May 10, 2011

The case for Euphratic

Gordon Whitaker presents an interesting theory about the presence of a pre-Sumerian Euphratic Indo-European language in Mesopotamia; as always I don't comment on linguistic theories, but I've often thought that the early Indo-Europeans of the Near East may have been swamped by people from their periphery (Arabia, the Caucasus, and Central Asia), so the theory is not that outlandish to my ears. One often thinks of the Indo-Europeans as doing the language replacement, so one often forgets that they were often the victims of language replacement as well (e.g., in large parts of Eurasia to Turkic, and the Near East to Semitic languages).

Sumerian appears seemingly out of thin air in Mesopotamia; that is peculiar if it was the aboriginal language of a highly productive (and populous) Neolithic group. The same can be said about Semitic languages that make their appearance into Mesopotamia soon after in the form of Akkadian. There is also the example of the Kassites, another language isolate, who attacked Mesopotamia (from Iran?) but were not as successful, nor were the Hurrians who spread their influence into northern Mesopotamia. I also like the author's suggestion that the picture of a pristine Sumerian-speaking aboriginal population contradicts all our historical knowledge about the region, where different groups have lived there side by side throughout recorded history.

It is difficult to detect a written substratum in Sumerian, as Sumerians were the first literate civilization in the world, and so we are unlikely to ever find written records of the putative Euphratic language. If such a language did exist, we can only search for it, much as Whitaker has done, as a substratum in the languages that followed it in Mesopotamia.

Bulletin of the Georgian National Academy of Sciences, vol. 2, no. 3, 2008

The Case for Euphratic

Gordon Whittaker
University of Göttingen, Germany

ABSTRACT. It will be argued that the cuneiform writing system, the Sumerian and Akkadian lexicon, and the place names of Southern Mesopotamia preserve traces of an early Indo-European language, indeed the earliest by more than a millennium. Furthermore, this evidence is detailed and consistent enough to reconstruct a number of features of the proposed Indo-European language, Euphratic, and to sketch an outline of Euphratean cultural patterns.

Link (pdf)

15 comments:

Daro said...

The only two Indo-European languages that are still present in the northern fertile crescent are the Armenian and Kurdish languages.

Onur said...

One often thinks of the Indo-Europeans as doing the language replacement, so one often forgets that they were often the victims of language replacement as well (e.g., in large parts of Eurasia to Turkic, and the Near East to Semitic languages).

There aren't many historically known cases of replacement of a Indo-European language by a Semitic language in a given territory, much less in a large territory. Historically, Indo-European languages, when they were present in a part of today's Semitic speaking territories, mostly existed as minority (often elite) languages. This was true for West Asia, and was true for North Africa. Only in Islamic Iberia we can speak of a large scale (by absolute territory size and population) historically known replacement of Indo-European languages by a Semitic language, but, of course, it was reversed and finally annulled following the Reconquista and the concomitant re-Latinization of Iberia.

Onur said...

I've often thought that the early Indo-Europeans of the Near East may have been swamped by people from their periphery (Arabia, the Caucasus, and Central Asia)

Such a thing might have happened during pre-historical times.

Andrew Oh-Willeke said...

It is far more parsimonious to assume, in the absence of evidence to the contrary, that Sumerian was the aboriginal language of Sumeria until it was swamped by the Semitic Akkadian language in the historic era.

There were a host of ancient non-Indo-European, non-Semitic languages in that are attested in the region from Iberia to Pakistan: Basque, Aquitainian, Elamite, Kassite, Hurrian, Hattic, Minoan, Lemnian, Rhaetic, Etruscan, the North Caucasian languages, the South Caucasian languages, and probably Harappan. There is not a single case in which an Indo-European language was attested anywhere by either Sumerian or Egyptian authors (the only literate cultures in existence at the time anywhere int he world) in the pre-Akkadian period.

