fifty'; s'ial%-v(e)is 'sixty' (the analogy of the preceding ci- of the word
for ' fifty ' caused the substitution of j'z- for s'a- ' six, ' which is from '^zecsa
(.tseK^sa"), compare, with initial stress, zecsans'l (.tscKsans'l) 'of a six-year-
old,' Latin sexennis) ; etc. The tens usually appear in the adjective form,
with the Indo-European adjective suffix -we~s, -wents, seen in Avestan
vi~saiti-vant- 'twenty-fold 'and in Greek rerpa's, -avTO<;, etc. (Brugmann, 11.
§182, 2, 11'. §356). This -we~s appears as -veis and -vis {ei and i = close
e~) on the Lemnos Stone {s'ialx-veis, -vis) ; as -us (.os < -wes < .we~s,
cf. Latin soror < .sweso~r) in the Mummy Ms. {cealx-us etc.) ; and as -Is
(a dull syllabic 1 -|- s) in Italian Etruscan {ceal%-ls etc.). In Southern
Etruria the final -s of this adjective suffix passed by analogy to the numbers
below mealxls ' ten ' when these were used as adjectives : thus maxs, huds,
etc. , by the side of the max, hud, etc. , found on the dice. This -s has been
mistaken by some scholars for an inflectional ending, and has been used as
shot to fire at the " Indo-Germanites " (Skutsch, Pauly-Wissowa, 805-47).
CORRECTIONS
Page 6, line 7, read Italic langtiage.
8, " has made.
9, ' ' for its disciples.
12, add r' represents an Umbrian variety of r.
35, read see fravvmv, page 16.
26, ' ' %anri6er.
T";, " 5, " no~o~nt {not nu~urnt).
16, " 23, " no''o~ntos (not 7iu~n~ntos).
THE TUSCAN TUNG
BY
GEORGE HEMPL
PROFESSOR OF GERMANIC PHILOLOGY
STANFORD UNIVERSITY
THIS volume furnishes the first authentic report " j
on Professor Hempl's discoveries as to the
nature of the Etruscan language and its relation
to other Italic dialects. The book gives a resume of the
Etruscan question; many inscriptions (in conventional
transcription, in phonetic spelling, and in translation) ;
chapters on the phonetic, morphological, and accentual
laws of the language; facsimiles of important inscrip-
tions, and various illustrations ; an index and glossary.
In an appendix are also given sample Venetic and
Messapian inscriptions, which for the first time reveal
the true nature of these languages.
The hook is to be issued during the Christmas
holidays, when it zvill be put into the hands &f the
trade and sold at the price of $3.00. Until that time,
St. John's Press zvill receive subscriptions at $2.25,
and zvill forzvard the book as soon as issued to such
cs send remittance zvith subscription.
St. John's Press
Stanford University
California