Etruscan as an Anatolian Language

G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Etruscan as an Anatolian (non-Hittite) Language THIS IS AN EARLY DRAFT OF THE PAPER PUBLISHED IN Atti del Sodalizio Glottologico Milanese, vol. V n.s. 2010 Edizioni dell'Orso - ISBN 978-88-6274-286-3 Currently (October 2013) I'm working on an expanded (and hopefully improved) version of this short paper, which should be completed in 2014 and will include: • a thorough analysis of the sources from which I extracted input data for my research • systematic references to such sources • an expanded description of comparison methodology • a new section on sound correspondences • a revision of several etymologies, plus some new ones • more detailed comparisons with Anatolian languages • a more in-depth treatment of bound morphemes • additional data from Lemnian and Raetian • a more robust set of statistical analyses of my findings Stay tuned on Academia.edu… 1 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 TABLE OF CONTENTS Key findings .........................................................................................3 Sources ................................................................................................3 Methodology ........................................................................................4 Regular sound laws .............................................................................5 Etymologies of basic lexicon ..............................................................9 Introduction......................................................................................................... 9 Likely etymologies ............................................................................................ 10 Tentative etymologies....................................................................................... 14 Basic lexical items with no pIE etymology ........................................................ 17 Non-basic terms with a pIE etymology ............................................................. 17 Etymologies of grammar elements ..................................................18 Likely etymologies ............................................................................................ 18 Tentative etymologies....................................................................................... 18 Grammar morphemes with no pIE etymology .................................................. 19 Shared innovations ...........................................................................19 Conclusions .......................................................................................19 Next steps ..........................................................................................20 ABBREVIATIONS IE Indo-European pIE proto-Indo-European 2 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Key findings Etruscan is an Indo-European language, belonging to a non-Hittite branch of Anatolian. This claim is based on the following evidence: • only lexical items for which Etruscologists have proposed a gloss have been used in this research (no attempt has been made at proposing new glosses for currently obscure or highly controversial Etruscan terms); • 138 of the glosses put forth by Etruscologists belong to basic lexicon, and 97% of such basic lexicon has an IE etymology; • 19 grammar morphemes have also been assigned a value by Etruscologists: 95% of them have an IE etymology • all such etymologies are based on regular sound laws; • a few notable innovations are shared by Etruscan and non-Hittite languages of the Anatolian branch of pIE. Sources The sources used in the present research are very few and readily available for inspection and cross-checks: B Giuliano Bonfante, Larissa Bonfante - The Etruscan Language: An Introduction, Revised Editon - Manchester University Press 2002 - ISBN-13: 978-0719055409 F G. Facchetti - L'enigma svelato della lingua etrusca - Newton Compton, 2000 FM G. Facchetti - Appunti di morfologia etrusca - Olschki 2002 - ISBN 8822251385 Joseph Greenberg – Indo-European and its Closest Relatives – Volume 1 – Grammar – Stanford 2000 – ISBN-13: 978-0804738125 Alwin Kloekhorst - Etymological Dictionary of the Hittite Inherited Lexicon Mallory, Adams - The Oxford Introduction to Proto-Indo-European and the Proto-IndoEuropean World - Oxford, 2006 - ISBN-13: 978-0199296682 Ranko Matasovi - Etymological Dictionary of Proto-Celtic - Brill 2009 - ISBN-13: 9789004173361 H. Craig Melchert - A Dictionary of the Lycian Language - Beech Stave Press 2004 Kàroly Rédei - Uralisches Etymologisches Woerterbuch -Budapest, 1986-88 - ISBN-13: 978-3447027359 Michiel de Vaan - Etymological Dictionary of Latin and the other Italic Languages - Brill 2008 - ISBN-13: 978-9004167971 G K M-A M ML R V 3 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Methodology How do you prove that language L belongs to family F? You must prove that the majority of native (=non-borrowed) basic lexicon + grammar of L can be derived from lexicon + grammar of proto-F through regular sound changes. An attempt to prove that L belongs to F can go wrong in many ways; here’s how the present research was designed to avoid each major methodological mistake: - a proof can be invalid if you skip some step in the method: the evidence presented in this research is based on a list of chronologically arranged sound correspondences, a set of lexical etymologies for over 95% of Etruscan basic lexicon, and IE etymologies for key Etruscan grammar morphemes; - a proof can be