Some appear of these ancient non-Indo-European, non-Semitic languages appear to be related to each other in language families. Basque and Aquitanian appear to form a language family and the Vasconic substrate associated with those languages may have had a range as expansive of the Atlantic Neolithic area.

Lemnian, Rhaetic, Etruscan are often viewed as a language family that may also have included Minoan.

An Elamite-Harappan connection is often hypothesized, and there is good reason to think that the Sumerian civilization provided the cultural seed for Harappan culture.

Hurrian and the North Caucasian languages show some strong linguistic similarities.

Whether through common genetic roots or areal influences, the Kassites appeared to have closer cultural links to the Sumerians than the Semitic Akkadians did.

The pro-writing and seal system of the Vinca script shows strong similarities to parallel and contemporaneous systems in Sumeria and the Indus River Valley.

The only places in the Near East where there is any evidence that there were ever Indo-European languages spoken are when some areas in the North adopted Hittite at its greatest extent from prior non-Indo-European languages, and immediately before that point Mittani, and in some coastal areas where Mycenean Greek appears to have been spoken in Levantine and Greek coastal areas that are late in time (many centuries after Sumerian had ceased use outside religious ceremonies). Moreover, there are strong internal indications that Mittani has its roots in Northwest India as a Sanskrit derived language from a common Rig Vedic religious heritage prior to its appearance in the Near East, not the other way around.

These non-Indo-European languages show some similarities. For example, they tend to be ergative and agglutinative, something not characteristic of either Semitic or Indo-European languages. It seems much more plausible that a macro-family of non-Indo-European, non-Semitic ergative agglutinative languages with roots in the Fertile Cresent spread in the early Neolithic from the Indus River Valley to Iberia and up the Atlantic Coast, that the time depth is such that is it impossible to demonstrate connection between the lingustic families that make up this macro-family, and that the only living vestiges that remain of these languages are the North Caucasian and South Caucasian and Basque languages and some place names, with Semitic and Indo-European languages having replaced the others in the historical period.

There is not a single instance of pre-Sumerian cuniform. The lexicon of proto-Indo-European does not support it as a Near Eastern culture - you won't for example, find birch or salmon in Iraq or Lebanon. There is also every reason to think that similarities attributed to a hypothetical substrate could have been borrowed into proto-Indo-European rather than the other way around.

We also have the Kurgan hypothesis that has multiple lines of evidence from archaeology, to ancient historic records, to linguistic relationships, to proto-Indo-European lexicon geomarkers to back it up. Any alternative hypothesis has a high bar to clear.

Nirjhar999 said...

Hmm. Some thoughts
1. Mitanni is also a indo-aryan ruled region from 1500B.C. in the area.but a peacock motif(Which is typicall indian) is also present there from 2100B.C.
2.mitanni is surely rigvedic featured as deities like mitra,varuna,indra,nasatya were invoked on the mitanni tabs.
3.unfortunately some supreme indologists consider it as indo-iranian as eka is aika and medha is mazda! I think its B.S. 2 consider it indo-iranian as in later prakit madhyam is mazzham and eka is oikam, for me its impossible that a languages words becomes simple to complicated! Its always the reverse.
4. The area u mentioned had never a single clan rule as you know different peoples like hattians, hurrians, hittites and also the med people took part to the musical chair.

eurologist said...

That's actually a pretty convincing article. I am still not sure who lived there first, or whether out of neighboring, different ethnicities eventually the Sumerians won over, or if they were always the larger group but started employing help from the IE speakers.

An interesting societal aspect is that the first written Sumerian would have presented itself as "secret" writing to Sumerian land owners or city-heads who employed these (evidently bilingual) writers. This could have well been on purpose, to avoid corruption and nepotism when doing inventories, registers, and trade lists etc.

Another aspect is that the writing itself may heave helped the IE speakers to learn Sumerian in the first place.

eurologist said...

