Some half baked ideas

Miguel Carrasquer Vidal mcv at
Tue Nov 26 10:36:43 UTC 1996

Dear historical linguists,
The following are three (slightly edited) articles of mine,
posted in the last few weeks on sci.lang and/or sci.archaeology.
I think some of the ideas are at least original, although I
unfortunately have had neither the time nor the means to look
for references or do a more solid investigation.
I thought it would be interesting to try these ideas out on
HISTLING, to see what your reactions are.  Dead ends?  Worth
investigating further?  References?  Any comments would be
most welcome.
The Etruscan-Phoenician Pyrgi Bilingual
[discussion on Woudhuizen's "decipherment" of the Lemnos stele]
I'm not convinced of the translation given by Woudhuizen in "Lost
languages", but I have no alternatives to offer.  I really don't
understand the Lemnos stele too well.
In the same book, Woudhuizen also tackles the Pyrgi bilingual,
three golden plaques from Pyrgi (Caere), one bearing an inscription
in Phoenician, two in Etruscan.  The Pyrgi bilingue I know better.
The Phoenician text says something like:
"To the lady Astarte. This is the holy place that Thefarie Velianas,
king of Caere has made and given, in the month of offerings to the
Sun, as his private gift (consisting of) the temple and its
Because Astarte has favoured her faithful in the three years of his
kingship, in the month of Dances [KRR], on the day of the funeral
of the Goddess.  And may the years of the statue in her temple be
the years of the stars here."
Woudhuizen translates the Etruscan as:
This temple and this statue, Thefarie Velianas, legislator of the
senate (and) people, has built (it) for the Lady Astarte and has
given the temple-complex(?) to her during (the month) Cluvenia on
account of two obligations: because she has favoured (him) on land:
in the third year (of his reign), (during the month) KRR, on the day
of the funeral of the Goddess, (and) because she has favoured (him)
at sea: during the praetorship of Artanes (and) the sultanate of
Achasveros (=Xerxes).
And may whatever (number) of stars yield to (whatever number) of
years for this statue.
Thefarie Velianas has built the precinct for the Goddess Athena
(and) has offered (it) as a sacrifice during the month of offering(s)
to the Sun.  May whatever (number) of stars be sporadic as compared
to whatever (number) of years for this temple."
This is what I make of the Etruscan:
ita tmia            this temple
ica-c heramasva     and this statue
vatieche            have been dedicated
unialastres         to Uni-Astre
themiasa            "caring for" (= curator, "curating"?)
mech thuta          res publica
Thefariei Velianas  Thefarie Velianas
sal                 Sol??
cluvenias           offerfeast ? (cleva=gift, offering [see B])
turuce              he gave
munistas thuvas     of his own gift (thuv- = own [*])
tameresca           temple building? (tam-eresca ~ tm-ia, cf. IE *dom-?)
ilacve              as well as ?
tulerase            enclosure? (tular=border)
nac ci avil         as three years
churvar             Churvar (month) (cf. Phoenician KRR)
tes'iameitale       ? she favours him ?
ilacve              as well as ?
als'ase             ?
nac                 as
atranes             ?
zilacal             of the "zilac" (= praetor)
seleitala           ? of the goddess ?
acnasvers           funeral? cremation? (verse=fire)
itani-m             and so
heramve avil        the statue('s) year(s)
eniaca              just-like
pulum-chva          the star-count (-chva is collective suff.)
nac                 so
thefarie veliiunas  Thefarie Velianas
thamuce cleva       made a gift
etanal              on the idus ?
masan               Masa (a month)
tiur                month
unias               to Uni
s'elace             he dedicated
vacal               (his) offering ?
tmial               of the temple
avil-chval          of? the year-count
amuce               may it be?  (-ce = aorist imperative?)
pulum-chva          star-count
snuiaph             great-er (-aph=comparative)?
I'm not claiming this is a full translation of the Pyrgi texts,
and the middle part of A is still very obscure, but I'm afraid
dragging in Xerxes and his uncle doesn't help much.  I think A
and the Phoenician plaque *are* bilinguals, although it's obviously
not a word-for-word translation of one into the other.  B seems
like a "reader's digest".
[*] Woudhuizen translates thuvas as "two", just like Zacharie
Mayani in his hilarious book "The Etruscans Begin to Speak"
Woudhuizen should know better: "thu" is "1".  Pity of IE *dwo:,
but that's just the way it is.
But why "his own"?  The following funerary inscription (TLE 619)
should explain:
"cehen suthi hinthiu thues' sians' etve thaure lautnes'cle caresri
aules' larthial precathuras'i larthialisvle cestnal clenaras'i ..."
Beekes translates: "This subterranean tomb for the first father
[etve?] for the family grave has been built by Aule and Larth of
the Precu family, sons of Larth and Cestnei ..."