invalid if the method is applied to poor data: this research is based on readily available sources for Etruscan and IE (see list, above); - a proof can be invalid if the method is applied to insufficient data: the present research is based on a standard list of 138 basic terms, a number which is usually considered adequate for such purposes (often, an even smaller list of 100 terms is used); - a proof can be invalid if you posit incorrect sound laws (e.g. phonetically implausible or inconsistent ones): this aspect is discussed in the section dedicated to sound laws; - a proof can be invalid if you do not apply sound laws correctly (i.e. sound laws would be good per se, but are used badly): every lexical etymology is accompanied by an explicit list of all sound changes that were applied to derive an Etruscan form from a pIE form; - a proof can be invalid if you allow for excessive semantic latitude in your comparisons: etymologies proposed in this research are largely based on identical meanings in Etruscan and pIE, or on very close meanings with similar reflexes in other IE languages; - a proof can be invalid if you mistake a loan for an inherited term: this is by far the hardest part, and all the “rules” 1 that have been proposed to distinguish loans from inherited words are probabilistic, not deterministic; but since basic terms are more resistant to borrowing than “cultural” and specialized ones, verbs are more resistant to borrowing than nouns, and grammar morphemes (especially bound ones) are more resistant than anything else, the present research is focused on basic lexicon, verbs and (bound) morphemes, to maximize the probability that these elements are inherited, not borrowed. - 1 to identify loans and distinguish them from reflexes, you can try and use the following evidence: - - - terms used for specialized (technical, cultural, etc.) concepts are more likely to be loans than basic terms are verbs are more resistant to borrowing than nouns morphemes (especially bound morphemes) are more resistant to borrowing than lexical items as a consequence of the above, you probably have many loans from language L2 to language L1 if you find many lexical similarities in non-basic concepts between L1 and L2, but little or no morphological similarity between L1 and L2 (e.g. Japanese and Chinese) loans resemble the source language much more than other languages in the same family (e.g. English words of French origin resemble French much more than they resemble Italian, Spanish, etc.) a word w1 in language L1 belonging to family F1 may be a loan from similar word w2 in neighboring language L2 belonging to family F2 if w1 has no hopeful etymology in F1, whereas w2 is analyzable (has an etymology) in F2 a term may be a loan if it violates phonetic or morphological rules (presence of non-standard phonemes or affixes; presence of standard phonemes with non-standard phonotactics; etc.) 4 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Regular sound laws The present research identifies the following 13 sets of regular sound laws connecting pIE to Etruscan. The high number of lexical etymologies also made it possible to arrange such sound laws in chronological order. Hereinafter, sound laws are followed by all Etruscan terms to which they apply, so readers can see at a glance which sound laws are supported by ample evidence and which ones are based on scant evidence at the moment. The effect of many sound laws is not always evident, since subsequent sound laws may have further changed their outputs. Therefore, in order to fully understand how a given sound law applies to a given Etruscan word, readers should refer to lexical etymologies, where all applicable sound laws are explicitly concatenated, in order to regularly derive Etruscan from pIE. Anyway, to make the following sound laws easier to understand at first glance, some examples are provided straight after major sound laws; readers are anyway urged to look up lexical etymologies to get a better understanding of how sound laws worked. 1 Loss of laryngeals hxC- > C-: lupu-, mi, mini, lautni, nur , rasna, cezp, nefts Chx > C: amin , far (a)n(a)-, mur-ce, šu -, uluter, zen-/zin-, mal-, san-, esan, ezh1e- > a-, e- (conditions for a/e alternation must still be identified): an, -em-, ar-, et/e , eca h1eh1 > e: netšh1i- > i-: ica, in, ipa, ita h2 > e: cese, eh1 > e, a (conditions for a/e alternation must still be identified): car-/cer-, far (a)n(a)-, hen, heva, mlak, repin i, vers-, zen-, cel-, pes, sem , rasna, esan, me lum, me , etera, far-/har-, far (a)n(a)-, nefts, rin-/ren-, a/ša, tam-, te-ce, val-, vatie-, mal-, sal-, san-, sar/-zar, es, ezeha > a, e (conditions for a/e alternation must still be identified): cau a, men-, hec-, sa-; (exception: uluter, influenced by 'dark' /l/?) eh2 > a: faneih2 > ei: leinh2e- > a-: am-, avil, ac-, atr-, acnan-, al(i)h3- > a-: amin h3e- > o-: ur an eh3 > o: tur, cezp ih3 > i: ziviha > i: -i ha > a: apa haei- > ai-: esuh2 > u: ur 5 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 2 Delabialization of labial plosives dw- > d-: zal (two < *zi-al + < *di- < pIE *dwi- two) d w- > d -: ez-, es k > k: car-/cer- (to make, do, build < pIE *k er- to make, do, build), cel-, -c, - , hu /hut g - > g-: ziv- (to live < *giv- < pIE *g ih3-we/o- to live) Exceptions: g - > w-: vatie-k - > -p-: ipa, popa 3 Anaptyxis CC- > CVC-: mulu-, šu -, cezp, hu /hut, ten-, uluter 4 Diphthongs eu > au: lautni ei > ai: tin (day < *tain- < pIE *deino- day), t(e)v-, tiu(r), cil (exception: lein) 5 Assibilation of palatalized plosives