Skip to main content

Full text of "Extinct Languages"

See other formats


Johannes Friedrich 


Baas-iaw* Center 
SIS ». Bals-wa"'* St. 

lB<Jl**a>t ii»i 



INDIANAPOLIS, IN 46202-5195 

i . 


Copyright, 1957, by Philosophical Library, Inc. 
15 East 40 Street, New York 16, N. Y. 

All rights reserved. 

Translated from the original German EntzfrTeiung 
Verschollener Schrilten undSprachen by Frank Gaynor 



Xb» Library 

iaiiasa Hsivsriity 

Downtown Csnttr 

518 N, Delawar* St, 

t»diartapalis, ladlana 

Printed in the United States of America 


Introduction - vn 

I. The Three Great Decipherments in the Study of the 

Ancient Orient l 

1. The Egyptian Hieroglyphics ..' * 

a. Land' and People, History and Culture 1 

b. The Principles of the Egyptian Writing 5 

c. The Decipherment of the Egyptian Writing ... 16 

d. The Meroitic Script and Its Study 26 

2. Cuneiform Writing 2 9 

a. Land and People, History and Civilization of 

Mesopotamia 2 9 

b. The Essential Features of the Cuneiform Writing 34 

c. The Spread of Cuneiform Writing to the East 

and to the West • ■ • 44 

d. Remarks concerning the History and Civilization 

of the Hurrians and Hittites 45 

e. Alphabetic Scripts Based on the Cuneiform Writ- 

ing 48 

f. The Decipherment of the Early Persian Cunei- 

form Script 5° 

g. The Decipherment of the Neo-Elamite Cunei- 

form Script 59 

h. The Decipherment of the Babylonian Cuneiform 

Script 00 

i. The Interpretation of Sumerian Records 68 

j. The Interpretation of Hittite and of Cognate 

Languages of Asia Minor 69 

k. The Interpretation of Human 79 

1. The Interpretation of Urartaean 81 

m. The Interpretation of Early Elamite 82 

n. The Decipherment of Ugaritic 83 




3. The Hittite Hieroglyphic Writing 

a. General Facts 

b. The Basic Principles of the Hieroglyphic Script 

and the Possibility of a Decipherment . . 

c. The Progress of the Decipherment 

II. The Decipherment and Study of Other Scripts and 
Languages of the Old World 

1. The Decipherment of Other Unknown Scripts and 


a. The Translation of the Lycian Language 

b. The Translation of the Lydian Language . '. 

c. On the Translation of the Language of Side . 

d. The Decipherment of the Numidian Script 

2. The Decipherment of Other Unknown Scripts 

a. The Decipherment of the Cypriote Script 

b. On the Decipherment of the Proto-Byblic Script 

3. The Translation of Other Unknown Languages 

a. On the Translation of Etruscan 

b. On the Translation of Other Languages of An- 

cient Italy ..... 

c. On the Translation of Phrygian 

III. Principles of the Methodology of the Decipherment 

of Extinct Scripts and Languages ...... 

IV. A Few Examples of Undeciphered Scripts 

1. The Sinaitic Script 

2. The Cretan-Minoan Script 

3. The Carian Script 

4. The Indus Valley Script 













2 37 




2 59 






In the history of human knowledge, the transition from the 
18th century a.d. to the 19th is of no less a significance than 
is the turn of the 15th century a.d. into the 16th, which is 
regarded traditionally as the change-over from the Middle 
Ages to the Modern Era. Whereas about 1500 a.d. the dis- 
coveries and the Renaissance were reshaping the knowledge 
and mental attitude of mankind, the era about 1800 a.d.— 
quite apart from the then nascent radical shift in political 
thought— is characterized by a whole series of new and radical 
facts of knowledge, notably in the fields of the physical sci- 
ences and technology, and in connection with the latter in 
the technique of communications as well, which would jus- 
tify the contention that in those fields the Modern Age be- 
gan about 1800 a.d. This change in the natural sciences went 
hand in hand with a parallel change-over in various human- 
istic sciences. That was, for instance, the time when archae- 
ology was given its new look, by Winckelmann, by the re- 
intensified study of original inscriptions, etc., and when the 
first steps were taken toward a true science of linguistics by 
the recognition of an Indo-European linguistic community, 
by a study of Germanic antiquity and by a systematic investi- 
gation and classification of all the recognizable languages of 
the world. 

Another thing that happened about the same time was 
—and this brings me right to the topic of the present book- 
that the human mind began for the first time to look back at 
the races which had existed before the beginnings of Greek 






history, at the races which had shaped the earliest history of 
mankind in the Orient before the Greeks, at their material 
and abstract thinking, and at the residues of that thinking 
preserved m the inscribed monuments which had survived 
from that remote period of antiquity to the modern age To 
the mind of the man of the 18th century, history had still 
begun, as it had for the Christian Middle Ages, with Homer 
and the tales of the Old Testament, and his knowledge of 
ancient tongues was restricted mainly to Latin, Greek, and 
perhaps Hebrew. Although a certain formal familiarity 'with 
the Old Egyptian monuments at least had been salvaged 
from remote antiquity into the modern age, the people of 
the 17th and 18th centuries still gazed with the same won- 
derment as had the Greeks and Romans at the odd pictorial 
characters with which those monuments were covered all 
over. But it never occurred either to the people of late an- 
tiquity or to those of the early Middle Ages to attempt to 
read this pictographic writing and to understand its con- 
tents. The knowledge of that script had been completely 
lost ever since it had ceased to be used. On the other hand, 
by today we have renewed our acquaintance with the Egyp- 
tian hieroglyphics and language, as well as with the cunei- 
form characters, once used in the Near East for writing a 
number of languages, but vanished from use and from the 
knowledge of mankind even earlier than the Egyptian writ- 
ing, and there are also other formerly forgotten scripts and 
languages with which we have become re-acquainted. The 
scientists who contributed to these re-discoveries thus re- 
stored to linguistic science the lost knowledge of a number 
of languages, some of them very ancient, and they laid the 
first foundation on which a historical study of writing at all 
became possible. But above all, they expanded the historical 
horizon significantly toward the past. Whereas the survey- 
able history of mankind had formerly comprised about two 

and a half millennia, it was now expanded to take in at least 
five thousand years. And not only do the political events of 
those long years unfold before our eyes, but so does also the 
material and intellectual culture of those ancient races; their 
homes, their garments, their ways of living, their religious, 
juristic and scientific thinking come to life anew and open 
for us an insight into the development of human life and 
thought from a perspective far wider in space and time. 

The decipherment of these old scripts and languages in 
the 19th and 20th centuries ranks with the most outstand- 
ing achievements of the human mind, and the only reason 
why it does not stand in the limelight of public interest as a 
co-equal of the radical triumphs of physics and technology 
and their related sciences is that it cannot produce the same 
effect on practical daily life which those discoveries can. 
.This inferior evaluation is also the reason why the unlocking 
of the secret of extinct languages and scripts is never de- 
scribed coherently, and it is therefore still hardly known at 
all to the general public. Yet, this subject is deserving of the 
most careful attention of the learned minds, and is absolutely 
worthy of a presentation per se. This is the aim of the pres- u 
ent book. I hope to be able to group the abundant material ,. 
to a certain degree so as to provide a clear and comprehensive 
view, by first discussing at greater length the outstanding, 
and to a certain extent classical, decipherments of relics of 
the ancient Orient, that of the Egyptian hieroglyphics, of 
the many branches of the cuneiform writing, and of the 
Hittite hieroglyphics which remained enigmatic for a long 
time, but are now laid open to study. Next, I shall discuss, 
more briefly and in a looser arrangement, a few other deci- 
pherments of interest, without any attempt at completeness. 
Only then will I consider it proper to set forth a few theoreti- 
cal reflections relative to the decipherment of extinct scripts 
and languages, such as follow readily from the previously ex- 


plained practice. And in conclusion, I shall append the pres- 
entation of a few still undeciphered scripts, and I shall 
attempt to answer the query as to why they still remain un- 




l. The Egyptian Hieroglyphics 
Egypt is the homeland of the mysterious pictographic char- 
acters which even the ancient Greeks contemplated with 
reverent wonderment and called hieroglyphics, "sacred 
iigns," because they suspected that they contained secret 
wisdom of the magician priests of Egypt. With the obelisks 
in Rome also, this notion of a magic significance of the 
hieroglyphics survived among the beliefs of the Occident, 
and also profound minds of modern times permitted them- 
selves to be influenced by it. Without a belief in a certain 
mysterious wisdom hidden within the hieroglyphics, a 
work of art like Mozart's Magic Flute would be inconceiv- 
able. This is why it is fitting that a presentation of the de- 
cipherments be introduced by a discussion of the Egyptian 
hieroglyphics. For the sake of clarity, also a brief geographic 
and historical survey will be useful. 

(a ) Land and People, History and Culture 
The cultural situation on African soil is rather simple; in 
the ancient days there was only one known civilized race 
there, the Egyptians whose mighty edifices and the picto- 
graphic writing on them still fill the modern visitor with no 
less amazement than they inspired in the ancient Greeks. 

Even in remote antiquity, Egypt was known as a gift of 
the Nile. Only the Nile Valley, about 500 miles long but 
only a few miles in width, is arable land, but extremely fei- 



tile at that, nurtured by the floods of the Nile, and flanked by 
barren desert on both sides. The Egyptians seem to have been 
a race of mixed blood, of African and Semitic-Asiatic extrac- 
tion; their language was a remote kin of the Semitic tongues. 
They considered themselves the original inhabitants of the 
land, and actually no other race can be demonstrated to have 
lived there before them. 

Originally, there must have been two separate kingdoms in 
Egypt: A Northern kingdom (Lower Egypt) in the Delta, 
and a Southern one (Upper Egypt) in the narrow Nile Val- 
ley, extending all the way to Assuan, at the first cataracts. 
King Menes of the Southern Kingdom united both realms 
about 28 50 B.C., and that event marked the beginning of the 
first of the thirty dynasties into which the Greek-Egyptian 
priest Manetho (about 280 b.c.) divided the entire history 
of the Egyptian monarchs up to Alexander. Beginning with 
the 3rd Dynasty, the city of Memphis, on the boundary line 
between the two original kingdoms (in the vicinity of mod- 
ern Cairo), was the capital of the Old Kingdom. The 4th 
Dynasty included the great pyramid builders, Cheops, Chef- 
ren and Mycerinus, and the era of the 5th Dynasty marks the 
beginning of the specific worship of the sun god Re. The 
reign of the 6th and 7th Dynasties (about 2350-2050 B.C.) 
was a period of political weakness. 

The Middle Kingdom introduced a new golden age, be- 
ginning with the 11th Dynasty. The city of Thebes, in the 
south, was the capital in those days. The political heyday of 
this era was represented by the reign of King Sesostris II, 
conqueror of Nubia (1878-1841 b.c), and the cultural high 
point by his son, Amenemhet III (1840-1792 b.c.) . A new 
decline ensued with the invasion of the Hyksos (15-1 6th 
Dynasties— about 1670-1570 b.c), an Asiatic race of bar- 
barians whose chief god is known to us under the Egyptian 
name Sth (Seth) and was a Near-Eastern weather deity. 


The expulsion of the Hyksos by Amosis ( 1 570-1 545 B.C.) 
marks the beginning of the New Kingdom (about 1600- 
715 b.c). Thutmosis I (1524-about 1505 b.c.) and above 
all Thutmosis III (1502-1448 b.c) were great conquerors 
on Asiatic soil. Thutmosis III conquered Palestine, and in a 
battle at Karkhemish, at the bend of the Euphrates, he de- 
feated the Humans, a race powerful in Northern Syria. Thus 
he created an Asiatic province of Egypt, which included 
Palestine and Syria and remained in existence for a long 
time. Also Egypt was unable to escape the influence of the 
highly advanced Syrian civilization; it manifested itself ma- 
terially in the importation of clothes, furniture, etc., and 
culturally in an acquaintance with Semitic deities, such as 
Astarte and Baal, and in the many Semitic words incorpo- 
rated into the Egyptian language. 

The rule of Egypt over Syria did not last forever. Under 
Amenophis III ( 141 3-1 377 B.C.) and Amenophis IV ( 1 377- 
1358 b.c), Syria suffered heavily from the attacks of the 
Habiru, an alien race of nomads, assumed to have been the 
Hebrews. An eloquent picture of this struggle is furnished by 
the correspondence of these two rulers with their Syrian 
vassals and with independent monarchs in Asia. This cor- 
respondence was found in the archives of El Amarna, Egypt, 
residence of Amenophis IV, in 1887, and to the amazement 
of the science of the late 19th century, it was found to have 
been written not in Egyptian, but in Akkadian (Babylonian) , 
on clay tablets, in cuneiform script— because Akkadian was 
the language of general communication in that era. 

The Egyptians soon had a new enemy to fight, the race of 
the Hittites, of Asia Minor, who took the place of the Hur- 
rians in northernmost Syria shortly after 1400 b.c Ramses I 
(1318-1317 b.c), Sethos I (1317-1301 b.c) and notably 
Ramses II (1301-1234 b.c) had to fight bitter battles 
against the Hittites for Syria. Also the battle of Kadesh 



(1296 b.c), hailed by Ramses II in a long epic poem as a 
great Egyptian victory, failed to bring a final decision. Ulti- 
mately, a peace treaty with the Hittite king HattuSili III, 
preserved in an Akkadian version in cuneiform script in the 
Hittite state archives in Bogazkoy, and in an Egyptian ver- 
sion in the temple of Ammon in Thebes, led to a mutual 
recognition of the political status quo. The long reign of 
Ramses II was otherwise another golden age of Egypt. 
•y A new danger threatened the civilization of the ancient 
Orient about 1200 b.c: An invasion of barbaric Indo-Euro- 
pean races from Europe, whom the Egyptians called "Sea 
Peoples." Their first, and most powerful, attack completely 
destroyed the Hittite empire. The Egyptians were able, un- 
der Ramses III (1197-1165 b.c), to defend their own coun- 
try at least, but they irrevocably lost Syria and Palestine 
where indigenous Semitic kingdoms arose then. A political 
decline of Egypt ensued. The rule of the Ethiopian kings 
Sheshonk, Taharka, and others (roth-yth centuries B.C.) 
was followed by a short-lived conquest by Assyria (67C-663 
b.c), another era of independence under the monarchs 
Psammetichus I, Necho and Amasis (663-525 b.c), and 
then came the conquest of the land by the Persians '(525 
b.c) whose place was taken by Alexander the Great in 332 
B.C., and by Rome in 30 b.c 

Unfortunately, it is impossible to present here a more de- 
tailed account of Egyptian civilization; for such information, 
the reader is referred to Agypten und agyptisches Leben im 
Altertum* by ErmamRanke (Tubingen, 1923). There he 
will find more information on the religion of Egypt with its 
richlyJiversiieji pantheon of numerous, partly animal- 
headed, deities (Re, Ammon, Isis, etc.), its strongly developed 
belief in life after death (Osiris, ruler of the dead in the un- 

* Egypt and Egyptian Life in Antiquity. 


derworld in the west), as well as QiLEgyP-riAn_arch.itecture f 
with thepyramids of the Old-Kingdom and the great temples, 
and pillared halls of- the Middle and New Kingdoms, on 
Egyptian science, mathematics, medicine, etc. 

(b) The Principles of the Egyptian Writing 

The script and system of writing of Egypt are the very im- 
portant cultural products which command our closer atten- 
tion at this point. Let it be mentioned in this connection, 
first and above all, that the Egyptians were the first users of a 
sort of paper as writing material; it was manufactured from 
pressed stalks of the papyrus reed, and paper still bears the 
name of that reed even in our modern languages . Their wr it- 
Jng utensil was a kind of brush, made of rushes, which they 
dipped into black or red ink. The direction of writing was 
not fixed; it seldom ran in the direction to which we are ac- 
customed (it is merely our custom to print Egyptian texts 
mostly from left to right, for our own convenience) , but in 
most cases horizontally from right to left, although often also 
from top to bottom, in which case, too, the vertical columns 
followed from right to left. It is to be noted that all the pic- 
tures of human beings and animals face toward the begin- 
ning of the line, also the feet walk in that direction, and the 
hands are stretched out so as to point that way. 