"The first father" (thues' sians') makes no sense (neither would
Woudhuizen's "second" father).  "Their own father (pater suus)" fits
much better, like it fits to translate "munistas thuvas" as "own (private)
gift" in the Pyrgi inscription (for "munistas" cf. Latin munus, muneris
(*munes-) "service, tax, gift").
Further comments: I have no Phoenician at all, so I'm relying on
the two (different) translations given by Beekes "De Etrusken spreken",
1991, and Best/Woudhuizen "Lost Languages from the Mediterranean", 1989.
I'm none too sure about the Italicisms "sal" and "munistas" and the
Phoenicism "churvar", but they just fell into their slots by treating
the text as a real bilingual (as in: the Phoenician says SH-M-SH, so
where is it in Etruscan?)
Armenian plural -k`
[discussion of IE numerals, Armenian c`ork` "4"]
The real mystery is that in Armenian, initial s-, medial -s- and final -s
normally drop (there are a few cases of initial h- derived from s-, which
suggest that /h/ was the intermediary stage).  A good example is the word
*snusos "daughter-in-law", which becomes <nu> in Armenian, genitive <nuoy>
(*snusosio > *(h)nu(h)o(h)i(o)).
The final -s of the nominative sg. and the genitive sg. has been lost.
But the plural -s (also in verbal endings -mk` (*-mos) , -yk` (*-tes))
remains as -k` (i.e. aspirated /k/).  Why?
Many explanations have been offered: the -k` is a (non-IE?) suffix quite
unrelated to -s, the -k` is a "reinforcement" of final -h under grammatical
pressure to maintain the distinction between sg. and plural, etc.  None of
it too convincing. [Reference: C. de Lamberterie BSL 74, 1979].
The only solution I can think of is that final -s after (long) -u- (in the
u:-stems and the o-stems [*o:s> us]) became labialized *sw, and spread from
there to the other nouns.  We know that all labialized voiceless consonants
(*sw, *tw, *kw) merged into *kw in Armenian, and give k` [or palatalized c`]
(e.g. k`oyr < *swesor "sister", k`un < *swopnos "sleep"; k`an < *kwam "than",
c`ork` < *ketwores; k`ez < *twe-ghe "you (acc.)").  It is interesting to note
that a /k/ (unaspirated k, usually corresponding to PIE *g) appears before
suffixed -n in some nouns such as jukn /dzukn/ "fish" (*ghdhu:s), mukn "mouse"
(*mu:s), armukn "elbow" (*arHm-us?) unkn "ear" (*(a)us-, pl. akanj^k` < *),
cf. akn "eye" (*okw-: one would expect *ak`-n).  Other u(:)-stems appear as
pluralia tanta (mawruk` "beard" (*smakru-), srunk` "shin" (*kru:s)).
Finally, there are some cases of -s > -r in Armenian (vb. 2sg. pret. -ir, -er,
certain -u stems like asr "fleece", t`anjr "thin"), maybe through *zw > *dw.
We know that initial *dw in Armenian develops to (e)rk-, as in the word
*dwo: > erku, which is another mystery in itself.
Further comments: I need to take another look the nouns and adjectives in
-r (n-stem pl. in -unk` I seem to recall?).  Unfortunately, I don't have
a Classical Armenian grammar handy at the moment.
Can somebody explain what Pokorny means when he says "arm. (mit expressiver
Geminata) <akn>, Gen. <akan> ,Auge etc.' "?  Any basis for that assumption?
Emesal a Sumerian Prakrit?
piotrm at (Piotr Michalowski) wrote:
>For non-Sumerologists I would like to note that these discussions often have
>referenec to eme-sal, a supposed "dialect" of Sumerian.  There are
>syllabically spelled eme-sal ("thin language") words found in certain myths in
>which women speak, and in liturgical texts that belonged to the repertoire of
>the gala priest (lamentation).  It is not clear, for technical reasons, if
>certain texts are to be read all in es or only partially,  There is no
>indication that this was ever a spoken natural dialect, and appears to be a
>specific way of pronouncing words in ritual.  It is attested relatively late,
>primarily in the later part of the Old Babylonian period in texts from
>northern Babylonia, that is after Sumerian was long dead and mostly, but not
>only, in an area in which it probably had never been spoken.  In the past it
>was considered a woman's language, but that was because the sal reading of the
>WOMAN sign was not properly distinguished from the munus reading.  sal is an
>adjective that means "thin."  emesal is not a good basis for the
>reconstruction of early Sumerian, as opposed to the "standard" language
>emegir, which means "native tongue."