de, di; te, ti; ge, gi > ze, zi: ziv- (to live < *giv- < pIE *g ih3-we/o- to live), zi - (to paint > write < pIE *dik- to show), sar/-zar, zal g- > z-: zen-/zin- (to make, produce < pIE *genh1- to generate) -gn- > -sn-: rasna (people, Etruscan < pIE *h3r g-no- kingdom) Exceptions: -g- > -c-: ac-g > - : -al 6 Fricativization of word-initial aspirated voiced plosives b - > f-: fan- (to decree, declare < pIE *b eh2- to speak), ruva (brother < *hr- < *fr- < pIE *b reh2brother) g - > h-: hen, heva, hia (here < pIE *g e- this) d - > -: uluter (twins? brothers? < pIE *d h1leha- breast, suckling + pIE *-ter kinship suffix), i, ez-, tali a, es (exception: te-ce) 6 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 7 Aspiration (and later fricativization) of unvoiced plosives t > , (t): ca(u) a, ar, ui, u(n), esan, tu i, far (a)n(a)-, me lum, ur an, šu -, lu , cil , et/e , etera, hu /hut, ita/eta, nefts, trin, ten-, uluter Also note ... - > t... - (aspiration dissimilation): tu i, tali a k, k > c, (> h) (conditions for c/ /h alternation must still be identified): cap-, clan, ca(u) a, ces-, cezp/* usf-, ica, eca, zi -, hin -, hut, t(e)v-, sar/-zar, cil , culs, -c, - , hec-/he -, ur, tezan, -ce p > p, , f (> h) (conditions for p/ /h alternation must still be identified): apa, cap-, pava, repin i, far (a)n(a)-, pes, puia, nefts, far-/har-, huš/hus-, popa, hec-, hil-, -pi 8 Devoicing of voiced plosives d > t: tur- (to give as a present < pIE *deh3-r- gift), tin, t(e)v-, tiu(r), tam-, tezan, -ti g > k: slica- 9 Treatment of other non-initial plosives -d - > -t-: lautni (freedman < pIE *h1leud - free), vatie, -ti, etera (exception: -ud V- > -upV-: lupu to die, to go < pIE *h1leud - to go, go away, pass away, die; compare Latin -Vd V- > -VbV-) 10 Sibilants s > s, š (conditions for s/š alternation must still be identified): usil/ušil, es-, se , sem , a/ša, šu -, spura, -s 11 Glides and rhotics wo > u: unu, un, tiu(r) r > ar: matu 12 Vowels ai > i, e: tin, t(e)v-, tiu(r), cil , es- (exception: avil) o > u: ui, u(n), puia, me lum, hu /hut, mur-ce, spura, tur-, matu, ur an VC- > C- (when V- is unstressed): netš- (bowels, entrails < *enetr- < pIE *h1en-h1eh1tr- inner organs), cezp, ca (< eca), ta (< eta) -V > -∅ ∅: -c, - , cezp -iwi- > -i-: cil 7 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 13 Consonant clusters -dt- > -s- (as in Latin and Celtic): pes, ces-dgn- > -cn-: acnan-tny- >-i-: puia (wife < *putnya < pIE *pot-niha wife) tr- > c- (i.e. [tr] > [t ] > [tš] > [ ] = <c>): ci (3 < pIE *tri- 3) -tr- > -tš-: netš- (bowels, entrails < *enetr- < pIE *h1en-h1eh1tr- inner organs) -t(s)l- > -š-: huš/hus-tp > -zp, -sf: cezp/* usffr- > hr- > r-: ruva (brother < *fr- < pIE *b reh2- brother) -nk- > -c-: ma /mac -nk > -n: rin-/ren-ns- > -z-: ez-nt- > -t-: ati, etera -mn > -mr 2 > -r: nur (9 < *numr-p < *numn- < pIE *h1newh1mn 9), sar/-zar -ls- > -s-: esan -ks > -s: a/ša -rt- > -t-: mutana, matu, etv- 2 Greek Agamemnon was borrowed as A memrun (B 192) and Mémnon as Memrun (B 201) 8 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Etymologies of basic lexicon Introduction The choice of which terms should be considered as “basic lexicon” is not arbitrary (and was not made after finding IE etymologies, to make the statistics “look good”!). It is based on a standard list of basic terms which was designed in December 2008 to be applied to long-range comparisons. This list was used to sift through Etruscologists’ works and extract all terms in the standard list for which a gloss has been proposed. This process makes sure that all meanings assigned to Etruscan terms have been independently proposed by Etruscologists. Had I proposed new meanings for Etruscan words for which no glosses by Etruscologists exist, I might have incurred into circular reasoning. The result of this selection is a list of 138 basic terms. Only for 4 of them (3%) was it not possible to find an IE etymology. For 53 of them (38%) an etymology was found, but presented semantic or phonetic irregularities whereby it had to be considered as tentative. For 81, though, i.e. 59%, an IE etymology with regular sound correspondences and a good semantic match was found: Etymology quality Total % likely 81 59% tentative 53 38% none 4 3% Grand total 138 100% Here’s the English translation of the terms with a reliable IE etymology: alive, to live; and, also, or; bend, incline; bowels, entrails; boy, child; city; cook; day; door; father; field; fire; four; girl; grandfather; grandmother; grandson; hand; he, she, this, that (relative); here, now; home; house; I, me; inside, to enter, to penetrate; it, that (relative); July; l bertus, freedman; love, Amor (god of love); minus (in numbers); monument; moon, month; nine; one; our?; people; relative and interrogative pronoun; seven; six; stranger, client, slave; sun; Sun (as a god); the, this; there; this; three; thus; to ask (for?); to decree, declare; to die, to go; to generate, beget; to give (as a present), dedicate; to give, offer, dedicate; to give, offer, make; to grab?; to hold (a public office); to lie, be placed; to live, alive; to live, dwell; to look, see; to make, do, build; to make, do, move, reach, elevate; to make, offer; to make, produce; to paint > write; to place, set up; to put; to say; to take, grab, contain; to want, desire; tomb (monument?); twenty; twins? (brothers?); -ty (suffix for tens in numerals); village; year; you It is conceivable that some of them are loans from neighboring IE languages, but it is very hard to believe that all of them are IE loans, i.e. that such an impressive stock of basic