About the written characters of the Egyptian script, it 
must be stated first of all, in general, that the pictographic 
script, to which the Greek Clemens Alexandrinus (died 
after 210 A.n.) already referred as hieroglyphics ("sacred 
signs" ) , was chiefly the script used on the monuments, but 
simpler, more cursive forms developed at a very early date 
for writing on papyrus. These simplified forms more or less 
lost their pictorial character and became similar to our let- 
ters. We call this book script hieratic writing (cf. Fig. 1). A 



o d I ' 


}& ^f .2. 

(i) /&.* «./ #,/ ^M (2) tpnn mrh.i st.wjrt.t (5) ps 
swr (4) k.t n.t tm rdj pr hfiw m hhw (;) jnr.t SwJ r rs n bibiw.f (6) n pr.n.jjm 

(1) Another (prescription) for the stomach when 
it is sick: (2) Cumin, goose fat, milk. (3) Cook, drink, 
(4) Another, in order not to permit that a snake come 
out of the hole: (5) A dry fish laid on the opening of 
its hole, (6) (then) it will not come out. 

Fig. 1. Hieratic writing of the Papyrus Ebers 
with transposition into hieroglyphics. (From Er- 
man, Die Hieioglyphen, pp. 37 and 76,) 

further cursivification of the hieratic characters in the early 
part of the first millennium B.C. resulted in the development 



of the demotic script which resembles our shorthand and is 
very difficult to read (Fig. 2) . These types of writing will be 
mentioned again on p. 16. All the three varieties of script are 
alike as to their inner principle, and therefore it will be suffi- 
cient to base the following discussion of the inner structure 
of Egyptian writing on the plastic hieroglyphic forms, made 
particularly impressive by their pictographic appearance. One 
more thing that I wish, however, to emphasize is that the 
following analysis of the Egyptian writing is a modern con- 
struction and has no footing in Egyptian antiquity. 

Fig, 2. Demotic Writing. (From Jen- 
sen, Die Schrifr, Fig. 46.) 



The Egyptian writing was not an alphabetic script, such 
as the one which we are accustomed by daily use to regard as 
something of a matter of course and naturally given. It con- 
sists of three distinct kinds of signs which impress us partly 
as strange, viz.: woid-signs, phonetic signs, and determina- 
tives. The word-signs, or ideograms, are signs which repre- 
sent the concept of the living being or inanimate object il- 
lustrated and otherwise perceptible by the physical senses, 
i.e., concrete, and they represent that concept regardless of 
the spoken form. The Chinese ideographic script is com- 
posed almost exclusively of such signs, and they abound in 
the Egyptian script, too. (A few examples are shown in Fig. 
3.) A script composed solely of pictographic word-signs 

h \ * 

Soldier eye giraffe horn swallow 


beetle flower sun 

mountain corner 


^ 9 

Bute sandal 




Fig. 3 . Egyptian word-signs for living beings and 
concrete objects (according to Jensen). 

would be understood, if need be, also by a person ignorant of 
the language, because it would represent only the underlying 
concepts behind the words, and not their spoken sound. Be- 
sides objects and beings directly perceptible by the physical 
senses, there are also sensorily perceptible acts, i.e., concepts 
expressed by verbs. It was possible to use simple word-signs, 


without phonetic references, for such concepts, too (Fig. 4) . 
Furthermore, sensorily not perceptible concepts and actions, 

i X 

to beat 

to fly 

to cat to go 

DO Nf * <PJT 

to fight 

to walk 

to weep 

Fig. 4. Egyptian word-signs for sen- 
sorily perceptible actions {according to 

i.e., nouns and verbs, could be expressed by some descriptive 
picture (Fig. 5). In order to write "south," they drew the 
picture of a lily, the flower characteristic of Upper Egypt; 

f I 

"* ft 15 

to rule 

to lead South to find old age 


Fig. 5. Pictograms of sensorily not perceptible concepts and ac- 
tions (according to Jensen and Erraan). 

"age" was indicated by a humped old man leaning on a staff, 
"cool" by a vessel with water running out of it, "to rule" by 
a sceptre, "to lead" by a commander's baton, "to go" by a 
pair of walking feet, "to fly" by a bird with its wings spread 
wide, "to eat" or "to speak" by a man who raises a hand to his 
mouth, etc. In these cases also, the pictogram expressed only 
the concept behind the word, but not its spoken sound. 



But that spoken sound was frequently a very important 
factor. For this reason, it occurred to the Egyptians very 
early, probably way back in the initial stage of the develop- 
ment of the art of writing, that a concept difficult to repre- 
sent pictorially could be symbolized by the picture of some- 
thing phonetically quite similar, but conceptually unrelated. 
This was as if we wanted to represent, say, the concept under- 
lying the verb beat by the picture of a bit, or the concept bad 
by the picture of a bed, or hate by drawing a hat. I have 
chosen these examples deliberately so as to make each ex- 
ample contain two words which are not completely identical, 
because the Egyptians also did not strive for any great ac- 
curacy. They drew the picture of a swaiJow (wr) to indicate 
wr, "great"; a house (pr) to indicate prj, "to go out"; a beetle 
(hpr) to indicate hpr, "to become"; a hond (mi) to indicate 
m&j, "to give biith"; a basket, dr, to indicate dr, "boundary"; 
etc. (Fig. 6a.) In fact, they could express midr, "ear," either 
by the actual picture of an ear alone, or by the juxtaposition 
of the signs for mi, "hond," and dr, "basket," neither of 
which has conceptually anything whatsoever in common 
with an ear. (Cf. Fig. 6b.) 

b > I. $ 2. ffi ^ 

a) wr, "swallow" and wr, "great"; hpr, 
"beetle," and hpr, "to become"; mi, "frond," 
and rasf, "to give birth"; dr, "basket," and 
(h, "boundary, 

b) midr, "eat": 1, Pictogram. 2, mS, "frond" 
dr, "basket," 

Fig. 6. Phonetic substitution bv words of similar sound 
(according to firman). 

As indicated by my vowelless rendition of the Egyptian 
words, the Egyptians apparently attached little importance to 
vowels. At any rate, they did not represent them in their 
script. This fact is evident from the words known to us both 


in Old Egyptian and in Coptic, the language of the Chris- 
_tian Egyptians._wJuch~is-- written with the Greek alphabet. 
E.g., JS)i>A "end" (pabu in Coptic), stands also for ph, 
"to attain" (poll in Coptic), and phtj, "fame" (pahrc in 

Phonetically considered, the signs used as phonetic sym- 
bols of words or parts of words are very different in size. Thus, 
hpr, "to become," and spr, "to attain," contain three con- 
sonants each, whereas wi, "great," and ph., "to attain," con- 
tain only two. (See also Fig. 7 for several especially frequent 
bi-consonantal signs.) In most instances, we are ignorant of 
the exact number and positions of the vowels to be interpo- 
lated between the consonants, but in any case, some of these 
phonetic compounds seem to comprise only one syllable, 


# 7 wt fe^t hi 

r"*"""! mts "^ hn W« \^ mr fJW, 

Fig. 7. Bi-consonantal phonetic signs. (According to Ernian.) 

others seem to consist of several syllables. It is a fact of par- 
ticular importance that a few especially short words con- 
tained only one consonant (cf. Fig. 8). Since we are not 
aware of the vowels to be pronounced with these consonants, 
these signs impress us as characters of a purely consonantal 
alphabet. But they were by no means necessarily such to an 
Egyptian, because the consonant, which has been drilled into 
our consciousness by school education until we have come to 
regard it as something naturally given, is by no means re- 
garded as such by primitive man. It was only in the Greek- 
Roman era that, through an acquaintance with European 


£1 C 

\ ' 
J * 








S £ 

Fig. 8. List of the monoconsonantal signs of the E m 
tian writing, the so-called "alphabet." (According to "fir- 

scripts, a sort of alphabet was developed which served pri- 
marily for writing Greek and Roman names, and therefore 
also took into consideration the vowels, to a certain extent. 
1 But the ancient Egyptians never thought of dissecting 

'"' their words into syllables, let alone into letters, and of even- 
tually abandoning their word-signs and cumbersome syllabic 
symbols in order to write with pure monoconsonantal char- 
acters. Such a reform would have had also the result that all 
those words which differed only in the vowel within them 
would have been represented by the same written form, and 
ambiguities would have occurred in the written texts, (Just 
think of writing the English words stake, stick, stock and 
stuck by the consonants stfc alone, the words wand, wind 
and wound simply by wnd, the words scare, score, scour and 
scurry simply by serf) A strong strain of conservatism was 
another factor that kept the Egyptians from discontinuing 
the use of the pictogram of a word once it had been adopted. 

v In fact, they added phonetic signs also to word-signs which 
expressed words sufficiently by themselves. Thus, §dm, "to 
hear," was expressed with sufficient clarity by 4 , the picture 
of an ear, but they nevertheless added the picture of an owl, 
denoting m, to it, making iboV The word wr, "great," was 

I represented clearly by **., the picture of a swallow, and yet 
the Egyptians liked to append the picture of a mouth,«= , rep- 
resenting an r, i.e.: ^. This wa s how the many pleonastic 
written symbols of the Egyptians came into existence. They 
never did get rid of this complicated mixture of different 
signs, in fact there were times when they just could not pile 
up enough pleonastic symbols to satisfy them. 
But even this complicated mixed method of writing was 
still not enough to make the vowelless Egyptian script un- 
ambiguous. How were the Egyptians to distinguish, e.g., jb, 
"kid," from jbj, "to be thirsty," when both words were 
written solely by the consonant signs / and b? In order to 
solve this problem, they resorted to the expedient of ap- 
pending the picture of a kid when they wanted to write jb, 
and the picture of a man raising his hand to his mouth when 
they intended to denote jbj. (Fig. 9a.} The sign of a house 
could mean both pr, "house," and prj, "to go out"; when a 
pair of walking feet was added, it was expressed clearly that 
pr;, "to go out," was meant. (Fig. 9b.) 

The appending of these unpronounced signs to the word- 
signs provided a convenient means of graphically distin- 

•>U% Ui -> 


Fig. 9. Different concepts pronounced alike: 

a) jh, "kid," and jbj, "to be thirsty." 

b) pr, "house," and pr;, "to go out" 

(According to Erman.) 

guishing words which would have otherwise been written 
.alike. We call these mute explanatory signs determinatives. 
They constitute the previously mentioned third type of 
Egyptian written symbols and play a very important part. 
Such a determinative was appended to most Egyptian word- 
signs, with relatively few exceptions. Several of the most im- 


portent determinatives are shown in Fig. 10. There is the 
drawing of a seated man which was added to names or desig- 
nations of men. You see also the drawing of a woman, which 
appeared after names or designations of women. You see the 










meat, limbs 





fixed but unpronounced determinatives), confusing as it 
\niay appear to the uninitiated at first, did enable the Egyp- 
tians to express their language in writing with satisfactory 
linambiguity. The combined use of the different kinds of 
symbols is illustrated further by the few sentences shown in 
Fig. 11. The failure of the Egyptians to adopt changes in 



light, time 

Son mine. 


descrts, foreign 





to break, 

to divide 






love for you. 

* ^ * & 


avenger mine, 


There protect 


Mi:-hpr-r l 


f SiWfi 

bands mine 

c nh dt 

may be live forever. 


I gtow 

limbs thine 






the protection 

abstracts . 

r\ www 


Fig. 10. Egyptian determinatives (according to Jensen). 

determinatives for mammals (picture of a pelt with tail), 
trees, plants, irrigated land (drawing of a ditch), countries 
( demarcated stretch of land ) , cities (ground-plan of a walled 
city, with two streets crossing each other) , water (three wavy 
lines), houses (ground-plan of a house with a door), meat 
(a piece of meat), time (picture of the disk of the sun), 
stones, lands (three mountains), to go, to see, liquids (ves- 
sel), to cut (knife), to tie (string), actions (beating arm), 
ships, to divide, minerals (grains), fire (coal basin with rope 
or chain for carrying it), and abstracts, i.e., spiritual things 
(book roll). 

Only this system of three different classes of written signs 
(word-signs, phonetic symbols of varying coverage, and suf- 




ndm * wjj 

Hov? sweftt 

A 1 I 


(is) friendliness thine 


5 h$± 


breast mi 


& k *.*z n d m&& 



holiness mine. 

T wonder 


about you. 

I place 


I lay might thine (an: 


1 1 * 









Q I t I 



1 1 


the terror before thee 




the supports 

www ri ^ 



of heaven. 

Fig. 11. Egyptian sentences (according to Jensen). 



this method of writing, which strikes us as so very compli- 
cated, was due not solely to their conservative nature, but 
also to the very concrete motive that their writing would 
have been ambiguous and misleading without these compli- 
cated elements. The same consideration has frustrated all 
attempts in modern China and Japan at replacing the com- 
plicated ideographic and syllabic script by the Latin alpha- 
betic writing. 

The Egyptian script changed in its external form only. We 
have already mentioned the simplification of the picto- 
graphic monumental writing into the hieratic book script, 
and the further cursivification of the latter to the shorthand- 
like demotic form. Fig. i (p. 6) shows a few lines from the 
famous Papyrus Ebers, a hieratic medical book dating from 
1600 B.C., now kept in the library of the Leipzig University; 
below the lines in hieratic script, you see the transcription of 
the signs into their hieroglyphic forms, followed by their 
transliteration in Latin characters and by their translation 
to English. In this connection, note that in the hieratic ver- 
sion the characters run, as in the original, from right to left, 
but the signs in the hieroglyphic transcription are printed in 
the left-to-right sequence to which we are accustomed in our 
own books. The direction of the demotic writing is shown 
in Fig. 2. 

{ c ) The Decipherm en t of the Egyptian Writing 

My readers may have grown impatient while following 
my presentation of our current knowledge of the Egyptian 
writing as seen after its decipherment, and now they are 
probably looking forward to an account of the process of 
the decipherment itself. But I had to approach the subject by- 
such a roundabout way, for otherwise I would have been un- 
able to describe the decipherment with the necessary clarity, 



nd furthermore, now I can afford to be so much briefer 
about many a point. 

People of antiquity did not rack their brains at all about 
a decipherment of the Egyptian script, the appearance of 
which was not unknown to them, because they had no in- 
terest in such things, and because they had the preconceived 
opinion that the hieroglyphics were no writing like all other 
writings, but concealed the secret wisdom of the philosopher 
priests, to be understood only by one likewise initiated into 
xnagico-mystic wisdom. This view was advanced, in late an- 
tiquity, by Horapollon, in his Greek book, HierogJyphica, a 
book that remained unchallenged for centuries. Thus, a 
veritable ban still lay on the hieroglyphics even in the early 
j>art of the modern age, and even their first great decipherer, 
Champollion, was unable to shake off its hold for many 
years. This was also the reason why, in the 17th century, 
^Athanasius Kircher, in his Sphinx mysragogica, could permit 
his imagination to run unchecked and interpret the simple 
phrase j^jj^ ****■> "Osiris says," as "The life of things, 
after the defeat of Typhon, the moisture of Nature, through 
the vigilance of Anubis." Of course, more serious scientists 
rejected such absurdities and considered the secret of the 
hieroglyphics unsolvable. Thus the German-Danish archae- 
ologist Zoega was unable to combat this skepticism in the 
late 18th century, although h_e_regarded the hieroglyphics 
soberly and correctly recognized the fact that the names of 
-the Egyptian monarchs were surrounded by an oval ring, 
which today we call a cartouche. (Cf. Fig. 13.) 

The interest in the civilization and language of Egypt 
then was enhanced unexpectedly by the historical events. 
One of the original aims of Napoleon's daring Egyptian cam- 
paign, parallel with the main political purposes of that under- 
taking, was to conduct an archaeological study of the land. 
Only then did the people of the West learn how many relics 



of Egyptian antiquity still lay preserved under the sands of 
the Egyptian desert. And it so happened just then that the 
soil of Egypt presented scientific research with the best 
means to decipher the hieroglyphic inscriptions which were 
becoming available by the thousands; that means was a bi- 
linguis, i.e., an inscription written in two languages, namely, 
in the unknown Egyptian tongue, and in the well known 
Greek language of Alexander's time. Strictly speaking, it was 
actually a trifr'nguis, a text written in three languages— in 
Early Egyptian, written in hieroglyphics, in Neo-Egyptian 
(demotic), in demotic characters, and in Greek, in Greek 
characters. It was found during the construction of entrench- 
ments at Rosetta j(in the Nile Delta) in 1799- However, the 
inscription was not intact; large sections were missing from 
the portion in Early Egyptian, chiefly at its beginning, and 
so were several words at the beginning of the first lines of the 
demotic portion, and the final words of the last lines of the 
Greek portion. (Cf. Fig. 12.) The Greek text was neverthe- 
less almost completely understandable. It contained a reso- 
lution of the Egyptian priests in honor of the young monarch 
Ptolemy Epiphanes upon his ascension to the throne on 
March 27, 196 b.c. He had done so much good for the tem- 
ples and their priests that the priesthood decided to honor 
him as a god and to erect a statue honoring him in every 
temple, with an inscription identical with this one. 