Interestingly, the author also states:

Of interest is the fact that this, like many other parallels to Indo-European, betrays a strong lexical affinity to that area of the IE dialect continuum from which the Western (or Northwestern) languages emerged.

Like, at least in some phylogenetic models, Hittite, Tocharian, Armenian, and Greek. Of course, this could at least in part also be a temporal rather than spatial effect: these people lived ~6,000 years ago - long before the emergence of specific Slavic or Indo-Iranian language features.

Nirjhar999 said...

I dont think that IVC was creeded by sumerians as:
1. The urban tech of the IVC was way ahead than of the latter.
2.Skeleton studies by guys like kenneth keneddy have shown conclusive south asian affinities(modern) of the IVC bones from 4600B.C. to 800B.C.
3.The language of the IVC have no trace of any sumerian effect.
4.The only relation between IVC and sumer is trade as melluha(see melluha in wiki) appears on sumerian tabs with some IVC trade objects 2.
5. The creed of IVC is mehrgarh which is also south asian as far as the dental anatomy have shown.

lars said...

mr Andrew:languages like hurric hattic and sumerian show a patern of tongues diffused by intruder tribesmen (in this case from caucasus and perhaps also from south asia for sumerian see sumerian/dravidian connection) since they lack internal diversity and they were attested for short periods then ceased to exist.
On the other hand Semitic south of Taurus and Indo-European north of taurus were both internally very diverse+had a still long lasting presence(for semitic there was such different languages as assyrian/babylonian/akkadian, eblaite, ugaritic, mariic, aramaic, amuric, canaanite etc... and for IE there was phrygian, luwian, palaic, armenian, mitanni, parthian, lydian ...and perhaps also kassic-kassites had indo-european named deities...)
Afrasan and Indo-European languages diffusion look like the result of demic diffusion of farmers+herders
Birch and Salmon as being supposedly pie lexicon does not stand since their initial meaning could have been merely tree and fish.
IE languages were not so northernly, for example old persian was more southern than mesopotamia and luwian was spoken as far south as the city of hamah in south central syria (there are many hyeroglyphic luwian written texts in that syrian region)
There are some dark points in the paper, for example the author states that conjunctions in and ina have no semitic counterparts+ignore semitic and even afrasan cognates of pie conjunction bi=by as well as ignoring that many of the quoted words are wanderworts present also in chinese, afrasan, dravidian etc...(such as the sumerian word gud=cow)
he connects sumerian ku=fish with pie dguh but ignores that proto semitic for fish is digg as well
in a nutshell the author states that there are 54 non sumerian (because polysyllabic) words, 26 of them are known semitic loanwords and thus the remaining 28 should be indo-european when words like "gud" and "dguh" are more problematic than that, please see below
for example semitic for fish is digg (it was used in the proto sinaitic script to refer to the letter d and had its form from a fish)
as for gud
The second borrowing is still more complicated. Gamkrelidze and
Ivanov see PIH *k>oou “bull, cow” as coming from the Sumerian *Nu[d]
(=gud, gu) and possibly the Egyptian ngÅw “longhorned bull” sometimes
shortened to ng or gw. Bomhard and Kerns identify the same root but
claim it as Nostratic on the basis of the Sumerian, and a Dravidian,
parallel. They do not mention the Egyptian forms.56 Gamkrelidze and
Ivanov claim that the sequence of the velar nasal /n/ and a pharyngeal
in Egyptian is comparable to the glottalized labiovelar in Indo-European.
They associate these forms with the “Old Chinese” *el/an(u)/to, on
ade/adi/r/'ad/ad/up to, to.
b/m/bE/bi/be/by, at
-/itti/hn/et/eti/yet, with.

lars said...