lexicon could have been borrowed altogether, because basic lexicon is resistant to borrowing. Also note that these reliable IE etymologies include several verbs – and verbs are even more resistant to borrowing than nouns. NOTE: etymologies for the following terms were suggested by Guido Borghi (p.c.): ac-, lein, ur, mal-, mlesi-, sal-, san-, slica-, es; the rest are mine. 9 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Likely etymologies ac- to make, offer < pIE *h2eg- to drive by 1, 5; compare proto-Italic ag-e/o- to do, act > Latin ag to drive, do; this root may have merged with an Anatolian root found in Hittite a- to do, make, Lycian a- to do, make (1 sg preterite active a a, agã) -al -ty (suffix for tens in numerals) < pIE *leg- to gather, collect by 5 (semantics: collection / group) amin love, Amor (god of love) < pIE *h3mh3- to take hold of by 1 (for semantics, compare Latin am to love) + -in participial ending an he, she, this, that < pIE *(h1)i-, h1e- he, she, it by 1 + same -n suffix (accusative?) as in un, in, mini; compare Latin is, ea, id he, she, it apa father < pIE *p a-t r father, *pa-pa daddy by 1, with a- by analogy with ati mother ar-, er- make, do, move, reach, elevate < pIE *h3er-, *h1er- to set in motion (vertically / horizontally) by 1. The two pIE roots probably merged early in many IE languages (M-A 391). This seems to have happened in Etruscan as well: from *h3er we have the meanings to erect, elevate > build > make (compare ur an); from *h1er- we have the meaning move.To these roots we may also add pIE *h2(e)r- to join, fit > make (> Armenian aorist ar-ar-i I made).The conflation of all these pIE roots explains the wide range of meanings attributed to this Etruscan verb. Compare proto-Italic *or-i- to rise, Hittite ar(a)i- to lift, raise, Sanskrit r- to move atr- home < pIE *h2ehx-tr- hearth by 1; compare Latin trium, Russian vatra hearth, Avestan tarš fire avil year < pIE *h2ei-wo- time by 1, 12; compare Lemnian avi = Etr. avil-s. For -il suffix see F 184. Compare proto-Italic *aiwo- period, age > Latin aevum period of time -c, - archaic form -ka and < pIE *-k e and by 2, 7, 12; compare Latin -que and, Hittite -kku and cap- to take, grab, contain (> cap-ra urn or vessel) < pIE *kap- to seize by 7; compare proto-Italic *kap-i- to take (> Latin capi to take); Albanian kap catch, grab, seize car-, cer- to make, do, build < pIE *k er- to make, do, build by 1, 2 ca(u) a Sun (as a god) < pIE *kehau- to burn; heat + pIE *-tV- participial ending by 1, 7; compare Greek kai to burn, Toch B kau day ces- to lie, be placed < pIE *kh2d-to- fallen by 1, 7, 13; compare Latin c sus fallen, cad ver corpse ci 3 < pIE *tri- 3 by 13 cil people, nation < *civlt < *kiw-il-it- < pIE *kei-uo- friend, intimate, socially close + *-il-itextension (as in Latin) by 4, 7, 12; compare proto-Italic keiwi- society > Latin c v lit s citizenship culs door < *klau-s(V) < pIE *kleh2u-d- to shut by 7 + either pIE *-to- participial ending or Etruscan -s- participial ending; *klau- > cul- either by metathesis, or by anaptyxis (*kulau-) and syncope (cul-); compare Latin claud to close, cl vis door key -em- minus (in numbers)3 < pIE *h1em- to take by 1; Etruscan imperative = bare root: take away! Compare Latin em I take, buy, Latin em here you are < imperative *eme take! enas of us < pIE *no- we + Etruscan genitive –s (but whence e-?); compare Hittite -nnaš our es- to want, desire < pIE *haeis- to want, seek out by 1, 10, 12 3 Found in ci-em za rum 17, esl-em za rum 18, un-em za rum 19 10 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 et, e thus < pIE *h1i / h1e- it + pIE *to- that by 1, 7; compare proto-Italic i-to- thus > Latin ita thus etera stranger, client, slave < pIE * h1e-tero- the other by 1, 7 (semantics: *other > stranger > slave); compare Umbrian etr- other; or else < pIE *nd ero- lower (semantics: *inferior > client, slave); compare HLuv. andara-, Lyc., Mil. ntre- lower, Latin nferus lower fan- to decree, declare < pIE *b eh2- to speak by 1, 6 + -an verbal extension (as e.g. in far an-, ur an-, renan-); compare Latin f - to speak, Greek ph mí < ph mí I say far-, har- inside, to enter, to penetrate < pIE *per- to pass through by 1, 7; also peras crossing might belong here far (a)n(a)- to generate, beget < pIE *perh3-i- to provide + -tV- + Etr. –an verbal extension, by 1, 7; compare Latin partus birth; in F 103 far an is glossed as genius, spirit; with the same semantics as Latin genius < gen- to beget hen, heva, hia, ia here < pIE *g e- this by 1, 6; compare Latin hi-c this huš, hus- boy, child < pIE *put-(s)lo- son by 7, 13; compare Latin p sus boy, Armenian ustr son (< *usl < *pu-tlo-); or else < pIE *h2wes- to dwell, live, be if *h2w- > * u as in Tar un; compare Hittite huš- to live, hušu- fresh, raw hu , hut 4 < * utur < pIE *k (e)tw(o)r 4 by 2, 3, 7, 12, with deglutination of *–ur, understood as the multiplicative suffix (as e.g. in tun-ur single, zel-ur double) ica, eca, ca this < pIE *(h1)i-, h1e- he, she, it + *ke- here by 1, 7, 12; compare Hittite k - this in it, that < pIE *(h1)i-, h1e- he, she, it by 1 with same -n suffix (accusative?) as un, an, mini ipa relative and interrogative pronoun < pIE *(h1)i- demonstrative stem + *k V- interrogative and relative stem by 1, 2; this is a compound, as shown e.g . by accusative inpa. Compare Oscan, Umbrian pi- interrogative pronoun. For a similarly-built compound (demonstrative + relative) also compare Greek hopoîos. Ipa later developed into a declarative conjunction, just like late Latin quod, quia ita, (e)ta the, this < pIE *(h1)i-, h1e- he, she, it + *to- that, this, it by 1, 7, 12; compare Latin iste that (near you) lautni l bertus, freedman < pIE *h1leud -(ero)- people > free + Etruscan -nV extension by 1, 4, 9; compare Latin l bertus freedman lein to die < pIE *leih2-, li-n- to stop, disappear by 1, 4 lupu- to die, to go < pIE *h1leud - to go, go away, pass away, die by 1, 9; pIE *-d - > Etruscan -pafter /u/ has a perfect parallel in pIE *-d - > Latin -b- (e.g. l ber free < pIE *h1leud ero-). Lupu derives from zero-grade root *h1lud - (otherwise **laupu should be expected). Compare OHG beleite burial, Greek loiteú to bury, Avestan ra to die. lu field < pIE *lu-to- dirt by 7; compare Latin lutum mud, dirt, clay, proto-Celtic *lut mud, dirt, Basque lu-r earth, land, soil, dirt -(u)m and, also, or has several Anatolian correspondences: Hitt. -ma and, but, Pal. -ma but(?), Lyd. -m(-) enclitic particle of uncertain meaning, Lyc. me sentence-initial particle man- to commemorate, manim monument < pIE *mon-eie- to make think of, remind; compare Latin monumentum, monimentum statue, memorial, tomb masan July < masu 5 (see mac) men- to give, offer, make < pIE *mehan- hand by 1; compare Hittite maniyahh- to hand over, Oscan manafum to give (1st sg. pf?), Basque eman to give 11 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 mi I, mini / mene / mine me < pIE *h1men- me by 1; K 112-113 reconstructs pIE *h1m(e)n- for the accusative and oblique cases, which then developed into *h1mn- > *h1m- in non-Anatolian pIE. For the nominative case, Hittite preserved pIE *h1egH, whereas HLuwian, Lydian and Lycian all analogically extended the oblique root to the nominative case as well. This is a key innovation that Etruscan shares with non-Hittite Anatolian. Compare HLuw. / mu/ I, / min/ me, Lydian, Lycian amu I mul(u)- to give, offer, dedicate < proto-Anatolian *mel- think; compare CLuv. malhašša- ritual, sacrifice, Lydian m w nda- offerings (for the dead), HLuv. /malwa-/ ritual offering, Lyc. mlesacrifical offering?, Mil. mla-, mle- sacrifice, offering – another impressive set of exclusive correspondences between Etruscan and Anatolian mur-ce lived, dwelled < pIE *morh2- to remain by 12; compare Latin moror to delay, remain nefts, nef(i) grandson < pIE *h2nep- t- grandson, nephew by 1, 7; this is not a Latin loan; in fact, na o - is also present in Lemnian, which is a sister (not a daughter) language of Etruscan, so cannot have borrowed from Latin netš-, ne š- bowels, entrails < pIE *h1en-h1eh1tr- inner organs by 1, 12, 13 nur 9 < *numr-p < *numn- < pIE *h1newh1mn 9 by 1, 13; sem- 7, cez-p, * us-f- 8 and nurall seem to share a common -p /- /-f ending (of uncertain origin; borrowed from Oscan / Umbrian p and?), so the root is nur-. According to M-A 308, 314-315, it is not clear whether pIE 9 ended in *-n or *-m. Etruscan provides evidence that it originally ended with both, ie. *h1newh1mn > *numn > *numr, with *mn > *mr as in Greek Agamemnon > Etruscan A memrun (B 192) and Mémnon > Memrun (B 201). In other IE languages, *-mn was simplified very early to either -n or -m (also see mini for early pIE *-mn- > *-m-). papa grandfather < reduplicated apa father; compare teta grandmother < reduplicated ati mother pava boy, youth < pIE *pau- small, child by 7; compare Latin puer boy, Basque haur child popa a cook < pIE *pok - a cook by 2, 7; compare Latin cocus a cook, Latin pop na eating house (< Sabellic), Myc. ato-poqo bread baker rasna people, Etruscan < pIE *h3r g-no- kingdom by 1, 5; compare Latin regnum kingship, reign, kingdom, Avestan r zan- rule, direction ren-an- to grab < ren hand (see rin) + denominative –an– extension (F 289) repin i bend, incline < pIE *(w)rep- to turn, incline by 1, 7; compare Greek hrép to incline oneself rin-, ren- hand < pIE *renk- to collect by 1, 13; compare proto-Balto-Slavic *rankâ hand, Burushaski -rén hand, Lithuanian renkù collect a, ša 6 < pIE *seks 6 by 1, 10, 13; it is not clear whether loss of final *-s is a regular phenomenon, but in this specific case it may be due to subsequent sem 7: *šas, sem > ša, sem . Compare Old Irish sé 6, Basque sei 6 sem 7 < pIE *sept 7 by 1, 10, with same - /-p ending as cez-p 8, nur- 9 spura city < pIE *sper-, *spor- to wrap around > to surround, enclose, fortify by 10, 12; compare Armenian p arem to enclose, surround šu (-) to put, šu i tomb (< *monument) < pIE *sth2- to stand; to place by 1, 3, 7, 10; compare Latin statuere to erect, statua statue sval alive, sval- to live < ziv- to live, alive (which see) + -al (adjectival?) extension tali a girl < pIE *d al- sprout by 6, 7 + - a feminine suffix (compare lautni- a l berta < lautni l bertus) 12 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 tam- house < pIE *dem- / dom- house by 1, 8; compare Lyc., HLuv. tama- house, building ar there < pIE *tor there by 7 te-ce (he) placed, set up < pIE *d eh1- to put by 1, 6; compare proto-Italic *f -k- to do (aor.), Umbrian face he did, Greek é-th -ke he put, placed, Hitt. ti- to put, place, Lyc., Mil. ta- to put, place tenu, tenve he held (a public office), ten as having held (a public office)< pIE *tn-eh1- to hold by 3, 7; compare Latin tenuit he held teta grandmother < reduplicated ati mother; compare papa grandfather < reduplicated apa father tin day < pIE *deino- day by 4, 8, 12 tiu(r) moon, month < pIE *dei-wo- < *dei- to shine by 4, 8, 11, 12; compare Luwian tiwa- Sun-god. For a similar semantic shift, compare pIE *leuk- to shine > *louk-sneha > Latin l na moon trin say! (or we say?) < pIE *t(e)r- say, speak out by 7; compare Hittite ter- to speak, to state, tarueni