Naturally, the report of this find aroused tremendous in- 
terest and rekindled the hope that the secret of the hiero- 
glyphics would be solved at last. But it was found soon that 
things were not as simple as all that. The almost completely 
intact demotic portion seemed to be more suitable for the 
undertaking than the seriously damaged hieroglyphic text. 
Although a few frequently recurring names in the Greek 
part could be identified with certain groups of signs recurring 
with equal frequency in the demotic part (thus, in 1802, the 


Fig. 12. The Rosetta Stone. (From Sethe, Vom BiJde zum Buch- 
staben, Table II.) 

Swede Akerhlad determined that the group of demotic char- 
acters reproduced in our Fig. 1 3 was the equivalent of the 



*»mt**~ (SHU 



Fig. 13. The name of Ptolemy in demotic 
and hieroglyphic characters. (From Erman, 
Die Hierogiyphen, p. 7.) 

name "Ptolemy"), this cursive script with its many inter- 
fused symbols (ligatures) contained so many unclarities that 
no real progress was achieved. Nobody even dared tackle 
the seriously damaged hieroglyphic text at first, especially 
since hieroglyphics still were being regarded as secret sym- 

Thomas Young ( 1773-1829),. the.English physicist who 
earned fame by his wave theory of light, was the first man to 
dare attempt the decipherment of the hieroglyphic text, too, 
and he identified the name of Ptolemy also in that portion, 
within the cartouche already discovered by Zoega. (Cf. Fig. 
13.) His reasoning that at least the Greek names of persons 
could not have possessed that character of secret symbols 
which was still ascribed to hieroglyphics in general, led him 
to dissect this group, as shown in Fig. 14. He interpreted 
another inscription as the name of Berenyce (cf. Fig. 15a), 

Young: ftp ^t X) (meaningless; actually jwjfor 0!) 
_£a oh (actually rn> for 11) /— ma (mi) 

Champollion: □ p o / # ] (an) && I (tw) j — ~ m (mi) 

Fig. 14. Analysis of the name "Ptolemy," according to Young 
and Champollion. 

which helped in the identification of a few more letters. In 
an article published in the Encyclopaedia Britannica in 1819, 
he tried to identify also the equivalents of individual Greek 
words in the hieroglyphic portion. But attempts at decipher- 
ment were still hampered by the belief in the symbolic na- 
ture of the hieroglyphics. 

The man who was the first to succeed in actually decipher- 
ing the hieroglyphics was the young Frenchman Jean Fran- 
cois Champollion (1790-1832) who, uncommonly gifted 

^ b (hi) <=> r A k (q) .Sas I (rw) [J t (J) 

""*" * §f\'(j/) -£[o(m) Up J^ «(0 

ine en dins -fait (Feminine (Determinative a: 

^ ending) ^ names of women) 

(Feminine ending -(a)t 
Q fore) 

(Determinative after 
O names of women) 
a) Berenyce 


1 .Jeopa tra 

Fig. 15. The names Berenyce and Cleopatra, and their analysis. 
(According to Jensen.) 

and mature at a very early age and raised as a prodigy, deter- 
mined as a mere boy of eleven to become the decipherer of 
the hieroglyphics. But he too was unable to rid himself, for 
many long years, of the belief in a symbolism of those signs. 
But he prepared himself for his chosen life task by careful 
study. First of all, he learned Coptic, the language of the 
Christian Egyptians which is written in Greek characters, 
and which— as we know today— is quite unsuitable as a bridge 
to an understanding of the ancient tongue both because of 
its quite impoverished vocabulary and its strongly changed 


De»utew B MK 




grammatical structure. Then he obtained reproductions of 
every accessible Egyptian inscription and papyrus, and at a 
cost of fifteen years of extremely tedious work, he compiled 
from them all the forms of the hieroglyphic signs, with their 
graphically simplified hieratic and demotic equivalents— but 
he did all this without yet daring proceed to the reading of 
one single character. 

An important achievement was, in 181 3, his reasoning that 
the ancient Egyptian language, like Coptic, must have ex- 
pressed the personal pronouns in certain cases by the end- 
ings -i for J, -k for thou (masculine), -f for he, -s for she, -n 
for we, -nn for ye, and -u for they. And he was able to ascer- 
tain at least that the hieroglyphic *l=^_ corresponded to the 
Greek words meaning he and him in the Greek portion of the 
Rosetta Stone. His studies of the history of writing enabled 
him then to recognize at once that the Coptic letter cj f (as 
well as some other Coptic characters ) was actually that very 
hieroglyphic *,=>_ /. But he was still burdened with the be- 
lief in the symbolic nature of the hieroglyphics. He still 
thought that in the hieroglyphic version of the name of 
Ptolemy, known since Young, the likeness of the lion, 
_2a (rw = I) , was to be regarded as a symbol of war— 
p ( t ) olemos in Greek— hidden within the name of Ptolemy. 

He did not abandon this erroneous belief until December 
2r, 1821. A simple count proved then that the still existing 
hieroglyphic text of the Rosetta Stone contained about three 
times as many signs as there were words in the Greek text. It 
was therefore quite unthinkable that each hieroglyphic rep- 
resented a whole word, and the hieroglyphic inscription was 
sure to have contained a considerable number of phonetic 
symbols as well. And only after arriving at this conclusion 
did he apply to entire names the methods of research into 
the history of writing, formerly applied to individual sym- 
bols only; he converted the demotic equivalents of the Greek 


names appearing on the Rosetta Stone, character by charac- 
ter, into their hieratic and further into their hieroglyphic 
forms, and, for instance, when applying this procedure to the 
name of "Ptolemy," he actually arrived at the same hiero- 
glyphic symbol group which appeared in the cartouche in 
the hieroglyphic portion of the Rosetta Stone (Fig. 13), So 
now he dared proceed to determine the meaning of each 
individual sign, as shown in Fig. 14. Analogously, going back- 
ward from the demotic form, he reconstructed the name 
"Cleopatra" as the series of signs shown in Fig. 15 (b), and 
the identical way of writing that name was actually found in 
an inscription written in hieroglyphics and in Greek, dis- 
covered in January 1822, so that now he had established tb 
values of a few more signs. 

The taboo was broken now, and Champollion identifi 
other inscriptions as the Greek and Latin names Alexandras 
Autokrator, Tiberius, Domitianus, Germanicus, Traianus 
etc. (Fig. 16}. But he still held the belief that only foreign 
non-Egyptian names could be written this way. Only o 


d)o (2 


:o^c=» Oojjs-a 


Fig. 16. The names Alexandras (a), Autokrator (b), Tfben 
(c), Domitianus (d), Germanicus (e) and Traianus (f) v 
glyphic script (according to Erman) . 

September 14, 1822, when he immediately recogni 
certain new hieroglyphic inscriptions the names of the a 
cient Egyptian rulers Re-mss {R'-m&), i.e., Ramses, an 
Thout-ms (Dhwrj-ms*), i.e., Thutmosis (Fig. 17), did i 
finally become clear to him that the ancient script also ha< 




Fie. 17. The names Ramses (a) and Thutmosis (b) 
in hieroglyphic script. (From Erman, Die Hiero- 
gJyphen,p. n.) 

used no mystic symbols, but phonetic chaiacters^and that 
his laborious research had re-discovered not merely .the way. 
of writing Greek and Roman names in Egyptian characters, 
but the .ancient writing itself. On September 27, 1822, he 
waj able to notify the Academy of Paris, in his famous Lerrre 
3 M. Daeier relative a J'alphabet des hiiwglyphes phone- 
riques-,* that he had succeeded in deciphering the Egyptian 

In the further decipherment, Champollion enjoyed the 
benefit of the enormous preliminary work which he had done 
by comparing the individual signs through so many years. In 
1825, he was already able to translate an Early Egyptian in- 
scription of Amenophis III. Further hieroglyphic inscrip- 
tions and hieratic papyri came next. He was able to learn 
also the correct meaning of the great epic poem glorifying 
the victory of Ramses II over the Hittites in the battle of 
Kadesh (cf. p. 4) . Let it be emphasized, however, in the 
words of Erman/ * that "it was no well organized knowledge 
that he bequeathed to his successors at the time of his prema- 
ture death {1832) . He had ingeniously comprehended prop- 
erly the words and sentences, but he never formulated a clear 
comprehension of the system of the script which he knew 
how to read." A decipherer, who skips over scruples and dif- 

* "Letter to Mr, Daeier, concerning the alphabet of the phonetic hiero- 

** Die Hierogfyphen, p. 12. (Sammlung GSschen, No, 608,) 


ficulties_ ingeniously, and a philologist, who ponders his re- 
sults carefully as he forges them into rules, are fundamentally 
different and must not be mistaken for each other. It should 
be, therefore, no surprise that Champollion's decipherment 
was by no means universally accepted at first. For three more 
decades, even scientists were not willing to admit anything 
more than the fact that at best a few royal names could be 
deciphered, but they insisted that everything else was pure 
fantasy. Only in 1866 did the discovery of another bilingual 
inscription, the lengthy Decree of Canopus, bring scientific 
confirmation of a whole series of facts which Champollion 
had ingeniously discovered. 

The new infant science thus had to mature for several 
decades before it eventually built a scientifically guaranteed 
Egyptian philology on the foundation left by Champollion. 
The details of this development are, at best, of interest to the 
specialist, but certainly not to the general public. For this 
reason, I shall mention only a few more principal facts at this 

The great and versatile German scientist Richard Lepsius 
was the one who ultimately clarified the Egyptian system of 
writing and, by his translation and analysis of the Decree of 
Canopus, silenced the doubts of serious scientists, too, a; 
regards the decipherment. Next to him, mention is due t 
several more careful, sober savants, namely: the Englis 
Birch, the astute Irishman Hincks, whom we shall meet 
again when we shall review the decipherment of the cunei- 
form writing, and the French de Rouge. Birch and Hincks 
are to be credited first of all with the correct explanation of 
the determinatives. The German Heinrich Brugsch was a 
man of a more tempestuous character; he was still an under- 
graduate in 1848, when he cast further light on the intricate 
demotic script which Egyptologists are inclined to shun even 



Among the personages of the subsequent period of philo- 
logical development, a few brief words of praise are due to the 
French Chabas and Maspdro, and to the English Budge and 
Gardiner. The credit for the really solid lexical and grammati- 
cal foundations belongs above all to German research, and 
primarily to the strict "Berlin School" of Adolf Erman 
{1854-1937) whose thorough studies of writing habits en- 
abled him to distinguish the individual linguistic periods and 
to establish the position of the Neo-Egyptian of the late 
second millennium B.C. as an independent language, rank- 
ing side by side with Early Egyptian. He steered lexical re- 
search and, above all, morphology— still hampered by lack of 
clarity as a result of the vowelless script— onto such firm 
tracks as the state of affairs permitted, and he was the man 
who formulated the now universally accepted view of Egyp- 
tian in phonological respects, too. Whereas some others, 
even in the twentieth century, still reckoned with the possi- 
bility that the Egyptian script partly indicated vowels, next 
to the consonants expressed, Erman consistently advocated 
the view of the purely consonantal nature of the Egyptian 
phonetic symbols. His school has remained authoritative in 
all linguistic investigations of Egyptian antiquity. 

( d ) The Meroitic Script and Its Study 

The use of the Egyptian system of writing was limited 
chiefly to Egypt itself, above all because it was so closely 
tailored to the structure of the Egyptian language that it 
would have been difficult to adopt it to another tongue. 
There is, however, a script which can be recognized to be an 
offshoot of the Egyptian, viz., the script of the so-called 
"Ethiopian" empire of Meroe, situated to the south of Egypt, 
from the 1st century b.c. till the 3rd or 4th century a.d. As 
for the shapes of the characters, the Meroitic script looks 
totally like Egyptian writing, and it even developed two varie- 


ties analogously to the Egyptian script, namely: a monu- 
mental script of a pictographic (hieroglyphic) character, the 
symbols of which partly look like those of the Egyptian hiero- 
glyphic script, and a cursive form (although it is found on 
monuments, too) which resembles the demotic writing of 


The striking feature of the Meroitic script is the small 
number of its characters; it consists of a mere 23 characters, 
in contrast to the hundreds of Egyptian symbols. This fact 
can be explained most easily by assuming that whereas the 
Meroitic script emulated the appearance of the Egyptian 
script more or less closely, it was essentially an alphabetic 
script (without word-signs, polyconsonantal signs and de- 
terminatives). One is all the more justified in thinking of 
the prototype of the Greek script because the Meroitic sys- 
tem seems to have used also vowel signs, although not con- 

Meroitic inscriptions have been known since about 1820, 
but they were regarded as undecipherable and untranslatable 
lqr_a long time, until the English Egyptologist Griffith suc- 
ceeded, between 1910 and 1930, in deciphering them to a 
certain extent at least. He started out from the inscription of 
Ben^g^ which is written on the whole in Egyptian, but with 
the names of the king and queen appearing both in Egyptian 
and in Meroitic. Using this inscription as a starting point, a 
more or less certain way of reading the symbols was estab- 
lished, whereas the translation of the unknown language of 
Meroe, not as yet demonstrable to be related to any known 
language, is still a very precarious matter. Fig. 18 shows the 
hieroglyphic and demotic alphabets of the Meroitic lan- 
guage, whereas Fig. 19 presents the reproduction of a demotic 
Meroitic inscription, with its transliteration and translation 
by Griffith. 




| i ante 








































































Fie. 18. The Meroitic alphabets. (From Jensen, Die 
Schiitt, Fig. 49,) 

: Wf/ : a&irtyi : tkti^-mn : iqe : X'krir : erkeU : am nit hey : ts&li 

"Isis (and) Osiris, protect Taktiz-Amon, (the one) begotten (by) 
Zekarer,bom (by) Amon-tares" 

Fig. 19. Meroitic inscription. (Jensen, Die Schiitt, Fig. 50.) 



2. Cuneiform Writing , 

The cuneiform writing of the Near East is far less well 
known to the general public than are the; hieroglyphics 
Egypt- Even the people of antiquity no longer had 1 a^y 
curate notion of it, only very ancient Greece had known 
as Assyria grammata, "Assyrian letters," The people of jthe 
modern era found it more difficult to get used to this harsh 
jumble of wedges, remotely reminiscent of the Chinese 
writing, than to the neat pictographic shapes of the bid 
Egyptian script. But the ancient Orient found the cuneiform 
characters to be by far the more important and practical sys 
tem of writing, and their use spread far beyond the bounda- 
ries of Babylon, their birthplace, and they were employed for 
writing numerous other Near Eastern languages. Let 
however, confine our consideration to the native land of the 
cuneiform script and contemplate Babylonia and Assyria 
somewhat more closely. 

(a) Land and People, History and Civilization 
of Mesopotamia 

Like the civilization in Egypt, the oldest, civilization, in 
the Near East also sprouted forth along rivers, in Mesppjo 
tamia, between the Euphrates and Tigris rivers which in 
those days still flowed separately into the Persian Gulf, 
These two rivers gave the country its name, for Mesopotamia 
is Greek for "the land between the rivers." In Mesopotamia 
however, the arable land was not confined, so completely tc 
the strips along the rivers, although the alluvial land, crisp - 
crossed by the many canals of the Euphrates, does consriutje 
the major part of it. 

The oldest identifiable inhabitants of the land were the 
Sumerians, an ancient civilized race of undetermined origin 
But from the early part of the third millennium B.C. on, the) 
shared the land with the Semitic Babylonians who hai 



migrated, as a race of nomads, from the Syrian desert, as did 
also the Amorites about 2000 B.C., the Aramaeans about 1000 
b.c, and the Arabs in the Christian era. 