mr Andrew:languages like hurric hattic and sumerian show a patern of tongues diffused by intruder tribesmen (in this case from caucasus and perhaps also from south asia for sumerian see sumerian/dravidian connection) since they lack internal diversity and they were attested for short periods then ceased to exist.
On the other hand Semitic south of Taurus and Indo-European north of taurus were both internally very diverse+had a still long lasting presence(for semitic there was such different languages as assyrian/babylonian/akkadian, eblaite, ugaritic, mariic, aramaic, amuric, canaanite etc... and for IE there was phrygian, luwian, palaic, armenian, mitanni, parthian, lydian ...and perhaps also kassic-kassites had indo-european named deities...)
Afrasan and Indo-European languages diffusion look like the result of demic diffusion of farmers+herders
Birch and Salmon as being supposedly pie lexicon does not stand since their initial meaning could have been merely tree and fish.
IE languages were not so northernly, for example old persian was more southern than mesopotamia and luwian was spoken as far south as the city of hamah in south central syria (there are many hyeroglyphic luwian written texts in that syrian region)
There are some dark points in the paper, for example the author states that conjunctions in and ina have no semitic counterparts+ignore semitic and even afrasan cognates of pie conjunction bi=by as well as ignoring that many of the quoted words are wanderworts present also in chinese, afrasan, dravidian etc...(such as the sumerian word gud=cow)
he connects sumerian ku=fish with pie dguh but ignores that proto semitic for fish is digg as well
in a nutshell the author states that there are 54 non sumerian (because polysyllabic) words, 26 of them are known semitic loanwords and thus the remaining 28 should be indo-european when words like "gud" and "dguh" are more problematic than that, please see below
for example semitic for fish is digg (it was used in the proto sinaitic script to refer to the letter d and had its form from a fish)
as for gud
The second borrowing is still more complicated. Gamkrelidze and
Ivanov see PIH *k>oou “bull, cow” as coming from the Sumerian *Nu[d]
(=gud, gu) and possibly the Egyptian ngÅw “longhorned bull” sometimes
shortened to ng or gw. Bomhard and Kerns identify the same root but
claim it as Nostratic on the basis of the Sumerian, and a Dravidian,
parallel. They do not mention the Egyptian forms.56 Gamkrelidze and
Ivanov claim that the sequence of the velar nasal /n/ and a pharyngeal
in Egyptian is comparable to the glottalized labiovelar in Indo-European.
They associate these forms with the “Old Chinese” *el/an(u)/to, on
ade/adi/r/'ad/ad/up to, to.
b/m/bE/bi/be/by, at
-/itti/hn/et/eti/yet, with.

Andrew Oh-Willeke said...

From the paper: "in one text from the mid-3rd millennium, the Ninmešarra of Enheduanna, only 54 polysyllabic lexemes out of a total of some 864 words occur. And if proper nouns and known Semitic loanwords are excluded, these 54 are reduced to a mere 28. Thus, Civil’s contention seems valid. As we shall see, many of the remaining polysyllabic terms in Sumerian betray an Indo-European origin."

Building a case for a pre-Sumerian Indo-European substrate from fewer than 28 polysylabbic words, some with dubious affiliations to Indo-European as opposed to other potential linguistic sources, and a supposition by Oppenheim from 1964, never supported by any evidence in the 51 years that have followed that “It is quite likely that the Sumerians had adapted for their own use an already existing system and technique of writing.", is not exactly a solid case, particularly given the admitted "continuity in the archaeological record" from the Neolithic until the development of Sumerian writing.

The premise that all 28 of the purported words are loan words also relies on the assumption that the usual trend of monosyllable words was univeral. It would hardly be surprising if some class of words (e.g. proper names and names for newly invented agricultural technologies) have more than one syllable because monosyllable options were used up.


Moreover, even if there are 28 non-Semitic loan words in ancient Sumerian, there is not terribly good reason to think that the direction of borrowing was from Indo-European to Sumerian, rather than, for example, from another non-Indo-European language that was an ultimate source for both Indo-European and Sumerian roots. One would expect areal borrowing of some words into languages in multiple linguistic families of the region. As noted by lars, a failure to look for Coptic borrowings is a particular deficiency, since that was a prestige culture that interacted with the Sumerians.