we speak u(n) 1 < *this one < pIE *to this by 7, 12; a perfect semantic parallel is pIE *h1oi-(no/ko/wo) one < pIE *h1ei- this one (M-A 310) u-i here, now < pIE *to- that, this, it by 7, 12 + locative –i uluter twins? brothers? < pIE *d h1leha- breast, suckling + pIE *-ter kinship suffix by 1, 3, 6, 7; compare Latin f lius son tur(u)- to give (as a present), dedicate < pIE *deh3-r- gift by 1, 8, 12; compare Greek d ron gift, Armenian tur gift, OCS dar gift tu i village, tu i-na public < pIE *teuteha- people by 7 4; compare proto-Italic *tout - town, society > Ven. teuta civitas, Umbrian tuta town, society. An alternative etymology might involve Hittite tuzzi- army < pIE *d h1-utiun thee (accus.) < pIE *wo- you by 11 + -n accusativev unu( ) you (accus. plur.) < pIE *wo- you by 11; compare Luw. unz- you usil, ušil sun < pIE *haeus- dawn + *séhaul sun (i.e. sun at dawn, rising sun), but vocalism is hard to explain if this word derives straight from pIE; this might be a loan < Sabine ausel sun; in fact, usil is attested in southern Etruria, namely in Vulci, in Caere and near Tarquinia val look, see < pIE *wel- to see by 1 vatie- to ask (for) < pIE *g ed - to ask, pray by 1, 2, 9 vers- fire < pIE *wer- to cook, boil by 1; compare Hittite war- to burn za rum 20 < *zartrum < zar 10 + *etr- other (see etera) + -um inanimate noun suffix (FM 76-77); semantics: 20 < *the other ten zen-, zin-, zina- to make, produce < pIE *genh1- to generate by 1, 5; compare Lat. gigno to produce, create, Basque egin to do, make, Sanskrit jan- to beget ziv- to live, alive < pIE *g ih3-we/o- to live by 1, 2, 5; compare Lemnian siv-, Sanskrit j v- to live, Old Persian va- to live, Latv. dz v- to live, OCS živ- alive, life zi - to paint > write < pIE *dik- to show by 5, 7; compare German Zeichen sign, zeichnen to draw 4 -u- vocalism may either be explained as zero-grade *tut- or as a loan from Umbrian (otherwise **tautshould be expected from *teut-, as in lautni) 13 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Tentative etymologies acnan- to beget < pIE *h2ed-gn- to beget in addition by 1, 13 + -an- common Etruscan verbal extension; *g should regularly yield **s or **z, so one must posit that *g was depalatalized in the *dgn- context. Compare Latin agn sc to be born in addition al(i)- to give (as a present), to offer < pIE *h2l-i- to kneel down by 1; semantics: kneel > offer (to gods); compare Hitt. haliie/a- to kneel down am- to be? to hold? < pIE *h2em- to hold, contain by 1: usually to be is ∅ in Etruscan. Am- is usually glossed to be, but most contexts where am- is used are compatible with a meaning to hold (a public office, or the content of a vase) anaven- to beget? to adopt (a child)? Compare Hitt. anna-wanna, Hluw. nna-wann(i) stepmother? Ci clenar [...] anavence (Ta 1.169) is usually glossed she generated three children, but if this etymology is correct the phrase should be intepreted as she adopted (became stepmother of) 3 children. But see nacnva for an alternative hypothesis. ati mother < *anti < *anati < proto-Anatolian *Honno- mother by 13; compare HLuw. /anati/ mother. If this etymology is correct, it has no non-Anatolian IE parallels, and is therefore additional evidence for an Anatolian origin of Etruscan cel earth, land, "Mother Earth" (a goddess) < *cultivated land < pIE *k el- to turn by 1, 2; compare Latin col to cultivate cezp, * usf- 5 8 < *kuzp < *kut- + -p < *kto- < pIE *h3(e)kteh3 8 by 1, 3, 7, 12, 13; with same -p//-f suffix as sem- 7, nur- 9; cezp has /e/ by analogy with sem . For a similar phonetic evolution, compare Toch A okät 8 clan, clen-, celen- son < *the growing one < pIE *k(e)l-n-H- to rise by 7 cver gift, offering, sacred object < *aku-er < ac- to make, offer (which see); for –er extension, compare fl-er victim, sacrifice from a verb fel- (F 148); for cv- < *aku- see cvil cvil, vil, Rhaetic akvil gift, offering < *aku-il < ac- to make, offer; compare cver ei(n), en not < * < pIE *ne not eitv-, etv- big? < pIE *h2erd( )u- high by 13 (but usually *h2e- > a-); compare Latin arduus tall, high etnam and, also, again, thus < pIE *h1eti beyond (> Latin et and); for -nam compare Hittite namma then, next, furthermore, of uncertain origin hec-, he - to put, place < pIE *pehag- / *pehak- to attach, fasten by 1, 7; compare Latin pang to insert firmly, fix, Greek p g-nu-mi to insert, drive in, thrust, fix, connect hil own?, hilar property? hilar (u) owner? < pIE **pel- to sell by 7; if the pIE etymology is correct, then the right (or original) meanings would be to sell, something sold, seller respectively hin -(a) below, under, behind < pIE *ke- this by 7; compare OHG hintana, Gothic hindana, Old English hindan behind ur full < pIE *kuh2-ro- < *keuh2- to swell by 1, 7 ma , mac, 5, masu, mašu 5 together < pIE *penk e by 13, with *p > m as in Greek elephant- > Latin elementum via Etruscan (B 73)? Or else < pIE *mehan-k e and the hand in the counting sequence 1, 2, 3, 4 and the (whole) hand = 5 5 found in the gloss osfer October (B 191) 14 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 mal- to look < pIE *melh3- to appear by 1 matu uncertain meaning < pIE *mr-to- dead by 11, 12, 13; based on the context (AT 1.96), the meaning dead could be appropriate. Compare mutana, murs me lum people, nation, district, territory < pIE *meh1-tlo- measurement by 1, 7, 12 + -um inanimate noun suffix (FM 76-77). Semantics: *land with measured borders me people, league, district < pIE *meh1- to measure by 1 (if the original meaning was district; compare me lum), or else < pIE *mVnus man (if the original