The political development of the populace was no less 
disunif orm than its origin. Insofar as we can see, the Sumeri- 
ans seem to have inhabited the land since time immemorial; 
at any rate, the English excavations in Ur and the German 
ones in Uruk, which unearthed relics of very ancient times, 
predating even the development of any writing at all, failed 
to disclose any indication of any older race. For a long time 
there was no unified government in Mesopotamia. Sumeria 
was a land of many small, independent city states to which 
the designation "Sumeria" is applicable in the racial sense 
only. From about 2500 b.c. on, there was a Semitic popula- 
tion next to the Sumerians, in Akkadia, to the north of Su- 
meria, wherefore most recently the designation Akkadians 
has been adopted and used parallelly with the reference to 
"Babylonians," and we speak now also of an Akkadian, i.e. f 
Babylonian-Assyrian language. One of the mightiest rulers 
of early Babylonia was Sargon I (about 2550-2300 b.c.) 
whom later ages regarded as the archetype of great monarchs 
(as did other nations regard Alexander the Great or Charle- 
magne). At the end of the third millennium b.c, the Su- 
merians enjoyed another period of second bloom under 
Gudea of Lagash (about 2050 b.c?) who united most of the 
land for the first time, but the united realm disintegrated 
again into small states later, during the struggle against 
mighty Elam in the east. 

The unification of all of Babylonia was finally accom- 
plished by Hammurabi of Babylon, an Akkadianized Amor- 
ite, the most brilliant monarch in Babylonian history, who 
made decisive the victory of the Semitic race over the Su- 
merians. The exact period when Hammurabi had lived re- 
mained undecided for a long time; the earlier authorities 


hesitated for a long time between 2100 (or even 2300) and 
about 1900 b.c, and only quite recently was it decided with 
finality that Hammurabi had lived from 1728 to 1686 b.c, 
a contemporary of the Assyrian king SamSi-Adad I who is 
definitely known to have lived 1749-1717 b.c From then on 
there was a unified Semitic empire in Mesopotamia, in which 
the Sumerians were gradually absorbed. Hammurabi insti- 
tuted great administrative reforms, recorded in his famous 
Code which was discovered by the French in Susa in 1903. 
Babylon became the capital of the realm known to us from 
that era on as Babylonia. Marduk, the youthful city god of 
Babylon, became the national deity, and the classical Baby- 
lonian language was used as the daily vernacular of the em- 

The empire of Hammurabi did not last long; the alien 
Kassites (or Cossaeans) subdued Babylonia and ruled it for 
a long time. But their rule (about 1600-1200 b.c) was 
gentle and peaceful. In the meantime, Assyria, in the north 
of the land on the Tigris, flourished and prospered ever more 
and more. Assur, a Babylonian city state, is known to have 
existed even in a much earlier era, and it had gone through 
alternate stages of political might and decline. In the 15th 
century b.c, it was a tributary of Mitanni, the mighty Hur- 
rian empire of northwestern Mesopotamia, but after the 
collapse of the latter it became independent and kept gaining 
in importance. Salmanassar I dealt a final death blow to 
Hanigalbat, the successor of Mitanni, about 1270 b.c Tu- 
kulti-Ninurta I laid siege to Babylon about 1230 b.c, and 
Tiglath-Pileser I was the first ruler of the Near East to push 
forward as far as the Mediterranean, about 1100 b.c 

From then on, Assyria was constantly ruled by a drive 
westward which is reflected in the Old Testament, too. Let 
us mention here the following outstanding personages 
among the rulers of the Neo-Assyrian empire: Salmanassar 



III {858-824 b.c.) who fought against Damascus, Tyre and 
Sidon, and also against Yehu of Israel; Tiglath-Pileser III 
(745~7 2 7 B - c wh° supported Ahaz of Judah against Fekah 
of Israel, and whose mastery Hoseah of Israel eventually had 
to acknowledge; Sargon II ( 721-705 b.c. ) , victor over mighty 
Urartu, conqueror of Samaria, who took the ten northern 
tribes of Israel into captivity; Sennacherib (704-681 b.c.) 
who made Nineveh his capital and laid an unsuccessful siege 
to Jerusalem; Assarhaddon (680-669 B - c -) wno even con- 
quered Egypt for a short while and thus brought the Assyrian 
empire to its largest geographical expansion; and finally, 
Assurbanipal (668-626 B.C.), destroyer of the Elamite em- 
pire, but better known for his literary interests and his 
library in Nineveh. 

The Indo-European Medes of Iran joined forces with the 
Babylonians in 606 B.C., and they succeeded in overthrowing 
and destroying the Assyrian empire. That victory made the 
Babylonians arise once again, after the long oppression by 
Assyria, and enjoy a brief second bloom of the Neo-Baby- 
lonian empire, among whose rulers mention is due above all 
to Nebuchadnezzar II (604-562 B.C.), destroyer of Jerusa- 
lem in 586 b.c, who took the Jews into Babylonian captivity. 
The Persian Cyrus defeated the Babylonian army and ended 
the independence of the country in 539 b.c Babylonia re- 
mained a province, first of the Persian empire, then of Alex- 
ander the Great and his successors, later of Rome and still 
later of the Parthian empire, until the Arabs finally made 
Bagdad the successor of Babylon. 

Nor is it possible to say here more than just a few words 
about the civilization of the Sumerians, Babylonians and 
Assyrians. The reader is referred to Meissner's comprehen- 
sive BabyJonien und Assyrien* where he will find detailed 
information on the Sumerian-Babylonian religion with its 

* 2 vols., Heidelberg. 1920, 1925. 


ancient divine triad, Anu (god of the heavens), Enlil (god 
of the air and the earth) and Ea (god of the subterranean 
watery depths) , with the goddess Mah or Ninhursag, "Queen 
of the Gods," as well as the younger triad, Sin (moon god) , 
Shamash (sun god) and Adad (weather god), with the god- 
dess Ishtar. As Babylon was becoming more powerful, also 
its own city god, Marduk, gained an important position in 
the pantheon, as did also Assur, the god of Assyria, when the 
power of the latter was on the ascent. Mesopotamia de- 
veloped a rich mythology and produced a series of epic 
poems; let it suffice to mention here merely the Creation 
Epic, in praise of Marduk, and the great Gilgamesh Epic, 
the deluge episode of which attracted such great attention 
because of the extent to which it agrees with the Biblical 
report of the Flood. Babylonian architecture did not pro- 
duce quite such impressive creations as those of Egypt, be- 
cause it worked with a perishable material, clay brick, but 
the Ishtar Gate of Babylon (reconstructed in the Berlin 
Museum) is nevertheless admirable, and so is the plastic 
art of the Neo-Assyrian period for its artistic perfection and 
true-to-nature details. Among the sciences of Babylonia, 
mention is due above all to mathematics and also to as- 
tronomy, jointly with astrology, which exerted a strong in- 
fluence on the West and gave us also the system of dividing 
time into weeks. The study of the history of Babylonian- 
Assyrian law is an important branch of the study of the 
Ancient Orient, because of the great codes of laws, among 
which the Babylonian Code of Hammurabi and the still 
older Code of King Bilalama of Esnunna (about 1884- 
1863 b.c) as well as the Middle-Assyrian Book of Laws 
are the most outstanding, and also because of the many 
thousands of private documents, originating from various 
regions of the Near East, which illustrate the living applica- 
tion of the system of laws. 



(b) The Essential Features of the Cuneiform Writing 

Of the many cultural achievements of Mesopotamia, its 
script and system of writing are those which are of interest 
to us here. A cursory look seems to reveal a profound differ- 
ence between this script and the Egyptian writing. Egypt 
used paper for writing material, for its literary records at 
least, whereas Babylonia used clay tablets. The characters 
were scratched into the soft clay with a wooden stylus, and 
the tablet was then fired to make it durable. Odd and in- 
convenient as this writing material may appear to us, the use 
of the clay tablet spread out from Babylonia, together with 
the cuneiform writing, and was adopted in remote parts of 
the Near East, in Syria and in Asia Minor; in fact, the clay 
tablet was used also for writing the Cretan-Minoan script 
and language in Crete and in prehistoric Greece. 








$ fea 


Fie. 20. Old Sumetian picto- 
grams and their development into 
cuneiform symbols. (Friedrich, 
Arcftiv Orienta'Ini, Vol. iq, Table 


^so the characters of the Babylonian-Assyrian script ap- 
pear to be basically different from the Egyptian symbols. The 
Egyptian writing originally employed clear, plastic, impres- 
sive pictograms which, however, later became simplified in 
daily use; the Babylonian-Assyrian writing looks like a hodge- 
podge of confusing combinations of wedges which remind 
the uninitiated of the writing of China. But at least this dif- 
ference is a secondary one: cuneiform writing also was in- 
vented by the Sumerians originally as a pictographic script, 
as demonstrated by a few examples in Fig. 20, and only the 
practice of writing them with a wooden stylus in soft clay 
made the individual pictograms degenerate very soon into 
wedge shapes, in which the old pictogTaphic form can be 
detected in a few instances only. There are spatial and 
temporal differences within cuneiform writing as such, too, 
between the intricate Old Babylonian, Old Assyrian (Cap- 
padocian), Middle Babylonian and Middle Assyrian cunei- 
form script, on the one hand, and the simpler Neo-Assyrian 
and Neo-Babylonian varieties on the other (see Figs. 21a, 
21b, and 21c for specimens), but these differences will not 
impress the laymen as strongly as the differences between 
the hieroglyphic and the hieratic and, especially, the de- 
motic scripts. 

On the other hand, the Egyptian script and the cuneiform 
writing are very similar as to their inner nature and principle, 
because the same three kinds of signs which we have observed 
in the Egyptian system of writing are present also in the 
cuneiform script, viz.: word-signs or ideograms, phonetic 
signs, and determinatives. Also in the cuneiform writing 
many words can be expressed by individual word-signs which 
denote only the concept in question, irrespective of the 
spoken sound of the word; the spoken word may sound quite 
different in Sumerian and in Akkadian, in given cases also in 
Hittite, Hurrian, Urartaean or Elamite, but the written sym- 


■ Fl / C \ 2 ^ Ear ty and late forms oi Bab y lonia n cuneiform writ- 
ing, (a) Old Akkadian inscription of King Sarganiiaralim, with 
{£>) transcription into Neo-Assyrian script (opposite paee). 
(Bohl, Akkadian Chiestomatky, I, pp. 40-41 ) 



bol of the concept is identical in all these languages. Thus, 
e.g., *f (originally the picture of a star) was the ideogram 
for the concept "heaven" in every language written in cu- 
neiform characters, and it was pronounced an in Sumerian, 
femu in Akkadian, nepis' in Hittite, etc. But the same sign 
was also the ideogram for the concept "god," and when used 
in that sense, it was pronounced dingir in Sumerian, iJu in 

mm ^ ^ 

iw-ttL m m 



Tffll* «f I 
<HiI ffH *- 

£11 HW >W 
4-ff ►»** 

(ji) W«a a-we4um (32) «awAar //«* (33)^ &fea//«a (34) H-ri-m 
(3 j) a-n-e-Ium Su-ii (36) td-da-ak (yj)& la Su-ur-ga-am (38) i-na ea-ti-iu 
(39) tm-bu-ru (40) id-da-ak 

"If a citizen has stolen a possession of a god or of a temple, such citizen shall be ' 
tilled. Also he who has accepted the stolen goods out of his hands, shall be killed." 

Fig. 21c. Early and late forms of Babylonian cuneiform writing: (c)J 
Early Babylonian script (Art. 6 of Hammurabi's Laws) with transcriptionl 
in Neo-As Syrian writing. 


Akkadian, smni- in Hittite, eni- in Human, etc. The ideo- 
s(arn »«-- meant "king" and was pronounced lugal in Su- 
merian, sarru in Akkadian, haSsu- in Hittite, ivri- in Hurrian, 
ereii- i° Urartaean. 

These words could be represented also by phonetic signs 
or by combinations of an ideographic symbol representing 
the stem of the word and phonetic signs for grammatical end- 
ings (Fig. 22). The noteworthy fact in this connection is 
that unlike the phonetic symbols of the Egyptian script, the 
cuneiform phonetic characters do not represent sometimes 
bigger and sometimes smaller consonant clusters without 
indicating any vowels, but stand for clear-cut syllables, in- 
cluding vowels. These syllables are, to use our classification 

^B-"^^t^- " HKHKT tin 






»f or rftff\? 

S arrant «^,m 


"3^ 3f> jww- txiT< 


Fig. 22. Examples of mixed writing (word stem repre- 
sented by ideograms, endings by phonetic characters) . 

of sounds, either consonant + vowel (e.g., ba, mi, ru) or 
vowel + consonant (e.g., ad, ir, uk) or finally (more rarely) 
consonant + vowel + consonant (e.g., bar, kid, lurn) com- 
binations. Each complex syllable of the last-mentioned type 
can be split into two simple ones, namely one of each of the 

4 o 


other two types, because bar could be written also as ba + ar, 
kid as ki + id, lum as Ju + urn, etc. But let it be emphasized 
in particular that a single, /one consonant can never he ex- 
pressed in cuneiform script. 

The unpronounced determinatives play an especially im- 
portant part in the cuneiform writing. They were mostly 
prefixed to the symbol of the word which they were intended 
to qualify, although in certain rarer cases they were placed 
after it, A specific determinative, consisting of a vertical 
wedge, appears before the names of male persons; another 
one, identical with the ideogram meaning "man," was pre- 
fixed to the designations of professions; a third one, origi- 
nally the likeness of a vulva, appears before the names of 
women and designations of female occupations. The above 
mentioned ideogram for "god" appears as a determinative 
before the names of deities. Other determinatives were used 

d A- 

J E»-lil 


Fig. 23. The names of three gods with determina- 

1. l Ha-am-mu-ra-bi 

2, ! Su-up-p i-Iu -it -si-ma 
j. T Pn-dtt-he-pa 

Fig. 24. Two names of men and a wom- 
an's name, with determinatives 


j, al»Ni-nu-a (nineveh) 

2. mat Mi-i§-ri (ecypt) 

4. al *Kar-ga-mii (karkhemish) 

Fig. 25. The names of "two countries and two cities, 
with determinatives 



Fig. 26. The name of a tree and of a wooden object, 
with determinate 



Fig. 27. The name of a metal and of a metallic ob- 
ject, with determinative 

Defore the names of cities, countries, trees and wooden ob- 
jects, metals and metallic objects, etc. (Examples appear in 
Figs. 25-27.) 

Like the ideograms, the determinatives also were identical 
in every language written with cuneiform characters. Thus, 
if we happen to find a text written in legible cuneiform script 
but in some unknown language, the names of men, women, 
gods, geographic names, etc. immediately reveal themselves 
as what they are because of their determinatives which are 
recognizable in that unknown language as well. Unlike our 

4 2 


scripts, the cuneiform writing thus offers quite a number of 
clues to the reader, and also to the decipherer. 

These facilities of reading or deciphering the cuneiform 
writing notwithstanding, we must not overlook an impor- 
tant difficulty of the reading of cuneiform writing, a diffi- 
culty that is produced by the polyvalence of many a sign. 
This polyvalence may consist in either of the following two 

( 1 ) One and the same sign may be, under certain circum- 
stances, an ideogram or a determinative or a phonetic syllable 
sign. Thus, the above mentioned »f is in the first place a 
word-sign denoting the Sumerian word an, the Akkadian 
word samu, both meaning "heaven," as well the Sumerian 
word dingir and the Akkadian word ilu, both meaning "god"; 
secondly, the same sign is a determinative used with 
names of deities, and thirdly, it is a phonetic symbol, repre- 
senting the syllable an. Further examples are given in Fig. 28. 

1. a) Word-sign (ideogram) for isu (wood) 

b) Determinative used before words denoting trees 
and wooden objects. 

c) Syllabic sign for iz, is, etc. 

2. a) Word-sign (ideogram) for mafu (country) and 

iadd (mountain) 

b) Determinative used before names of countries and 

c) Syllabic sign for tar, mat. Sat, nat, gin, etc. 

Fig. 28. Written symbols which may be used as 
ideograms, determinatives and syllabic signs. 