Nirjhar999 said...

Well every language have a common ancestor

lars said...

Orel proposes an Afroasiatic root *gaw(calf bull)attested in Berber, East Chadic, and Omotic also about the word sukur please see below PIE *sekhiir- 'axe, poleaxe' (Lat. secilris 'axe, battleaxe', OCS sekyra 'axe,battleaxe')Semitic forms Akkad. sukurru 'javelin'von Soden 198 1 :III.1 266), Hebr. segor 'axe': Georgiev 1 953,In Indo-European the word is restricted to the Ancient European area; it must have been borrowed from Semitic while the speakers of these dialects were still in the Near East for the word dapar please see below (notice that this stuff is written by the same gamkrelidze who presented whitteker's paper)PIE * t 'aph- ' sacrifice ' : Sem. *- dabh- ' offer as sacrifice; split' , *dibh'sacrifice , sacrificial animal' (see Fronzaroli 1 965 :IV.255): Akkad. zibu ' sacrifice, offering ' , Ugar. dbh ' sacrifice' , mdbh 'altar', Hebr. zebah ' sacrificial animal' , zibeh. ' offer in sacrifice' , mizbeh. ' altar', Syr. debhi ' sacrifice' , Arab.dabah ' offer as sacrifice ' , dibh.- ' offering, sacrifice'That the Indo-European word is a borrowing is shown byits foreign accessive consonant sequence and root vowel *a.also for birch and slamon please see below 1/salmon:A number of dialects have a special term for one species of fish, the salmon:OHG lahs (Ger. Lachs), Lith. lafifa In Tocharian the cognate word Toch. B laks means 'fish' in general. The narrow dialect distribution of the word, and its absence from a number of dialects,demonstrates its late rise in its specialized sense as referring to a species of salmon, a spotted reddish fish found in rivers of the Caspian basin (Salmo trutta caspius Kessl.) and northern Europe (Salmo salar L. The word for ' salmon' ,*lak"hs-, can be traced to the Indo-European root *lak"h- 'red; spotted' : Skt.liiksii 'red lacquer',Pers raxf ' spotted;red and white As the speakers of Indo-European dialect groups migrated into new ecological environments, new words arose to refer to previously unknown animals found in the new territories. As a rule such words were formed from derivatives of inherited roots, the derivatives receiving specialized meanings.The word for ' salmon' in the dialect group just mentioned seems to have arisen in just this fashion.2/birch:The presence of a term for 'birch' in Proto-Indo-European shows that the Proto-Indo-Europeans knew this tree,in the mountainous regions of Asia Minor.
For the conjunctions in,ina and bi please see below
These together with similar Indo-European and Afroasiatic prepositions (discussed below) and the parallel uses of ablaut or vowel change led him to conclude,“while we should continue to enter a plea for caution, it is
becoming more and more likely that the Semitic, Hamitic and Indo-European languages were originally one.
Eblaite/Akkadian/Egyptian/Canaanite/Arabic/PIE/English
in, ina/ina/m/lE/fi 'inda/en/in
ìna/ana/n/r >el/ila/an(u)/to, on
ade/adi/r/'ad/hatta/ad/up to, to.
b/m/bE/bi/bi/be/by, at
'att/itti/hn/et/attat/eti/yet, with.
According to Lipinsky's "Semitic languages" the Hamito-Semitic(ie afrasan)form of the preposition "in" is likely to have been *im/*m-, wich could have nazalized version(ie in "n") as in akkadian and eblaite reflexes.
Also afrasan for "to,on" is there is also "'in or l" a preposition in or n in Egyptian, which covers the semantic range of Semitic l and ana, and most likely derives from the same common Hamito-Semitic preposition, just as Tigrinya ne- is related to Ge'ez la-. All these forms are best explained by a basic 'in/l- preposition, with dialectal variants.
In fact, Egyptian ir might be considered as resulting from in, if a Hamito-Semitic non-geminated liquid n could become r in non-initial position, like in modern Gurage and in Margi, a language of the Chadic family.

lars said...