Both should be taken into account, if possible...  Whether Emesal is a dialect,
a sociolect, a religiolect or a sexolect is interesting in itself, but the point
is: it's not a mere invention, and certainly not a Babylonian one.  The "sound
are too complex, partly reflecting hesitations that also occur in Emegir (n ~ l,
g~ ~ m, g ~ b) partly unique (AFAIK) to Emesal (n ~ sh, de,di > ze,zi).  If all
n-'s would become sh-'s, well and good, but that is not the case.  Or if it only
affected "cultic" words, but that doesn't seem to be the case either.  Nor is it
just another case of palatalization (cf. nundum ~ shumdum "lip").  I'm not sure
Emesal was either, "a literary dialect" would be the safest term.  It's not more
archaic than Emegir, except maybe in its treatment of Emegir "n" (=n, l or sh).
The palatalizations (udu > eze "sheep", etc.) certainly look like innovations.
It is very interesting to compare with Sanskrit/Prakrit, in the 5th. century AD.
Quoting from Michael Coulson's introduction to "Teach Yourself Sanskrit":
"By now Sanskrit was not a mother tongue but a language to be studied and
consciously mastered.  This transformation had come about through a gradual
process, the beginnings of which are no doubt earlier than Pa:n.ini himself.
Something of the true position must be reflected in the drama, where not merely
the characters of low social status but also the women and young children speak
some variety of Prakrit."
The Prakrit of the plays is an artificial "dialect" itself, only loosely based
on the local vernaculars.  Still, it reflects most of the phonological
of Middle Indo-Aryan, while at the same time standing closer to pre-Pa:n.inian
than Sanskrit itself in some respects (e.g. the l ~ r variations).  We even have
religious connection, given that Pa:li Prakrit was chosen by the Buddhist and
Jainist religious reformers as their lithurgical language over Sanskrit.
Maybe Emesal is a "tempolect" as well, a stylized vernacular "spoken" by female
literary characters and chanted by "folk-religion" priests, at a time when the
literary dialect was already a dead language.  The difference with Prakrit would
be that the vernacular was dead too (or at least dying).
Further comments: some closer study of the role of Prakrit
in Sanskrit drama would be necessary to see if the analogy
with Emesal holds.  A preliminary (phonetic) study of
Emesal is given below.
Appendix: Analysis of the Emesal vocabulary in Thomsen
The following is an analysis of Emesal phonetics in relation to the main dialect
(Emegir), based on the Emesal wordlist in Marie Louise Thomsen's "The Sumerian
language", pp. 285-294:
Emegir          Emesal           English
[Words appear to have different roots]
(en             umun)           "lord"
(guza           ash.te)         "throne"
(g~itlam        mu.udna)        "spouse"
(nig2           ag~2)           "thing"
(nitadam        mu.udna)        "spouse"
(tum2/de6       ga)             "to bring"
[Phonetic alternations]
a.gar3         "field"
alim            e.lum           "aurochs"
a.nir         "lament"
arad            e.ri            "slave"
dag~al        "wide"
dig~ir      "god"
dish  , di.ta    "one"
dug3            ze2.eb          "sweet", "knee"
dugud        "heavy"
dumu            "child, son"
en3..tar        ash...tar       "to ask"
engar           mu.un.gar/.g~ar "farmer"           Am.anki         "Enki"
En.lil2         Mu.ul.lil2      "Enlil"
e2.g~ar8        a2.mar, e2.mar  "figure"
ga-             da-             "verbal prefix"
-gin7           di.im3          "like"
g~a2-e          ma(-e)          "I"
g~al2            "to be"
g~ar, g~a2.g~a2 mar,      "to place"          "rite"
G~a2.tum3.dug3  Ma.ze2.eb.zib   "the goddess Gatumdug"
g~eshtin        mu.tin          "wine"
g~eshtug2       mu.ush.tug2     "ear"
g~eme2          "slave girl"
g~ish           mu(.ush)        "tree"
g~idru        "scepter"
g~ir2            "dagger"
g~iri3          me.ri           "foot"
ha-             da-             "verbal prefix"
ha.lam     "to destroy"
i3              u5              "grease"
igi             i.bi2           "eye"
inim         "word"
kalam       "land"
lu2              "man"
mer              "anger"
munus           nu.nus          "woman"
nam            "-hood ?"
nag~a            "soap"  "tortoise"
nig~ir        "herald"
nin             shen, "lady"
nir             "prince ?"
nirah           she.ra.ah       "the snake-god Nirah"
nir.g~al2    "prince"
nu.gig          mu.gi4.ib       "hierodule"
nundum    "lip"
sag~            she.en          "head"
sipa            "shepherd"
sig4            she.eb          "brick"
si.g~ar         si.mar          "bolt"
sum           "to give"
sha3.g          sha3.ab         "heart"
tud2            ze2.ed          "to hit"
udu             e.ze2           "sheep"
unu3, utul      mu.nu10, nu12   "shepherd"
Not counting the words that are apparently based on different roots, we get the
following "phonological rules".  Note that Thomsen is trying to show the
_differences_ between Emesal and Emegir, so that there is probably a bias
against "zero alternations" (no change).