(2) The second type of polyvalence, which occurs less fre- 
quently in the earlier cuneiform script, but very often in the 
Neo-Assyrian and Neo-Babylonian writing, consists in what 


is termed the polyphony of the cuneiform characters, viz., 
that several phonetically different syllables may be repre- 
sented by one and the same sign, as illustrated by the few 
examples shown in Fig. 29. Thus, the sign meaning bar may 
be read also as mas', the sign meaning ud may stand also for 
tam or par or lah or his", the sign meaning kid may be read also 
as sab or lil, etc. The reader must rely on his knowledge of 
the language and on his familiarity with that particular kind 
of text in order to decide more or less accurately which 
syllabic value is the right one in the given grammatical or 
syntactical situation. 


1 . kid, sab, HI. 

2. pit, gir 

3. lal, lib, tub, pab, nar 

Fig. 29. Characters representing several different syl- 
lables (polyphony). 

The modem reader is likely to wonder also in connection 
with the cuneiform writing why the Babylonians did not dis- 
card their complicated system of partly polyvalent written 
characters and develop at least a simple and clear syllabic 
writing. The fact is that such attempts were made in the late 
Neo-Babylonian era, and it may well be assumed that the 
Semitic system of alphabetic writing, already known then in 
Babylonia, too, was taken in consideration as a pattern. But 
by that time it was too late— the cuneiform writing was just 
about to yield the stage to the more convenient alphabetic 
script. At any rate, in the heyday of the cuneiform writing, 
its users knew no alien script more convenient to handle and 
capable of exerting a stimulating influence in this direc- 
tion, and all the inconveniences notwithstanding, the cunei- 
form script was still relatively the most convenient way of 
writing the Sumerian, Akkadian and other Near Eastern 



languages. The conclusion that it was a fairly convenient way, 
more convenient in any case than the Egyptian writing, is 
demonstrated by the very fact that the cuneiform script, un- 
like the Egyptian system of writing, did not remain con6ned 
to its native land but was adopted by a number of neighbor- 
ing races for the writing of their quite different languages. 
People of the modern era have therefore not been very wrong 
in referring to the cuneiform writing occasionally as the 
"Roman letters of the ancient Orient." I shall now present 
a brief summary of these adoptions of the Babylonian cunei- 
form writing by other races, and then I shall say a few words 
about its historical and cultural significance. 

(c) The Spread of Cuneiform Writing to the East 
and to the West 

The influence exerted by the Babylonian civilization and 
script to the east of Babylonia was slight. In those parts, 
Elam, a state in southwestern Iran, was the only nation in 
contact with the Sumerian, later Babylonian, civilization; 
that contact started in the middle of the third millennium 
B.C., and in ancient times the Elamites borrowed not only 
the cuneiform script but also the Akkadian tongue for record- 
ing their documents. Only later did they decide to write 
texts in their own Elamite language, a tongue neither Indo- 
European nor Semitic, using the Babylonian cuneiform char- 
acters. In the first millennium B.C., when after the Medes 
the Indo-European Persians entered Iran from Armenia, 
Elam was their first cultural center. Thus, the Persians re- 
tained at first the Elamite written language along with the 
Elamite government, and only under Darius did they create 
an "Early Persian" script which was composed of wedge 
shapes but was a not quite pure alphabetic writing (cf. also 
pp. 50 et seq.). Their inscriptions in honor of their rulers 


w ere thereafter composed in three languages: Old Persian, 
Akkadian and Elamite. 

The cultural influence of Babylon toward the west was 
stronger. The Babylonian cuneiform writing was first 
adopted to the west of Babylon, in the early part of the 
second millennium b.c, by the Hurrians of northwestern 
Mesopotamia. The Hurrians borrowed the cuneiform script 
for writing their language, a tongue neither Semitic nor Indo- 
European, and they soon passed it on to the nations of 
ancient Asia Minor which were culturally, in particular 
religiously, strongly dependent on them, and thus first of all 
to the Hittites, the dominant race of Asia Minor, of Indo- 
European origin and speakers of an Indo-European language. 
The Hittites, in turn, adopted the Babylonian cuneiform 
script to write not only their own language, but also the 
languages of their likewise Indo-European neighbors, the 
Luwians and Palaians, as well as the non-Indo-European 
tongue of the Khattians (or Proto-Hattians), the ancient 
race which founded a civilization of its own around the city 
of Khatti (or Hatti), the present-day Bogaz-koy, in the third 
illennium b.c, and later bequeathed it to the conquering 
ndo-European Hittites. Finally, the inhabitants of the land 
I Urartu, in the mountains of Armenia, were young collat- 
eral relatives of the Hurrians. From the 9th to 6th century 
b.c, the Urartaeans wrote their own language with the Neo- 
Assyrian cuneiform characters, in other words in a script 
imported from Assyria in all likelihood in that late period of 
their history only. 

(d ) Remarks concerning the History and Civilization 
of the Hurrians and Hittites 

A few remarks concerning the history of the "Western 
laces" seem to be indicated here. The Hurrians appear to 
have come forth from the Armenian mountains, to migrate 

4 6 


toward northwestern Mesopotamia about 2000 b.c, and al- 
though they were not an Indo-European race themselves, 
they seem to have been led mostly by an Indo-European 
ruling class, of a specifically Indie character. Their conquests 
took them to Syria and Palestine, and probably also far 
into eastern Asia Minor. It is not clear yet whether there was 
any link between the Humans and the empire of the Hyksos 
in Egypt as well. The Humans founded small states all over, 
which disappeared quickly again; only the kingdom of 
Mitanni was temporarily more powerful under King TuS- 
ratta, about 1400 B.C. But Tusratta was defeated by King 
Suppiluliuma of the Hittites about 1575 b.c, and Mitanni 
lost its importance, and soon its political independence, too. 
The ascending state of Assyria conquered the land and 
adopted the traditions of the Hurrians, including their de- 
sire to reach the Mediterranean. 

The Humans played an important role in the history of 
the Near East, for in all probability they were the ones who 
imported the horse from southeastern Europe and popular- 
ized its breeding and use in that part of the Orient. In the 
domain of religion, a strong Human influence on the Hittites 
is unmistakable. They must have been the ones who passed 
the Babylonian cuneiform writing, which they had taken 
over from Babylonia", on to the Hittites, and the latter ob- 
tained from the Hurrians also various intellectual assets, 
Human and Babylonian myths and epics, such as the Gilga- 
mesh Epic, along with the script. 

The Hurrians vanished from the stage of history by 1000 
b.c, at the latest, and only in the land of Urartu, in the 
mountains of Armenia, did the racially cognate race of the 
Urartaeans maintain its position as a rival on equal footing 
of the Neo-Assyrian empire from the 10th to the 7th century 
b.c, to be eventually supplanted by the Indo-European 
Armenians in the 6th century b.c. The Urartaeans, who ex- 


celled in the art of the working of metals, conducted military 
campaigns which took them deep into the Caucasus, too, so 
that inscriptions in the Urartaean tongue, written in the Neo- 
Assyrian cuneiform script, are being discovered not only in 
eastern Anatolia and northwestern Iran, but also in what is 
Soviet territory today. 

In eastern Asia Minor (Cappadocia), the civilization of 
the Khattians (or Proto-Hartfans) , a non-Indo-European 
race, developed around the city of Khatti, or Ham' or Hat- 
tusa, about 2000 b.c, or even earlier. It was taken over by the 
conquering Hittites, an Indo-European race, in the early part 
of the second millennium B.C. The Hittites had migrated to 
Cappadocia from Europe, by some still unknown way, and 
they fused with the Proto-Hattians. The Hittite language, 
an Indo-European tongue, became the dominant vernacular, 
but Proto-Hattian retained its importance as the language of 
the cults of the most important state deities. The Hittites 
were the dominant power of the Near East notably in the 
14th and 1 3th centuries b.c Their king Suppiluliuma (about 
1380-1350 b.c.) destroyed the Human kingdom of Mitanni 
and commanded the respect of Egypt as well. His son, 
MurSili II (about 1 345-131 5 b.c ) , fought bitter wars against 
Arzawa and other kingdoms in Asia Minor to defend the 
hegemony won by his father. Hattus'ili III (about 1282-1250 
b.c ) ended a long war with Egypt by signing the peace treaty 
with Ramses II of Egypt which we mentioned before (p. 4) . 
The Hittite empire in Asia Minor crumbled under the on- 
slaught of the barbarian "Sea Peoples" (cf. p. 4) about 
1200 b.c Only in northern Syria did the so-called "Neo- 
Hittftes" survive until the 8th century b.c, as witnessed by 
their inscriptions in the so-called "Hittite hieroglyphic" 
script (cf. pp. 69 et seq.), notably in the city of Karkhemish 
at the bend of the Euphrates in Syria, but they were even- 
tually absorbed in the expanding Assyrian empire. 



(e) Alphabetic Scripts Based on the Cuneiform Writing 
Aside from the outright borrowings of the Babylonian 
cuneiform writing by races speaking other tongues, we know 
of two more cases of the creation in the ancient Orient of 
new scripts utilizing the wedge shape as their basic element. 
In the city of Ugarit on the northern Syrian coast, today the 
heap of ruins called Ras Shamrah, French excavations, begun 
in 1929, have unearthed the center of a small state inhabited 
by a Western Semitic populace. However, Ugarit, as a harbor 
town, was subject also to alien, Human and even Cretan, 
influences. As for the history of Ugarit, not much more is 
known today than the fact that its king became a vassal of 
the Hittite monarch Suppiluliuma, probably after the defeat 
of Tusratta. In the domain of literature, of great significance 
is the discovery of a number of clay tablets in the library of 
a temple, bearing Ugaritic mythological texts, epic poems in 
the Ugaritic language, honoring their deities; these tablets 
date from approximately the 1 5th or 14th century b.c. These 
texts appear on clay tablets, and also the shape of the script 
resembles the cuneiform writing. But the Ugaritic script is 
an alphabetic system, using only thirty signs (without any 
word-signs or determinatives), and is the oldest specimen of 
alphabetic writing in the Near East (cf. Fig. 30) . The find- 
ings of the most recent years point to the conclusion that the 
Semitic alphabet of individual letter signs was already known 
to the people of Ugarit, in the same sequence in which it is 
known to us from later times and which influenced the con- 
ventional arrangement of the European alphabets. 

The youngest script based on the wedge shape was in- 
vented by the Persians under the great Darius. The Early 
Persian writing, as used in the Old Persion portions of the 
trilingual inscriptions in honor of the ancient Persian rulers, 
has only one feature in common with the Babylonian cunei- 



























































































Fig. 30. The Ugaritic alpha- 
bet. (De Langhe, Les Textes de 
Ras Shamra-Ugarit, vol. I, p. 

form script, viz., the wedge shape which constitutes the 
principal element of the characters. But this is a mere ex- 
ternal resemblance, because the Early Persian script is an 
alphabetic system, although not quite purely alphabetic. It 
consists of thirty-six phonetic symbols, and it still shows a 
few elements of the syllabic system of writing {cf. Fig. 31}. 
A few additionally created ideograms are obviously artificial. 
The Semitic alphabetic writing, known to the Persians in 
its Aramaic variant, was indubitably a factor of great influ- 
ence on the creation of this script. 



( f ) The Dec/pherment of the Early Persian 
Cuneiform Script 

We have followed the development of the cuneiform 
writing from its origin as a Suinerian pictographic script, 
throughout its dissemination over the Near East, until its 
second heyday in ancient Persia. In order to describe the 
progress of the decipherment of the cuneiform script and of 
the reconstruction of the languages written by it, we must 



Sym- lPhoneti< 
bol j Value 


Phonetiol Sym- 
Value | bol 



o, a 




6, 6a 


W before 1 


*. £ 








v, u 




», no 






t before 

it, 1)2 


U, Jlfi 


I. la 


k before 





ns, «W 






d before, 






g 'before 


d before 







A, Ja 






0r, #ra 


i, ia 


p» p& 







Fig. 31. The Early Persian alphabet. {Jensen, Die Schrift, Fig. 69.) 

follow the opposite course and contemplate, first of all, the 
Early Persian script which, being an alphabetic system, is 
more readily susceptible to decipherment, and only then are 
we to proceed to unravel the enigma of the other languages 
in the chronological order of their decipherment. 

I must point out, however, before proceeding any further 
that the knowledge of the cuneiform writing was lost much 
earlier than that of the Egyptian hieroglyphics. Even the 
Greeks had no longer any correct notion of it; only Herodotus 
(IV, 87) mentions it as 'Aa-vOpta ypaniMTa (Assyria gram- 
mata), "Assyrian characters." Thus only in the modern era, 
beginning roughly in the early 17th century a.d., did Europe 
begin to hear about this "nail-shaped" (as it was first called) 
script, through the reports of individual travellers. The first 
report with a short specimen of the script, consisting of five 
groups of symbols, was contained in a letter written by 
Pietro della Valle from the city of Shiras in Persia to a friend 
in Naples in 1621. The first reproduction of a complete 
Early Persian inscription was made public by Chardin in 
1674. The name "Keilschrift" (cuneiform writing) seems to 
have been used first by Engelbert Kampfer in the late 17th 

In the course of the 18th century, other explorers reported 
on several inscriptions honoring kings of ancient Persia, in 
their complete trilingual form, and in 1762, Count Caylus 
even published a report on a quadrilingual (Old Persian, 
Elamite, Babylonian and Egyptian ) alabaster vase of Xerxes, 
but since even the Egyptian writing was still undeciphered 
in those days, a decipherment of the cuneiform characters 
was out of the question. Carsten Niebuhr brought back 
especially reliable copies of trilingual inscriptions from his 
trip to Persepolis in 1765, and he published them in 1788. 
Niebuhr already recognized the three completely different 
systems of writing in the inscriptions, viz., first, a quite simple 



script (Early Persian) consisting of altogether forty-two 
characters (according to Niebuhr's count), secondly, a some- 
what more complicated writing (the Elamite script), and 
thirdly, one especially rich in characters (the Babylonian 
cuneiform writing). 

Following in Niebuhr's footsteps, Olav Gerhard Tychsen, 
the Orientalist from Rostock, made an important discovery 
in 1798: He established the fact that in the first, simplest, 
script a single diagonal wedge served as a sign separating two 
words from each other. Another important accomplishment 
of Tychsen (among several mistaken conclusions) was the 
assumption that the three scripts represented three different 

Another step forward was accomplished by the Danish 
academician Friedrich Miinter, by his Versuch uber die 
keilformigen Jnschriften zu Persepolis,* in 1802. Working 
independently from Tychsen, he also interpreted the single 
diagonal wedge as a separation mark between words, and he 
presented plausible historical arguments to warrant the con- 
clusion that the inscriptions in question originated under the 
Old Persian monarchs of the House of the Achaemenides, 
and that therefore their language should be close to that of 
the Avesta, the sacred book of Persia. He suspected, further- 
more, that the first portions of the inscriptions were in an 
alphabetic script, the second portions in a syllabic script, and 
the third portions consisted of ideographic word-signs. He 
reasoned that all the three portions of any given inscription 
were probably identical in content, for multilingual inscrip- 
tions had been a very common custom in the ancient world, 
and also because whenever a word recurred in the first portion 
of an inscription, a corresponding recurrence of symbols 
could be observed every time in the second and third por- 
tions of the same inscription as well. He assumed, quite 

* An Essay on the Cuneiform Inscriptions of Persepolis. 


correctly, that certain repeatedly recurring groups of symbols 
meant "King" and "King of Kings." He was less lucky in his 
attempt at determining the phonetic values of the characters; 
only by sheer accident did he identify the symbols for a and 
b correctly. 

The man who succeeded in making the Early Persian 
script really legible, beyond such rudimentary findings, how- 
ever, was no trained Orientalist, but a young German high- 
school teacher, Georg Friedrich Grotefend (1775-1853) of 
Gottingen. He was practically ignorant of Oriental lan- 
guages, but he had practiced diligently the decipherment of 
artificially composed secret scripts. His situation was thus 
totally different from that of Champollion: Champollion 
spent fifteen years in painstaking studying and preliminary 
training, after which he succeeded almost despite expecta- 
tions, whereas Grotefend plunged right into the project, 
without any great preliminary training, and without a bi- 
lingual inscription to go by, such as there were available for 
the study of Egyptian, and yet it took him a mere few weeks 
to score quite a considerable success. At any rate, however, 
Grotefend also had to have certain auxiliary data available, 
for no decipherment is feasible without some clues to go 
by. Also Grotefend recognized the separation mark and the 
three types of writing. And also he deduced that the first 
portion of each inscription was written in an alphabetic, not 
a syllabic, script because there often were as many as ten 
symbols between two successive separation marks, and the 
existence of words of ten syllables was an improbability. 