As for "ad"=to "The prep 'ad is attested in Palaeosyrian as a-di (Tell Beydar) and a-di-ma (Ebla), used both in the local sense "to(wards)"and the temporal sense "until"It's found in Ugaritic ('d),in Heb ('ad, 'âdē),in Phoenician ('d), in Aram ('ad), in South Arabian ('d, 'dy, 'dn),and it survives in Ge'ez as an adverb 'ādi, "still more". Its augmented form 'dky occurs in Lihyānite. There is no doubt that this preposition is related to the Heb substantives 'ad,"lasting future", and 'od, "duration", and probably to the South Arabian verb 'dwly, "to march", and to its Arabic equivalent 'adā ('dw), "to speed".However, the preposition 'ad is not attested in Arabic where its function may be exercised by hattd"For "with""The old preposition *'tt and its variant *wtt derive from a noun used in ancient Egyptian as ist, "belongings",and they are probably related to Arabic 'atta,"to be abundant";'atāt,"furnishings".The preposition is first attested in Palaeosyrian, at Ebla (áš-tu, áš-tù, áš-tùma,áš-tá,áš-ti,/'att-f) and at Tell Beydar(áš-te4, áš-tum)likewise in Old Akkadian(iš-tum, iš-tu, iš-te4, i.e. /'itt-f)which also uses the form itti resulting from the assimilation 'itt- > 'itt-The usual translation of this particle is"from", "since","with""
"while Canaanite languages— Amarna correspondence,Heb,Phoenician—use the form itta-,itt- resulting from the assimilation 'itt->'itt-.An assimilated form of the preposition is probably preserved also in South Ethiopic as tá-,"from", "in", "with", attested in Amharic, in Argobba, and in most Gurage dialects; e.g. Amharic let tâ-qān tágwazdn, "we traveled day and night", lit."night with day(the Semitic day starts at 6 p.m.)
The preposition "with", "and", may appear as 3t in Tuareg, but it is usually voiced into d in the Libyco-Berber dialects.It is used as a postposition itti "to, at, in", in Cushitic; e.g. Oromo inni Gobātti autobusarrā bu'e, "he got off the bus at Goba".The sumerian word for god dingir is polysyllabic too(it is not analyzable as din+gir)that word has been connected to yenisseian/ket denger=god,high(from there it was borrowed by turkic-whose verbal roots are monosyllabic too-)denger is a loan from an IE language(perhaps scythian;indic druga name of hindu deity)connected with ie reflexes like kurdish drej=long, english long etc.all from pie dhlugo=long The Sumerian case is similar to Turkic in that while in both languages the basic lexical stock is native sumerian and turkic, cultural words on the other hand seem to be mostly afrasan and indo-european so it looks as the sumerians were the result of south asian warrior migrants(they appear beardless in their depictions in contrast with beardy akkadians and beardy gods so they could be from the south indid racial stock add to that the fact that the most likely assumption is that sumerian is either para-dravidian either para-munda either para tibeto-burmese)[I think the connection dingir/pie dhlugo=high,long is phonetically more consistant than the author connection of dingir with deyos=god]The current scholar communis opinio sees the clearly un-sumerian words as loanwords yet we know that migrating warrior tribes tend to adopt cultural words from the language of the conquered native farmers+city dwellers as well as from neighboring folks and in our case it seems that those involved languages are semitic diffusing from south and west and IE diffusing from north and east as well as perhaps being the native pre sumerian folk and indeed during the first stage of the sumerian logograms the early tablets present no compelling internal evidence concerning the identity of the language indeed besides sumerian some of those logograms are IE and Sem pt3 later

lars said...