REF# RULE     (# of cases) EXAMPLES
B0   b > b    (1)          nigbunna > shenbunna
P0   p > p    (0)          --
P1   p > b    (1)          sipa > subu
D0   d > d    (8)          dig~ir > dimer
D1   d > z    (4)          dug > zeb, udu > eze
D2   d > -    (0/1)        arad (ir?) > eri
T0   t > t    (3)          tar
T1   t > z    (1)          tud > zed
G0   g > g    (2/3)        nugig > mugib, g~eshtug > mushtug
G1   g > g~   (0/1)        engar > mungar/mung~ar
G2   g > n    (1)          nigbunna > shenbunna
G3   g > d    (3)          agar > adar, -gin > -dim, ga- > da-
G4   g > b    (7)          dug > zeb
K0   k > k    (2)          kalam > kanag~
M0   m > m    (4)          dumu > dumu
M1   m > n    (2)          munus > nunus, g~eme > gin
M2   m > g~   (5)          inim > eneg~
N0   n > n    (10)         inim > eneg~
N1   n > sh   (9)          nin > shen
N2   n > m    (3)          nugig > mugib, nundum > shumdum,-gin > -dim
N3   n > l    (1)          nig~ir > libir
G~0  g~ > g~  (0)          --
G~1  g~ > m   (20)         dig~ir > dimmer
G~2  g~ > n   (1)          sag~ > shen
G~3  g~ > b   (1)          nig~ir > libir
G~4  g~ > g   (1)          g~eme > gin
S0   s > s    (3)          sipa > suba
S1   s > sh   (2)          sag~ > shen, sig > sheb
S2   s > z    (1)          sum > zeg~
SH0  sh > sh  (2/3)        shag > shab
SH1  sh > d/t (1)          dish > did, dita
SH2  sh > -   (1/2)        gish > mu
Z0   z > z    (1)          g~arza > marza
H0   h > h    (1)          Nirah > Sherah
H1   h > g    (1)          halam > geleg~
H2   h > d    (1)          ha- > da-
L0   l > l    (8)          g~al > mal
L1   l > n    (1)          kalam > kanag~
R0   r > r    (19)         dig~ir > dimmer
P1: p/b confusion common in Emegir as well.
D2: dropping final consonant common in Emegir as well.
G1/G2: these look like g/g~ confusions.
G4: occurs sporadically in Emegir as well.
M1: nunus < g~unus?; gin: -m > -n is common in Emegir as well.
N2: nu.gig ( < lu.gig ), Emesal probably has > mugib;  -gin/-gim also
 in Emegir.
N3: n/l confusion common in Emegir as well.
G~2: -m > -n is common in Emegir as well.
G~3/G~4: these look like g/g~ confusions.
SH1: Emegir also has dili, probably just different suffixes.
SH2: dropping final consonant common in Emegir as well.
H1/H2: g-/h- confusion occurs sporadically in Emegir as well (cf. ga-
/ha- prefixes).
L1: l/n confusion common in Emegir as well.
This leaves the following "sound laws", which can be considered as typical of
Emesal dialect:
D1   d > z    (4)           dug > zeb, udu > eze
T1   t > z    (1)           tud > zed
S1   s > sh   (2)           sag~ > shen, sig > sheb
S2   s > z    (1)           sum > zeg~
G3   g > d    (3)           agar > adar, -gin > -dim, ga- > da-
H1   h > g    (1)           halam > geleg~
Possible reflexes of distinct Proto-Sumerian phonemes?:
A: [*w, *ngw ?]
   M2   m > g~   (5)           inim > eneg~
   G~1  g~ > m   (20)          dig~ir > dimmer
B: [*gw ?]
   G4   g > b    (7)           dug > zeb, igi > ibi
C: [*L (Welsh "ll") ?, cf. n ~ l]
   N1   n > sh   (9)           nin > shen, nundum > shumdum
Additionally, there are some interesting hesitations in Emegir (Thomsen,
pp. 44-46) as well:
h ~ r                    rush(u) ~ hush(u) "red"
g ~ b (see G4 above)     buru4 ~ "raven", abrig ~ agrig
h ~ g (see H1, H2 above) ha- ~ ga- "vb. prefix", HA=ku6 "fish"
n ~ l (see L1, N3 above) nu ~ lu2 "man", nu- ~ la- "not", limmu "4" ~
                         nimin "40".
Miguel Carrasquer Vidal
mcv at

More information about the Histling mailing list