His actual decipherment, presented before the Learned 
Society of Gottingen on September 4, 1802, was based on 
the inscriptions reproduced in our Figures 32 and 33. Like 
Miinter, Grotefend assumed that the inscriptions had been 
composed by Persian kings of the family of the Achaemen- 
ides, and he conjectured also that the first portions of the 





inscription were written in Old Persian, the language of the 
reigning dynasty. Also Grotefend studied the inscriptions 
with the purpose of finding the names of the kings with their 
titles and genealogies, already lcnown from other ancient 
sources, and specifically from sources in Iran itself, viz., from 
the inscriptions of the later Sassanian kings. And like 
Miinter, also Grotefend tried to identify the word recurring 
in Fig. 32 under Nos. 2, 4, 5 and 6, and in Fig. 33 under Nos. 
2, 4, 5 and 7, as the words meaning "king." He interpreted 
the second inscription, along the Sassanian pattern, tenta- 
tively as follows: "X, the King, the great (?), the King of 
Kings, Vs, the King's, son, the Achaemenide (?)—." The 
translation "Y's, the king's, son" was based on the considera- 

n m £T k— g<ft% \ a <<* « m K- Ki "5 K- v 

II HI ~1 'V t~ V» V\ > \M> VV HI TV" TV TT TV- ^ 

K TO if *- m =\ m -m v^ti « ?tt T< r Kg ti k- -v 

«. *9 £ /.-r £ v° /t/ // 1? £; -i.» ?/ « // ti 1^ a 11 /■•-/ 

m \ 9 ft <tT il v°<k «TI m "W =K Ti" << ft T<- v 1 <K 

TIT -N H Ml |T >y\ ^H ITT 1M -^ TI ^\ IT yV \M, 

g ^ff -TiT »W VJ;W if- £T »M gw <t <ff E< <„ << 

(Transliteration) (,) D(a)-a-r(a)-y(a)-v(a)-u-S(a) (2) x(a)-i(a)-a- 
y(a)-$(*)-i-y(a) (3} v(a)- Z (a)-r(a)-k(a) <4> *f«J- 
£(a)-a-y(a)-&(a)-i-y(a) (;) x(a)-l(a)-a-y(a)-&(a)-i-y(<i).a-n(a)~ 
a-m(a) (6) x(a)-i{e)-a-j(<t)4(a)-i-j(a) (7) £(«)$( a) ~j(t()-*- 
n(a)-a-m(a) (8) Vi-!-S(a)-t(a)-a-s(a)-p(a)~b(a)-y(tz)-a (9) >fa}- 
»-/(W (10) H(a)-x(a)-a-m(a)-n(a)-i-s(a)-i-y(a) (u) h(a)-y(a) 
{iz)i-m(a)-mfa) {13) t(a)-£(a)'r(a).-m(a} (14) a-ku-u-n(a)-u-i(a) 

(Pronunciation) DarayavauS xlaya&iya va%rka xlaya&iya x£aya&)ja- 
nam xSdya&iya dahyunam Visiaspahya pufa HaxamaniUya by a imam 
talaram akunaul * 

* x = ch in German ach; s" = sh in English she; j- as in English yes; 8 = 
tft in English the; c =ch in English chin; c as in French ca. 

"Darius, the great king, the king of kings, the king of the lands, Hystaspes' 
son, the Achaemenide, (is the one) who built this palace," 

Fig. 32. Old Persian inscription of Darius. (Messerschmidt, Die 
Entzifferung der KeiJschrift, Fig. 1.) 

tion that the word "Y" of the second inscription appeared 
at the beginning of the first inscription, thus in all probability 
as the name of a ruler, but it appeared in the second inscrip- 
tion after the title "King of Kings" with an ending aug- 
mented by one sign; Grotefend concluded from this circum- 
stance that he was looking at a possessive case linked as such 
with the subsequent word meaning "Son." 

As the next step, Grotefend went through the list of the 
names of the Persian kings, known from Herodotus, and 
checked which names seemed to be most likely to be repre- 
sented by the characters appearing in the inscriptions. Cyrus 
and Cambyses seemed to be out of the question because the 
two names under study did not begin with the same letter, 
and also because they were not of different lengths, but ap- 
proximately equal in length. Finally, an important fact was 
that the father of the author of the second inscription, who 
was the author of the first inscription, also bore the title 
"king," whereas the father of the first inscription did not. All 

i«TT « fj m E| % jjj v 2 «TT 3 m » W U K- *?■** H >l 

m — ■-"'*- .„ — .- *•- _y — , ... *. — - — ■ — WK ~t — ^ 

t= «N0fl « m K- K» if K- * s «rr « to «- r<» njK-m 

:< m -M ^nf?rtr^ *1> <K <w << < «tt « m #■ !<t 

n" K- <-< K- iii ^u <ff ft ^<^ «" m -w tCv << n T^ n 

(Transliteration) X('a)S(a) -y(a) -a-r(a)-l (a) -a (2) x , a) -$(a) -a- 
J(a)-&(a)-i-y(a) (j) v(a)- K (a)-r(a)-k( a ) (4) x(*W*}-*-y(*)~ 
Q(*)-i-j(a) (j) x(a)-$(a)-a-y(a)-&(a)-i-y(a)-s-n(a)-a-m(<i) (6) 
D( e )-a-r(a)-y(a)-v(a)-b(a)-u-S(a) (7) x(*}J(*)-*-j(a)4(*)4- 
j(a)'b(a)-y(a)-a (8) p(a)-u-f(a) (9) H(a)-x(a)-a-m(a)-n(a)-i- 

(Pronunciation) XiayarSa xlaya$iya va^rka xl&yafyiya xSSya&iyanam 
lj&rayavahaul xlaya&iyabya ptifa Haxamaniliya 

"Xerxes, the great king, the king of kings, Darius', the king's, son, the Achae- 

Fie. 53. Old Persian inscription of Xerxes. 
Entziflerung der Keilschrift, Fig. 2.) 






these facts considered together led to the conclusion that 
the author of the second inscription must have been Xerxes 
and the author of the first inscription must have been his 
father, Darius, whose father, Hystaspes, had not been a king. 
The next logical step was to find the Persian forms of these 
three names, which were presumably somewhat different 
from the forms handed down by the Greeks. Giotefend in- 
serted the Avestic forms of the names, and thus he suc- 
ceeded in determining the phonetic values of 15 letters— al- 
though four of those 15 were incorrect because he failed to 
establish the exact forms of the Old Persian variants of the 
names. With the phonetic values thus obtained, he at- 
tempted to read the word meaning "king," arriving at the 
pronunciation khsheh . . . (instead of the correct sound, 
xsayafliya). The reading and translation of the two inscrip- 
tions deciphered by Grotefend is also shown in Figs. 32 
and 33. 

Grotefend thus succeeded within a very short time, with- 
out any bilingual inscription to help him, in laying the 
foundation of the understanding of texts written in charac- 
ters of totally unknown phonetic values. This accomplish- 
ment was made possible, apart from his genius, by the fact 
that he was familiar with the line of Persian kings from other 
sources, as well as by the fact that the Early Persian script 
consisted of only thirty-nine characters and was constructed 
along principles akin to those underlying our alphabetic 
scripts. In the case of a syllabic writing, using a greater num- 
ber of symbols, the range of possible phonetic values would 
have been much wider, and the prospect of finding the right 
values would have been poorer. 

The fact that Grotefend's later activities no longer pro- 
duced as happy results as his first decipherments is to be 
attributed not solely to Iris insufficient training as an Orien- 
talist, but above all to the circumstance, emphasized by 

Grotefend himself, that a decipherer and an interpreter (i.e., 
a philologist) must not be mistaken for each other. His de- 
cipherment should have been developed and elaborated fur- 
ther by trained, professional Orientalists, but those were the 
very people who failed to give him the credit that was his 
due. In fact, the scientific journal of Gottingen did not even 
print the full report of his discovery, but merely a short item 
mentioning it. Only in 1815 was there published a detailed 
report in Heeren's Ideen uber die Politik, den Verkehr und 
den Handel der vornehmsten Vblker der a/ten Welt* This 
is how it came about that his decipherment was given little 
attention and interest at first. 

It was only in 1826 that the Danish professor Rask was 
able to identify the ending of the genitive plural in the phrase 
"king of kings," and only in 1836 could the phonetic values 
of most of the Early Persian characters in one of the inscrip- 
tions be defined by Burnouf, a French student of the Avesta, 
and still more completely by the German Christian Lassen, 
professor of Sanskrit in Bonn, on the ground of a list of peo- 
ples. Of particular importance was the further finding of 
Lassen that in the Early Persian script, similarly to the usage 
of the Indie alphabets, the vowel a was not indicated by any 
special sign (but a was indicated by the addition of a), so 
that, e.g., a p may be read as the consonant p or as a sort of 
syllabic sign for pa. Thus, the ancient Persians wrote xsayfliy 
vzrk for xsayatfiya vazrka ("the great king"), hxamnifty for 
Haxamanis'iya ("the Achaemenid"), daryvus' for Darayavaus' 
("Darius"), etc. 

In the meantime, however, an English investigator went 
to work on the decipherment of the Early Persian writing, 
independently from Grotefend, and favorable circumstances 
enabled him to make substantially greater progress. Henry 

* "Thoughts on the Politics, Communications and Commerce of the Most 
Outstanding Nations of the Ancient World." 



Rawlinson (1810-1895), a British officer, entered the 
Persian government service as a military adviser in 1835; 
and as such he had the opportunity to undertake extensive 
journeys in the country and to study Early Persian inscrip- 
tions right in the field. Of decisive importance was his dis- 
covery of the big Darius inscription on the cliff of Bisutun 
(referred to mostly, incorrectly, as Behistun), by far the 
longest and most substantia] of all Early Persian inscriptions. 
In 1835, he had only the sketchiest idea of Grotef end's de- 
cipherment; he had not even the opportunity to have the 
relevant publications sent to Persia. Thus, as he stated, he 
undertook the entire decipherment anew, independently on 
his own, along similar considerations. He used two other in- 
scriptions, copied by him in his own hand; those two inscrip- 
tions were likewise records left by Darius and Xerxes, so that 
he reached the same conclusion as Grotef end. His knowl- 
edge of the Behistun inscription, however, permitted him to 
gain a better and more profound insight into the Old Persian 
language than Grorefend's. He soon recognized it as a close 
relative of the Avestic and Old Indie (Sanskrit) languages, 
and therefore he was able to use these languages for the 
interpretation of the Old Persian words and grammatical 
forms. In this case, it was possible also to achieve correct, 
incontestable results with the aid of simple homophones in 
the closely related languages, by what is called the etymolog- 
ical method, which is all too often misleading when the re- 
lationship between the languages under consideration is less 
close, as will be discussed later. The publication of the big 
Behistun inscription by Rawlinson in 1846 signified a mile- 
stone along the path of these research esf and studies. Subse- 
quently, the last details of the Early Persian script and the 
Old Persian language as well were cleared up during the 
following decades, by Rawlinson, Hincks (the Irishman 
whom I already mentioned in connection with the Egyptian 


writing, and whom I shall mention again in discussing the 
decipherment of the Babylonian-Assyrian script) and Jules 
Qppert, of Paris. 

(g) The Decipherment of the Neo-EIamite 
Cuneif arm Script 

After the unlocking of the secret of Old Persian, the lan- 
guage written in an alphabetic script and closely related to 
Avestic and Old Indie, the trilingual inscriptions of the 
Achaemenids could be regarded as records written in three 
languages one of which was known, so that the decipher- 
ment of the other two languages appeared to be easy on the 
basis of its knowledge. The second portion of each inscrip- 
tion, evidently written in a syllabic script (the Neo-EJamite 
version, as we know today), was logically the nest to be de- 
ciphered, because even though this script with its 111 differ- 
ent characters seemed to be more complicated than the 
alphabetic Early Persian writing, it was still simpler than the 
writing used in the third portion, with its many hundreds 
of symbols. The absence of separation marks between words 
turned out to be an obstacle in the attempts at decipherment. 
Also when trying to decipher the Neo-EIamite portions of 
the Achaemenide inscriptions, the first step was to identify 
the corresponding names in the Early Persian and Neo- 
Elamite portions, followed by the interpolation of syllabic 
phonetic values in the Neo-EIamite text. The first attempt 
in this direction was made by Grotef end in 1837, and he 
established the fact that names of male persons were identi- 
fied by a vertical wedge (the determinative for the names of 
male persons, as we put it today) placed before them. A 
better progress in this field was feasible only after the publi- 
cation in 1853 of the Elamite text of the big Behistun in- 
scription by Professor Norris of London, for it resulted in the 
increase of the number of proper names from the formerly 





known 40 to 90. The result was that the phonetic values of 
most of the Elamite syllable signs could be determined, and 
on the basis of the Old Persian translation also the individual 
words could be clarified lexically as well as grammatically. 
Nevertheless, during the subsequent decades the study of 
Elamite lagged far behind the study of the other languages 
written in cuneiform characters, and there are still several 
unclarified points in the Elamite grammar. 

(h) The Decipherment of the Babylonian 
Cuneiform Script 
The investigators turned with far greater interest to the 
study of the third and most complicated portions of the 
Achaemenid inscriptions, the Baby/om'an-Assyrian (Aca- 
dian ) parts. It so happened that it had been established in the 
meantime that the same script had been used on monuments 
as well as on clay tablets, many of which became known by 
the end of the 18th century a.d., to be followed by an un- 
ending series of discoveries of more and more such relics in 
the 19th century. Thus there seemed to exist a whole rich 
literature in that language, in sharp contrast to the conditions 
relative to the Early Persian and Neo-EIamite versions. But 
the world could finally look forward to the revelation of the 
most important historical and cultural data about ancient 
Babylonia and Assyria when the French Consul Botta began, 
in 1845, the excavation of the palace of the Assyrian king 
Sargon in Khorsabad, and again in 1845 when the English- 
man Layard began to excavate the ruins of Nineveh, for both 
excavations unearthed a great many monuments inscribed in 
the same third variant of cuneiform writing. 

The Achaemenid inscriptions had to be used as the point 
of departure also in the endeavor to decipher this most im- 
portant language of the entire literature written in cuneiform 
characters, and the first step had to be once again the attempt 

to identify the Akkadian equivalents of the proper names 
occurring in the Old Persian version and then to use them 
for the determination of the phonetic values of syllable signs. 
Of course, that was easier said than done. Not only did the 
writing contain more than 300 different signs, and not only 
was there no separation mark at all, but— as today we know- 
one and the same word was written in one instance by several 
phonetic syllable signs and in another instance by an ideo- 
graphic word-sign, and the system of ideographic representa- 
tion extended even to the writing of proper names. Such a 
peculiar system of writing was obviously bound to discour- 
age the first investigators who had absolutely no knowledge 
of this writing convention. Thus, Rawlinson himself made 
the following admission, in 1850, when he had already been 
able to interpret a longer historical inscription correctly as 
to its essential points: "I will admit freely that when I had 
learned the secret of every single Babylonian symbol and 
every single Babylonian word to which I had found any clue 
at all in the trilingual inscriptions, whether by direct evi- 
dence or through a key, when I tried then to apply the key 
thus gained to the interpretation of the Assyrian inscriptions, 
I was tempted more than once to give up these studies once 
and for all, because I was losing all hope for the achievement 
of any satisfactory result." 