+there are other authors who claim a semitic origin+the author contradicts himself;indeed he connects laryngeal h2 with sumerian "u" but some lines later he connects sumerian "u" with the laryngeal h3(when claiming an ie origin of sumerian "u"=ewe from ie "h3ouis")phonetically both words fits within sumerian phonetic paradigms+both have afrasan parallels(while lacking eastern IE ones) see below from the aforementionned paper "19 Egyptian Cw.t “Kleinvieh”|Cushitic: (N)Beja eewu “capricorn” ‡‡ IE*H3ewi- „sheep“.26 Semitic: Syriac yaCCā „avis quadem, pterocles al. coturnix“; Tigre wiCe „sorte de passereau“ || Egyptian CwC “ein Vogel”, Cjw “Graukranich-Jungvogel”‡‡ IE *H2woi-s nom. : *H2wei-s gen. “bird”
Also for his connection of sumerian "igi" with IE "h3oku", first it's very unlikely that a folk borrows its UNIQUE word for "eye" from another language+here too there are afrasan parallels(from the same paper please see below:
15 Semitic: Ugaritic Caq “eyeball”; Hebrew Cāqā id.IE *okʷ- (**H3ekʷ-)“eye”.He connects sumerian "hanbur" with pie "h2andh"; the connection is weak both phonetically and semantically+ie h2and is usually assumed to be an afrasan semitic loanword please see below
PIE *Handh- 'edible plant' (Skt. andha- 'plant from which soma was made',Gk. anthos 'flower', anthinon . e'idar 'vegetable food, lotus', Arm. and 'field ') :Sem. */:tint-(at-) 'wheat, grains'Ugar. /:ttt 'wheat'(cf. Aistleitner 1 963 : 101), Akkad. (Babyl.) utetu 'barley grain', (Assyr.) ututu 'grain crops; grain', Hebr. hitta 'wheat', Aram. hinti 'wheat', Arab. hinta 'wheat', S.Arab. (Soqotri) hinteh, (Mehri) heyt 'wheat'
The Indo-European word is restricted to the GreekArmenian-Aryan dialect area and testifies to contact between this dialect group and the Semitic linguistic world in some part of the Near East. Note that the Semitic emphatic (pharyngealized or glottalized) *t is rendered with the IndoEuropean voiced aspirate *dh. It must be assumed that in a number of Semitic
dialects this sound was already perceived as voiced in ancient times as for ner see below
22 Semitic *naCar- “young man; lad, fellow” ‡‡ IE *H2ner- “man”.
Words like "tukun","larah","zarah"...are assumed to be akkadian loanwords please see the paper below
http://anthrocivitas.net/forum/showthread.php?p=163882
His suggested IE etymologies for some sumerian cities look very speculative and most times it looks like wichever language on earth can provide its own etymology to those toponyms.
It remains that a number of his suggested IE etymologies for some sumerian words are solid ones.
As an example the Whittaker claims in another paper(euphratic adjectives)that sumerian "huluh" is from pie h3olh1-o-wo-s wich means destructive"(notice that phonetically[that word is prununced 3ol'owos 3 stands for laryngeal h3 and ' stands for laryngeal h1]and also semantically [connecting fear with destruction] the paradigm is very lax)while Civil(word no 172)connects it with Akk "galatu" with the same meaning and with another Semitic parallel in Arabic "hawal"=terror;here phonetics is slightly better and the semantic is perfect yet he does not provide an afrasan cognate(this is in part due to the lack_contrary to overstudied indo-european_of available works dealing with proto afrasan stems)however what is clear is that apparently there is no_universal_paradigm that is able to distinguish Sem from IE reflexes(however that is partly true for other languages such as hurric see for example the sumerian word "hizur" claimed hurric loanword by most sumerologists but IE by whittaker)
http://oi.uchicago.edu/pdf/as27.pdf