meaning of me was people); but neither etymology accounts for mla , mlac, malak beautiful, good < pIE *mel- good by 1; this etymology could be improved if other IE reflexes with a -k- extension could be found (unless -ak is an Etruscan affix). An interpretation of mler as ml-er well (F 189, 231), if correct, would confirm that the Etruscan root is ultimately ml-. Otherwise, if p > m is valid (as in mac 5 if < *penk e), a parallel with Latin pulcher is possible, and would also provide a parallel for -k- extension. mlesi- hill? < pIE *melh3- hill (but whence -s-?) murs, murš sarcophagus, urn < pIE *mer- to die. Compare matu, mutana mut(a)na sarcophagus < pIE *mr-to- dead by 13 + -na adjectival suffix (F 58). Compare matu, murs. nacn((u)v)a in ati nacn((u)v)a grandmother; various alternative comparisons may be attempted with Anatolian (none of them conclusive); one is Hitt. anna-wanna, HLuw. nna-wann(i) stepmother, from pan-Anatolian anna mother + -wannV (also found in HLuw. t ta-wann(i)stepfather); if proto-Anatolian *Honno-kwanna can be reconstructed (compare sound law 1.4.3.13, K 69-70: -Rk C- > -RuC-), we'd have in Etruscan *annak na > nacn(v)a. In this case, the right meaning would be stepmother, and the etymology for anaven- (which see) would be wrong. Alternative comparisons with similar phonetic shapes might include Hitt. anna-neka- sister (neka) by the same mother (anna), Hitt. nakkiyann- dignity, importance, power; Hitt. nekna- brother, Hitt. pappan(n)ikna- brother sharing the same father (< *pappa- father and nekna- brother), possibly to be compared with Etruscan hapax apa nacna (Vt 7.2). To check whether any of these possible correspondences makes sense, one will have to analyze each context in which ati nacna occurs. neri water; compare modern Greek nerò water, if not < nearò(n) fresh (water) pes land? pIE *ped-to- fallen by 1, 7, 13; compare Latin pessum to the ground, Greek pédon ground, floor puia wife < pIE *pot-niha wife by 7, 12, 13 pul splendid, birght < pIE *pVl- ? bright, splendid? Etruscan pul is probably thesame root as in pulum stars ; the pIE reconstruction is my conjecture, connecting Latin place to be pleasing and pulcher beautiful (both without good etymologies, and both appearing to have a -k- extension) pul-um star < pIE *pVl- ? bright, splendid? (see pul) + -um inanimate noun suffix also (FM 76-77). Note that a very similar word, fulum- va, does not seem to mean star (B 176, 178, 193). If fulumva is a different word, it may be connected to fler offering sacrifice restm, reštm (cultivated?) land: a loan < Latin restibilis sown or planted every year ruva brother < *hr- < *fr- < pIE *b reh2- brother by 6, 13 (but usually *eh2 > a) sa- this? pIE *so, seha that by 1 sal- make, carry out < pIE *sel- seize by 1 san- deceased? < pIE *senh2- to reach, accomplish by 1 15 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 sar(-), ar, -zar 10 < *zear < *zehamr < *dehamn < pIE *dekmt 10 by 1, 5, 7, 12 se , sec daughter < pIE *seu(hx)- to beget by 10, or pIE *seuk- to suck? si(n)- to hear, sense? < pIE *s(e)nt-i- to notice; compare Latin senti to sense, feel slica- to separate? < pIE *sleig- to split by 8 špan- plain (noun): a loan from some Italic source with an s-like prefix (e.g. Latin expandere to spread out?) < proto-Italic *pand- to spread out, extend < pIE *pt-n(è)-h2- to spread spel cavity: compare Latin sp lunca cave, Greek sp laion cave es anger? < pIE *d wes- to breathe > spirit by 1, 6, 7 esan morning, dawn < *tel-sa-n < pIE *telh2- to raise by 1, 7, 13 (and *-ls- > -s-); -sa- might be either Etruscan -as(a) active present participle suffix (F 289) (semantics: the raising (of the sun)) or the same extension of this root as in Latin toler - < *telas - < *telh2-st(e)v-, tva to show, see? Compare Lyc. *tewe- eye (ML 64), Mil. tewe- to encounter, see? ez-i(n(-)) to kill, sacrifice < * ens- < pIE *d wenh2- to die by 01, 02, 06, 13 + *-s- causative suffix; compare Greek thánatos death; it is not clear whether Lyc. tezi-, ezi- sarcophagus, coffin is related, too i water < pIE *d eh1(y)- to suck > drink by 6; compare Sanskrit dh tave to drink tiss, tis lake? is probably a ghost word; this seems to be a hapax, and its segmentation is debated, too (F 210 posits either neitiš or itiš or tiš) ur an(-) to make < pIE *h3er-to- risen by 1, 7, 12 + -an- verbal extension; semantics: to erect > to build, make -vis to look, examine? (in netš-vis haruspex see netš-) < pIE *weid- to see, know zal, zel-, es(a)l- 2 < *zi- + < *di- < pIE *dwi- 2 by 2, 5, + -al suffix as in Latin du lis of two; compare Greek di-, OHG zwi- 2 zuslev- animal < *ziv-s-lev- *live animal < ziv- live (which see) + pIE *leuhx n animal; compare Tocharian B luwo, Tocharian A lu animal 16 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Basic lexical items with no pIE etymology The following terms are part of the basic lexicon, but no IE etymology could be found for them. ce a above?? The meaning above is not certain: in B 165, F 160, 165, 185 ce a is glossed as on behalf of, in favour of and would be a postposition. In B 215 it is glossed ceremony, right, law. A reduplicated form of eca this (*eca-eca > ce a) might also be compatible with some contexts (e.g. B 115) muc- inside? sanisva, sanišva bones? Compare san- deceased? i-(e)m, i-s each, all Non-basic terms with a pIE etymology Non-basic terms are more likely to be borrowed than basic terms, so any attempt to classify Etruscan should be mainly (or even exclusively) based on basic lexicon. Yet the following nonbasic terms have interesting IE etymologies