For the better understanding of the reader, I reproduce 
here the original text of an inscription, with its transliteration 
and translation (Fig. 34) ; this text is the Babylonian portion 
of the Xerxes inscription, the Old Persian text of which is 
shown in Fig. 33 on page 55. As in the Old Persian text, the 
equivalent of the word "king" can be recognized to be repre- 
sented by the ideographic word-sign recurring under the 
numbers 2, 4, 5 and 8. Thus, according to the pattern of the 
Old Persian version, the two symbols appearing under No. 3 
must mean "great" (read today as rabu*, i.e., rabu, "great" 



with the phonetic complement "u"), whereas the first six 
symbols in the first line must stand for the name of Xerxes 
(read today as J Hi-H^-ar-Ji ) . The vertical wedge at the be- 
ginning of the first line revealed itself as the determinative 
used before the names of male persons, which Grotefend had 
already recognized as such. The same determinative appears 

ft «3 «tf«— <$p°- <tft**^Y ^v_2t=^ 

3 tf^ fc*ffi= 4 sg> 5 ^ T^H 6U 

7 T g^T 7^ 

PT 7? 

, W'tf'-y-'-"'^ (2) law* (}) r«*/(« (4) <far {;) Sarram'M&i (6) m&r 
(7) 'Da-a-n-ta-a-muS (8) S«m (9) A-ba-ma-an-niSSis 

"Xerxes, the great Icing, the king of kings, the son of Darius the king, the 

Fig. 34. Babylonian inscription of Xerxes. (Meissner, Die Keif- 
schtiit, Fig. 3 . ) 

before No. 7 = "Darius'V'IW-n-^-rf- mtrjand before No. 
9 = "Achaemenide'7 'A-fra-ma -an-nil-K-*) . The word mean- 
ing "son" cannot follow the name of Darius in this version, 
as it did in the Old Persian text, but must precede it (as No. 
6) whereas the symbol preceding No. 6 indicates the plural 
of the word "king" (plural sign, MES) . 

The meaning of each individual word was thus ascertained, 
but still regardless of its pronunciation. For the determina- 
tion of the spoken sounds, it was again necessary to use the 
proper names as the point of departure, for the proper names 
could not be represented by ideographic word-signs, but had 
to be in syllable signs, since the name of Xerxes consisted of 
five characters, and the name of Darius consisted of six, not 
counting the determinative. The determination of the pho- 



netic values of these syllable signs seemed nevertheless 
difficult, for the Babylonian forms 0:: these names might have 
sounded different from the Persian; versions. (In fact, as we 
know now, they did actually sound different.) Another diffi- 
culty consisted in the fact that many of the signs appearing 
on the clay tablets from Babylon were different, often sharply 
different, in shape from the forms used in the Achaemenid 
inscriptions; we have already pointed out the sharp differ- 
ences among the Old, Middle and New, Babylonian and 
Assyrian cuneiform scripts (cf. pagbs 35 et seq.). Thus, the 
investigators of the cuneiform writing faced here a problem 
similar to the one which confronted Champollion as he com- 
pared the hieroglyphic, hieratic and demotic characters. 

It is impossible to mention here more than the most im- 
portant points of the widely ramified detail work, of no 
significance for the general public, involved in the decipher- 
ment. In the 1840's, Grotefend identified the names of 
Darius, Xerxes, Cyrus and Hystaspes in the Babylonian texts, 
and he realized also that a group of symbols appearing on 
bricks found in Babylonia had to represent the name of 
Nebuchadnezzar— only he was unable to read it as yet. (Cf. 
pp. 53etseq.). 

The Swedish Isidor Lowenstern was the first to advocate 
the view (in 1845} that this was a Semitic language. He was 
of the opinion that the phonetic symbols of the Babylonian 
cuneiform writing were simple consonant signs, because— he 
argued— also the later Semitic alphabetic writings (the 
Hebrew writing, the Arabic writing:, etc.) indicated the con- 
sonants only, leaving the vowels unrecorded. In advocating 
this view, however, he made the peculiar observation that 
-for every consonant there had existed several, apparently in- 
discriminately interchangeable signs. Thus, for instance, he 
found no less than seven different signs representing r— 
which actually are the syllable signs ar, k, er, ur, ra, 11 and ru. 


The brilliant Edward Hincks^whose name I have men- 
tioned repeatedly on the preceding pages, was the man who 
recognized that these signs did not represent consonants, but 
syllables. In 1850, he was able to state decisively that the 
Babylonian cuneiform writing contained "not one single 
sign standing for a simple consonant, but signs representing 
a consonant preceded or followed by a vowel." Hincks ascer- 
tained also that in addition to the "simple" syllable signs of 
the ab, ir, etc. and da, ki r etc. type, the script contained also 
symbols for the complex consonant + vowel + consonant 
type, such as kan, raur, etc., and that each such complex sign 
could be split into two simple ones (i.e., ka-an, mu-ur, etc.) , 
for these two methods of writing alternated in frequently 
recurring words (cf. Fig. 55). Hincks is the discoverer also 
of the polyvalence of the Babylonian cuneiform symbols; he 
saw that the same sign could be used as a word-sign, a syllable 
sign or determinative, and he recognized also the determina- 
tives of names of deities, countries and cities, etc. correctly as 

1. a) 


r. a) 1st 

2. a) gir 

3. a) turn 



= b) Is-wr 
— b) gi-ir 
= b) lu-um 

"m <MT«T 

Fig. 35. Alternative compound and simple syl- 
lable symbols. 

Botta, the excavator of the palace of Sargon in Khorsabad, 
is the man to whom the principal credit is due for the recog- 
nition that one and the same word could be represented in 


one instance by one single ideo jra.;>hn word-sign and in 
Another instance by a group of symbols which must be re- 
garded as syllable signs. This conclusion was warranted by the 
fact that among the numerous inscriptions in the palace of 
Sargon there were quite a few which were identical in con- 
tent, and it was a frequent experience to End a group of 
phonetic symbols in one of them at a place where an ideo- 
gram appeared in another. Thus also the spoken words rep- 
resented by ideograms could gradually be determined. 

A final important point of knowledge was discovered by 
Rawlinson who was in the position to work with abundant 
material, and who is fully deserving of the honorary title, 
'Tather of Assyriology," given to him by the British. The 
important point to which I refer here is the peculiarity known 
as phonetic2l polyphony, the fact that the symbol 4J- , mean- 
ing ud, can be read also as tarn, par Mb, minis' (cf. page 43) . 
He was in the position to state in his publication on the 
Babylonian text of theBehistun inscription (1851) that, "It 
can be proven beyond all doubt that the very great majority 
of the Assyrian symbols are polyphonous." The list of 246 
characters which he included in that publication is on the 
whole still valid today and is the basis of our current lists 
of characters. 

The reading of the proper nimes • still remained the 
hardest nut to crack for a long time to come. Thus, the 
name of Nebuchadnezzar, Nabu-iudum-usur ("O /God/ 
Nabu, protect my boundary mark"), appeared written as; SiiJmaini-asarid, Salmanassar, was written as 
Di.MA.Nu.BAit. The difficulty was solved only when vocabu- 
laries of the ancient Babylonian savants were discovered in 
Nineveh, in which such ideographic methods of writing were 
explained. It was seen then that ax,ak was an ideographic 
symbol for the name of the god Wa-bi-ufn, sa.du represented 
the word kudurru (boundary mark ) , f is stood for nasara ( to 



protect), the imperative form of which was usut, di was the 
ideographic symbol for iu/mu (welfare)— thus ni-ma-nu 
stood for Julmanu = gift of welcome— and bar represented 
asaridu (first ) . Generally speaking, without the grammatical, 
lexical and graphic lists prepared by the ancient Babylonians 
and Assyrians themselves on their language, the decipher- 
ment of the Babylonian cuneiform writing would probably 
have been a more laborious task than it actually was. 

At any rate, an adequate basis for the reading and transla- 
tion of the Akkadian language was created about the middle 
of the igth century a.d., and the framework erected merely 
had to be completed by details. Hincks had recognized even 
formerly that two inscriptions, one written in Old Babylo- 
nian script, the other in Neo-Babylonian characters, were du- 
plicates of each other. This discovery enabled him to compare 
a whole series of Old Babylonian signs with their Neo- 
Babylonian equivalents and thus to lay the groundwork for 
a cuneiform palaeography. 

The subsequent research made very rapid headway, and it 
is truly amazing how little time was needed to achieve a 
complete understanding of the texts. Of course, investigators 
less close to the subject still maintained an attitude of scepti- 
cism about polyphony and ideography, not known to them 
from the more familiar systems of writing, and consequently 
they still distrusted the new science. In order to settle the 
question of the reliability of the decipherment, the Royal 
Asiatic Society of London therefore resorted to a special 
expedient: In 1857, Rawlinson, Hincks, Fox Talbot and 
Oppert happened to be in London at the same time. All the 
four scholars were given a copy of a longer text which had 
just been discovered, with the request that they work on it 
independently from each other. Their letters were then 
opened in a formal meeting, and the Society was able to find 
with satisfaction that all the four solutions agreed in all their 


essential points. Thus the young science. of Assyriology could 
now truly be said to stand on a firm foundation. 

Decipherment was no longer a much discussed subject in 
the further course of the research. In the second half of the 
19th century and in the early years of the 20th, careful detail 
work was the main thing and it eventually built Assyriology 
up to a full-fledged philological science. The Semitic charac- 
ter of the language having been firmly established, wide use 
was made also of the phonetically identical or similar words 
of the Hebrew and Arabic languages for the determination 
of the meanings of Akkadian words. And in fact, there were 
many words found which were totally identical in sound and 
in meaning in Akkadian, Hebrew and Arabic, such as, e.g., 
the Hebrew and Akkadian ki (how); the Akkadian and 
Arabic la and the Hebrew Jo (not); the Akkadian biru, the 
Arabic baitu and the Hebrew baj'it or het (house); the 
Akkadian and Arabic kalbu and the Hebrew keleb ( dog) ; the 
Akkadian sarapu and the Hebrew saraf (to burn); the Ak- 
kadian eberu and the Hebrew 'abar (to transgress) ; etc. Only 
occasionally wiD the meanings differ, as e.g., in the case of 
the Akkadian amaru ( to see ) and the Hebrew 'amar ( to say ) . 
Thus, the etymological method, the determination of the 
meanings of words of an unknown language according to 
the meanings of phonetically identical or similar words of a 
known related language, was in most instances successful in 
the domain of the Akkadian tongue. 

The German scholar Friedrich Delitzsch, who founded 
a strongly methodistic school in Leipzig, Breslau and Berlin 
and made Germany the center of cuneiform research, was 
the chief Assyrian philologist about 1900. That was when also 
the Americans began to be interested in the new science, 
and the first representatives of the now highly developed 
American Assyriology were trained then by Delitzsch in 



The 20th century learned, above all, to divide the disuni- 
form structure of the Akkadian language spatially and tem- 
porally into distinct Babylonian and Assyrian dialect groups. 
The fruit of this research, encouraged chiefly by Benno 
Landsberger, is the Grundriss der akkadischen Grammatik* 
a book evidencing amazing diligence and knowledge, by 
Wolfram von Soden, published by the Papal Bible Institute 
in Rome in 1952. A dictionary of modest size, also by Wolf- 
ram von Soden, can be expected to be published within a 
few years. Thus, German science has an honorable share in 
having produced the present broad structure of Assyriology. 

(i) The Interpretation of Sumerian Records 
Akkadian was the last of the languages of cuneiform litera- 
ture the decipherment of which had to begin with the most 
fundamental details. The translation of Sumerian does not 
mean the decipherment of a new script, but merely the in- 
terpretation of another language, for Sumerian was written 
in the same Babylonian cuneiform script as Akkadian. This 
translation of the Sumerian language was accomplished very 
slowly and gradually. In fact, in the first decades of the his- 
tory of Assyriology it was even doubted that Sumerian had 
been a distinct language at all, rather than just a mystic way 
of writing Akkadian. As late as in the last decade of the 19th 
century, young F. H. Weissbach found it necessary to devote 
a book, Die sumerische Frage** (Leipzig, 1898) , to proving 
that Sumerian had been a distinct language. 
^ But there were many difficulties even about the purely 
linguistic understanding of Sumerian. This peculiar lan- 
guage, heretofore not considered definitely related with any 
known tongue, became extinct as a living vernacular soon 
after Hammurabi's time and continued to be used only by 

* An Outline of Akkadian Grammar. 
** The Sumerian Question. 





Orient/ as 
01 iposed \ 
learr t lis 1 

c va ue> 






th e chi$ 






the Babylonian priests as a liturg: 
was learned in the religious seminin 
a "Church Latin of the ancient Oi 
reason, even the Babylonians corippsp 
aids to help the student priests 
they compiled lists of rare phonet 
merian, also grammatical paradig' 
above all, they recorded numerous 
deities and incantations, with thefr 
in Babylonian. These study aids 
modern scholars to a gradual insigl 
difficult and peculiar language, 
them, we would probably still be 
Sumerian tongue. Up to World Wa: I 
gual texts with their often imperii 
merian were the only understood sj ecirr ens 
Only the pioneering translation 
monolingual Old Sumerian royal 
other such records* and Poebel's Sjunjieiid^i 
gradually pave the road to an und;rsta:id 
monolingual texts, too, in the intcrp re 
kenstein is most outstanding at this torne 



in additio 

(j) The Interpretation 

Languages of Asi; 1 

The 20th century brought along 
comprehension of Sumerian, also 
a language once spoken in eastern Asii Nlirio: ' 
Winkler discovered the state archi 
the ruins of Bogazkoy, 94 miles es 
preserved there were written on 

* Die sumerischen und akkadischen Kf)ni^si^t^^jjujt^i ; 
Vorderasiatische Bibliorhek, Vol. 1 , Part 1 
Sumerische Grammatik. Rostock 19:3 

iscri rtipns 

of Hntrifjejnfl 
: Minor 






hires .m 







of Ankara 






of Cognate 

re -discovery 



End therefore 


fete. .For this 


fcktjrkct language; 

occurring in Su- 


tjejxte, hymn,s| to 

' " :■ translations 

guides of, thj 


:ipt been 

Waffled by th 

lat language. 

Ipangih of the 

sf '■■ Gudea and 

Qiarhrnsjir** did 

o; : 'the; ancient 

ofi which Fa 


:en Ifti 


to t: 





1906, Hugo 


the ljecpras 




cuneiform writing, but only a small part of them were in the 
Akkadian language, for their great majority' were in Hittite, 
This find held out the hope for a translation of the Hittite 
language-or rather, of the Hittite language of cuneiform 
litezatuie, because (as mentioned on page 47 and as it will 
be discussed on pages 71 et seq.) there are inscriptions in 
"Hittite" hieroglyphics, too. 

At the time when World War I broke out, Friedrich 
Hrozny, a Czech scholar then teaching in Vienna, was busilv 
preparing copies of the Hittite cuneiform records appearing 
on the clay tablets kept in the museum of Istanbul, com- 
missioned for this work by the Deutsche Orient-Gesellschaff , 
and in the course of this activity he succeeded amazingly 
quickly in gaining an insight also into the language which, to 
his greatest amazement, he found to be of an Indo-European 
structure. Also the Hittite records written in cuneiform 
characters presented solely a problem of linguistic interpre- 
tation, but no problem of deciphering the script, and for this 
reason the word "decipherment"-a term that should be re- 
served for the rediscovery of the lost key to forgotten scripts 
—should be avoided when referring to this language, and 
only the expression "translation," or "interpretation," should 
be used. In other words, the Hittite texts were to the cunei- 
form palaeographers as a Hungarian, Finnish or Turkish text 
written in legible but untranslated Latin script would be to 
most Europeans today, and by no means did they represent 
to them the same enigma which a Chinese or Japanese text 
with its alien script and alien language would represent to a 
European layman. 

How was it possible to regain, within relatively so few 
years, the knowledge of this language, dead and lost for mil- 
lennia? At first, there was no bilingual, Hittite and Akkadian, 
text available; only later were a few such texts discovered, and 
then they were merely instrumental in confirming the al- 


ready established translations of a few words. Higher hopes 
were attached at first to a number of fragments of dictionaries 
also discovered in Bogazkoy, consisting of word lists of the 
type already known from Babylonia, listing Sumerian words 
with their Akkadian equivalents, but completed by the Hit- 
rites by a third column showing the translations in their own 
tongue. These hopes were only partly realized because these 
dictionaries indicate mostly the meanings of rare words 
which occur but seldom in the texts, whereas they are mostly 
useless with respect to every-day vocables especially neces- 
sary for a first idea. Moreover, the dictionaries furnish very 
scant information as to the grammatical structure of the Hit- 
tite language. 