or comparanda as well: maru(n) maron (name of magistracy): compare Lyc. maraza- judge, arbitrator, Mil. mara- law *pursena king's name, Porsenna: compare Hitt. paršna- leopard-man sac- sacred: compare Latin sacer sacred, Hitt. š ki- sign, omen španti dish < pIE *spond- to libate; compare Hitt. išpant- to libate, pour, sacrifice, išpantuwalibation vessel tar un-, tar na-, etc. Tarquin(ia) < pIE *terh2-u-ent- > proto-Anatolian *trH ent- overcoming; compare Hitt. tarhu- to prevail, conquer, be powerful; Luw. Tarhunt- Storm-god, Lycian trqqñt/trk nt-/ Storm-god zil- to rule, hold office: compare Hittite šalli- head, chief, notable 17 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Etymologies of grammar elements I could find 20 grammar morphemes for which Etruscologists have proposed a gloss: 95% of them have an IE etymology: Etymology quality likely tentative none Grand total Total 11 8 1 20 % 55% 40% 5% 100% Likely etymologies -ce active past suffix in verbs: compare Greek -k- aorist and perfect suffix, and possibly also Lycian - a 1st p sg preterite active ending < proto-Anatolian */-Ha/ -i feminine suffix in proper nouns < pIE *-iha fem. suffix by 01 -i locative < pIE *-i locative -al genitive < pIE *-(o)lo- belonging / pertaining to; this pIE possessive suffix is found in Lydian, Hittite and Latin; denominal adjectives in Lydian are used instead of genitive; e.g. siuv-ala of the God -n accusative (in pronouns only) < pIE *-m accusative ending > Hitt. -n accusative -pi to, towards (postposition; follows accusative case) < pIE *h1opi at, by; compare Latin ob towards, Greek epì on, at -s genitive < pIE *-s genitive -s absolutive (in male nouns, in archaic southern Etruscan only) < pIE *-s nominative Lemnian -sio patronimic suffix < pIE *-syo genitive - (i), -t(i), -te in (postposition; follows locative case) < pIE *-d e, -d a, -da locative suffix found in adverbs of place and time by 8, 9 -tra outside (postposition) < pIE *trh2-nt-s crossing; compare Latin tr ns across, through, Umbrian tra on the other side of Tentative etymologies -an(i)- denominal verb extension: compare Hitt. -anna- / -anni- imperfective (i.e. progressive, iterative, durative, distributive and ingressive) suffix < pre-Hittite *-ann(a)i-as, -asa active present participle < pIE *-ent-s / *-ent-ia active present participle (masc., fem.); compare Greek participles with similar phonetic developments -is ablative; compare Latin –is abl. plur., Hitt. –aš dat./loc. plur. - e passive past marker: compare Hitt. -hha- 1st person sing. preterite middle ending, Lycian 1 sg. pret. middle ending - agã 1 sg. pret. middle ending -r human plural: compare -r third person plural in verbs, found in Italic, Indic, Tocharian and Hittite? -ri for; in favour of (postposition; follows locative case) < *-r locative (and occasionally allative) suffix in adverbs -(u)s verb suffix < pIE *-ske/o- > CLuw. -zza-, Hluw. -za-, Lyc. -s- imperfective suffix 18 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Grammar morphemes with no pIE etymology - va, -cva, -va non-human plural has no apparent IE parallel, but ‘Eurasiatic’ parallels can be found in -ku plurals in Eskimo, Aleut, Chukotian and Gilyak (Nivkh) Shared innovations Etruscan shares the following morphs with non-Hittite Anatolian languages: • -al genitive: compare Lyd. • mi I, mini, mene me: compare HLuw., Lyd., Lyc. • mulu to give, offer, dedicate: compare CLuv., HLuw., Lyd., Lyc., Mil. • maru(n) a magistrate: compare Lyc., Mil. • ati mother: compare HLuw. • tev- see: compare Lyc, Mil. whereas the following are shared with Hittite as well: • -m and: compare Hitt., Lyc. • tar (u)n- Tarquinia: compare Hitt., Luw., Lyc. • et-nam also: compare Hitt. Conclusions Borrowing of very high percentage of basic lexicon and grammar morphemes is unrealistic. The most likely explanation of regular correspondences between Etruscan and pIE lexicon and grammar is that Etruscan is Indo-European. Etruscan also shares some notable innovations with non-Hittite Anatolian languages. Interestingly, non-linguistic evidence seems to point in the same direction: recent genetic studies (Piazza et al., 2007-2010) confirm that Tuscan women, men and cows (as well as Raetian cows) display statistically meaningful resemblances with their Anatolian counterparts. Etruscologists also consider Lemnian and Reatian to be related to Etruscan. Interestingly, Lemnos is very close to Anatolia. As for Raetian, it may have been separated from Etruscan by Celtic invasions. 19 G Forni - Etruscan - DRAFT - v03f2 Copyright © Gianfranco Forni 2010 Next steps I believe that the case for Etruscan as an IE language has been made, and it is now a matter of working out the details (amending errors, improving sound laws, finding new etymologies, etc.) – something that Etruscologists and Indo-Europeanists (notably, Anatolists) are better equipped to do than I am. The real acid test for this theory is now to use pIE as a tool to interpret Etruscan terms which have remained obscure so far: if this attempt succeeds, the theory of Etruscan as IE will turn out to be not just descriptive, but predictive as well. As a contribution to this exciting enterprise (for which I do not have adequate skills), I provide below a table, to help future researchers to trace back wordinitial Etruscan consonants to their possible pIE origins: Etruscan word-initial consonants c possible pIE origin k, k, k , tr, (g ?, g?) k, k, k , (g ?, g?) z ze, zi g de, di dwe, dwi te, ti ge, gi g e, g i p f b, p h k, k, k , g , p, (g ?, g?) t d, dw, (t) t, d , d w v w, g m m, p-?? 20