The principal work was thus to be done by combinatory 
research on continuous, contextually coherent Hittite texts. 
The most reliable means for building up an understanding 
of the contents of such records consisted in the peculiar 
method of ideographic writing, quite characteristic of the 
Hittite language. The Babylonians and Assyrians already had 
written their languages partly phonetically, partly with non- 
phonetic ideograms, partly also mixing the two classes of 
symbols, writing the stem of the word by an ideogram and 
the endings phonetically. The Hittites took over this custom, 
and they added also a characteristic feature of their own, 
namely that they interspersed also phonetically written Ak- 
kadian words and entire groups of words in the Hittite text. 
Whether these words and groups of words were also pro- 
nounced in Akkadian or perhaps in Hittite, is still undecided. 
At any rate, a written Hittite text always contains constituent 
elements taken from three languages: Sumerian ideograms, 
many of them with Hittite (or also Akkadian) flexional end- 
mgs, Akkadian words and word groups, and finally phoneti- 
cally written Hittite words. As an illustration, I quote here 
Article 11 of the Hittite laws, transliterated in Latin charac- 


ters, followed by its English translation. The use of lower- 
case italics in the transliteration indicates that the word or 
part of a word appears written phonetically in Hittite, where- 
as roman capitals indicate a Sumerian ideogram, and itali- 
cized capitals an Akkadian word or part of a word : 

tak-ku Lti.ULO LU el.lum na-a$-ma qik-su ku-il-Ai tu- 
wa-ar-ni-i%_-%i nu~ut-h zo gin kubabbar pa-a~i. 

"If somebody breaks a free man's hand or leg, he gives 
him (as a penalty) 20 shekels of silver." 

The stem of the word meaning "man" is written here by 
the Sumerian ideogram lu.ulu lt: to which the phonetic Hit- 
tite -an ending of the accusative case is added; el.lum, 
"free," is written in Akkadian, and so is, "his 
hand," whereas in the word gir-su, "his leg," the stem is 
written by the Sumerian ideogram gir, and the ending by 
the Akkadian possessive suffix, -su, "his." The amount of 
the penalty, 20 gin kubabbar, "20 shekels of silver," is writ- 
ten in pure Sumerian, and the words takku ("if"), naJfma 
("or"), kuisli ("somebody"), tuwarnizzi ("he breaks"), 
nu-s's'e ("now to him") and pai ("he gives") are written 
phonetically in Hittite. 

In observing texts which have been preserved in several 
copies, like the Hittite laws, we find sometimes that a given 
word appears phonetically written in one, while it is repre- 
sented in the other version by an ideogram the meaning of 
which is known and thus helps the interpretation of the 
phonetically written Hittite equivalent. Thus, Article 15 
of the laws appears in one copy as tak-ku uJ.ulo 1 -"-*/ el.lam 
iS-ta-ma-na-ai-la-an ku-iS-ki if-kal-U-a-ri 1 2 GIN KUBABBAR 
pa-a~i, whereas in the duplicate is'famanas'san is replaced by 
GESTU-an ("ear"), so that the translation is: "If somebody 



slashes a free man's ear, he gives 12 shekels of silver." The 
word for "not" is written in one copy of the laws in Ak- 
kadian, u.ul, whereas in another copy it appears in Hittite, 
as na-ar-ta. 

Occasionally, the ideographic symbols and Akkadian signs 
may be so predominant in a sentence that the meaning of 
the entire sentence can be determined on their basis and also 
the lexical meanings and grammatical functions of the Hit- 
tite forms appearing among them can be ascertained. This is 
illustrated by the following example, taken from the descrip- 
tion of a religious festivity: 

2 DUMU £.GAL AM A LUGAL SAL.LUGAL ME.k QA.TI pt-t-da-an-^i 

lugal SAL.LUGAL 5u ME3 -s v M v ar-ra-an-yi , 

("Two palace officials extend hand-water to the king [and] 
queen, king [and] queen wash their hands.") 

Here the ideographic elements, dumu ("palace of- 
ficial/s/"), lugal ("king") and salxugal ("queen"), the 
Akkadian words ana ("to," used as the sign of the dative 
case) and me.e qa.ti ("water of the hand"), as well as the 
mixed ideographic-Akkadian su mes ("their / 
in Akkadian/ hands /su ME9 /") are clear and help also 
in the interpretation of the two verbal forms appearing in 
phonetic Hittite characters, pedanzi ("they extend") and 
arranzi ("they wash"), both of which are identified as pres- 
ent tense, third person plural, by the ending -anzi. 

The consistent utilization of the semi-ideographic method 
of writing (ideographic word stem and phonetically written 
ending) furnished first of all the possibility to determine 
which endings belong to the noun and which to the verb, 
and in what functions. lugal-uS ("the king"), LUGAL-un 
("the king," used as direct object), SAL.LUGAL-ri ("to the 
queen"), etc., contain declensional endings, whereas the 





endings appearing in gul-iui [cuL-ahhun] ("I struck"), 
Du-at ("he made"), etc., are conjugational suffixes. The sys- 
tem of ideography was useful above all in the laws which by 
their very division into short sentences of the "if some one 
does such and such a thing, he shall pay such and such a 
penalty" type restricted the range of possible meanings very 
narrowly and often, when they contained a sufficiently great 
number of ideographs, became fully interpretable. This was 
why Hrozny used the laws as the basic texts for working out 
the first grammatical and lexical picture of the Hittite lan- 

One must bear in mind in this connection that while the 
uniform, systematically recurring representation by ideo- 
graphs of the most frequently used words will readily reveal 
the meaning of such a word to the decipherer, it will still 
leave the spoken form of that word a secret, until and unless 
some fortunate accident gives a clue to it. We are, however, 
still ignorant of the spoken form of just such common, every- 
day words, many of them important also for comparative 
Indo-European philology, like son, brother, sister, wife, horse, 
dog, etc., for the very reason that they were always written by 

The meanings of a number of Hittite words and the most 
important grammatical functions of the language were thus 
determined relatively quickly by the combinatory method, 
i.e., by logical conclusions based on the relationship of the 
words to the rest of the sentence. And as long as Hrozny 
used solely the combinatory method, he obtained indispu- 
table, lasting results. Encouraged, however, by the results 
accomplished in the Akkadian field (cf. p. 58) by the etymo- 
logical method, the interpretation of meanings based on the 
identity or similarity of the spoken forms of words, Hrozny 
then felt that he was Justified in applying the same method 
to the Hittite language as well, i.e., in deducing the mean- 

jngs of unknown Hittite words from words of identical pro- 
nunciation in other Indo-European tongues. The ease with 
which this method may creep into combinatory research is 
well illustrated by the following example: 

The sentence nu NiNDA-an e-iz-za-at-te-ni wa-a-tar-ma 
e-ku-ut-te-ni means "Now you eat bread; water, however, 



In this sentence, the ideogram ninoa, for "bread," was 
positively identified, and it was logical to assume that the 
verb governing "bread" as its direct object meant "to eat," 
although this judgment might have been influenced subtly 
also by the Indo-European etymology with the New High 
German essen, the Latin edere, etc. Bearing in mind the well 
known parallelism us membrorum of the ancient Oriental 
languages, Hrozny correctly deduced that the second half of 
the phrase, linked to the first one by -ma, meant "water, 
however, you drink"; yet the sound of the noun wa-a-tar, 
identical with the pronunciation of the Germanic watar, 
"water," another etymologically correct conclusion, is a fac- 
tor in this instance, too. Now then, the method of judging 
by such identical spoken forms, without any objective sub- 
stantiation, implies the great danger that the interpreter will 
be misled by false identities of sound once in a while. This is 
what happened repeatedly to Hrozny, too. Thus, he trans- 
lated the verbal root da as "give," because of the spoken 
sound which was identical with the sound of the stem of the 
Latin verb dare, whereas later combinatory research deter- 
mined that its meaning was the very opposite— "take." The 
word appa means "back," but thinking of the Greek ap<5, he 
translated it as "away"; he translated piran ("in front") as 

around" because it sounded like the Greek peri, the noun 
arkuwar ("prayer"), on the analogy of the Latin arcere, as 

'defense," interpreted nawi ("not yet") as "new," and so 
°n and so forth. 



Thus, Hrozny's combinatory researches— fundamental as 
such, and also correct in many an instance— do contain many 
an element of uncertainty because of the infiltration of the 
etymological method, and the young discipline of Hittitology 
might have gotten on the wrong track if F. Sommer had not 
pointed, in his monograph entitled Hethitisches* tempera- 
mentally at the impending dangers and had not steered the 
research back onto strongly combinatory paths. The au- 
thority of Sommer, supported by several younger researchers, 
must be given the credit for the fact that the new discipline 
developed into a strict and secure philology within a few 

There is not much that can be said as yet with respect to 
the linguistic neighbors of the Hittite tongue in Asia Minor. 
The non-Indo-European Pwto-Hattkn— the language of the 
original inhabitants of Hatti (Kbatti), the characteristic 
feature of which is a flexion operating chiefly with prefixes, 
instead of suffixes— was perpetuated by the Hittites as the 
ritual language of the most important state deities, like the 
Sumerian language in Babylonia, and therefore the student 
priests had to learn it in school as a dead language. For this 
reason, the Hittites provided many Proto-Hattian texts with 
Hittite translations, and these bilingual texts offer to modern 
research the opportunity to make a headway, step by step at 
least, into this strange language, while monolingual Proto- 
Hattian texts are still as good as unintelligible to us for the 
time being. In Fig. 36 I quote a few sentences from the best 
known Proto-Hattian-Hittite bilingual text, an adjuration 
spoken at the laying of a new cross-bar in the palace. The 
reader might use these sentences to test his own translating 
ability, but I present here at least a brief summary of the most 
important word equivalents: 

* Leipzig, 1920; Bogazkoy Studies, No. 4. 


(II 40) wa-ai-ha-ab-ma H-S'u-ttr ai-ka-ah-hi-iir iu-il-wu 
(41) lmu Ha-at-tu-uS ti-it-ta-ah-%i-ta-at iu-ii-wa (42) Ta-ba-ar-na ka-a- 
at-tl ta-ni-wa-aS = hittite (43) D1NGIR MES KUR™ ma-ni-ia-ah-hi- 
it da -a -ir -ma -at UR0 £ta-at-tu-Si (44) sal-li G1! StT.A da-a-ir-ma-at nu-zs 
La -ba-ar-na-a5 LUGAL-u[s c-eS-zi] 

"The gods allotted the countries; but for Hattusa they took the great 
throne, but they took it, and Labarna is king, " 

[proto-hattian] — 

(III 19) a-an-tu-uh li-e-^u-tt-ub li-ei-te-'ra-dh 'b'a-la Ifi-e-ie- 
jp-Se-ip] (20) ba-la ax-ne-eS ka-a-\\a*an'Wa-lu-id-Au-u-mi — hittite. 
(zi) da-a-as-ma-as-za TUG" 1 -* k^GAR.TAGH^ KLi E. SIR« LA -ia 
(22) na-aS-sa-an da -a -is GIS DAG-ti 

(23) [a-an-tu-uh?] W}-t?*&-da-ta-nu pa-la li-e-iz-Z.i'b'-'i' 
(24) pa-la [an-ne]-e$ ka-a-ha-an-n!a-$u-id-du-u>t=^B-VTTiTV..[z%) [da]-a-a5?- 
[ma-zk GAP.KIN.AG? ZAf?.Z£/?-ia na-at-£a-an da-a-iS Gi*DAG-ti 

"But he took the garments, the drapes (?) and the [the shoe]s and 
placed them onto the throne, — But he took cheese (?) and rennet (?) and 
placed it onto the throne." 

' ' \\\ fll 


(40) a-la-ah ta-aS-te~nu-ti-tva bi-e-m-il (41) H-ga-a-ru ta-ui- 
fe~e-ta-nu-ii-wa = hittite. (42) nu-wa-kan i-da-lu-us an-da? li-e ti-iz-zi 
(43) i-da-lu-uS-wa-kan TJKU-as Sri an-da li-e li-iz-zi 

'And an evil one shall not come in, an evil man shall not come 

into the 


[proto-hattian] — 

( ; 1 ) ma-al-hi-ib-hu ? te-e-ta 



iu-u-ul a-ia-ah-bi (j 

[2) ta-ai- 
iu-u-ta lu-u-la ^Su-li-in-kai-ti ha-at-ti (;j) ia-ni-wa-al u-un-hu-bi = 
hittite, (54)na-as-ta a-as-su an-da tar-ni-es-ki-id-du (J5) i-da-lu-ma- 
kan an-da li-c rar-na-a-i (j8) d Su-li-in-kat-ti-is-5a-an LUGAL„us 
an-da e-es-zi 

"Then he shall let in the good one, 
King Sulinkatti is inside." 

shall not let in; 

Fig. 36. From a bilingual Proto-Hattian-Hittite text. (Keil- 
schrifturkunden aus Boghazfcdi, Vol. II, No. 1, column II, lines 40- 
44; column III, lines 19-25, 4°-44f 5 l ~S^- 



washab-ma = Hittite dingir* 1 * = "gods" 

eiwur = Hittite kur" e| = "countries" 

asKahhir = Hittite maniiahhir = "they allotted" 

suwa probably = Hittite dair-ma-at = "but they took," but it 

may have another meaning 
titahzilat = Hittite salli Gll grj.A = "the great throne" 
katti (katte) = Hittite lugal-us" = "king" 
teniwaS — Hittite eszi = "there is" 
antuh = Hittite datf-ma-as-za — "he took them (-as") however 

(-ma) to himself (-za)" 
le-zub = Hittite -rtfo? 1 -* = "garments" 
le-sterah = Hittite Ktj! GAR . T A G *''* = "drapes (?)" 
bala (pala) = Hittite -ia and nu = "and" 
le-sepsep = Hittite Kui E .siR? 1 '* = "shoes" 
annes" = Hittite dais" = "he placed" 
ka-hanwasuiddun = Hittite °» S DAG _/; = "onto the chair" 
Je-udatanu = Hittite = "cheese" 
a-sah = Hittite idalut = "evil" 
tas'-re-nuwa (tas"-te-ta-nuwa) = Hittite ande le uizzi = "he shall 

not come in" 
he-uil = Hittite E-ri = "into the house" 
isgaru = Hittite idalus UKtj-as = "an evil man, villain" 
maihib = Hittite ass'u = "good" 

te-tah-sui = Hittite anda tarneskiddu = "he shall always let in" 
a-sah-bi = Hittite idalu-ma = "but the evil" 
taS-tuta sula = Hittite anda le tarnai = "he shall not let in" 
unhubi = Hittite anda = "inside"? 

Like Hittite, Luwian and PaJaian also are Indo-European 
languages, yet in these instances, too, caution must be exer- 
cised in applying the etymological method. In a few in- 
stances, Hittite parallel texts could be found for certain 
Luwian texts, and thus it was possible to undertake a certain 
combinatory work on the Luwian language. Fig. yj presents 
a sentence from the best known such bilingual texts, where 
the concordance, by the way, is not quite literal. Luwian texts 



Luwian: (zi) ^Sa-an-ta-aS LUGAL-wl & An-na-ru-um-mi-en-xi 
(13) aS-faa-itu-tva-an-tn ku-in-^i wa-aS-Sa-an-ta-ri (24) ^Lu-u-la-hi-iti- 
%d-a$-far Jiu^u-up-pa-ra-^a ku-in~x* hi-ii-hi-ia-an-ti = hittite. (I 36) 
e-hu A MARDUK kat-ti-ti-ma-at-ta d In-na-ra-u-wa-an-ta-a3 (37) 
li-wa-du e-eS-ha-nu-wa-an-ta ku-e-eS u-e-eS-Sa-an-ta (38) lXl 1£Ei Lu- 
u-la-hi-k-aS-Sa-an hu-up-ru-u§ ku-i-e-e5 iS-hi-ia-an-ti-il 

(Translation from the Hittite part) "Come, Marduk (= Luwian "King 
Santa"), but let with you come the 'robust (?) (gods),' don the blood' 
stained (clothes), (also) the Lulahi (gods) who (are) swathed in . . ." 

Fic. 37. From the Luwian-Hittite QuasibiUnguis (KeiJschriftur- 

foinden aus Boghazkoi, Vol. IX, No. 31, col. II, lines 2: 
lines 36-38) 

-24 = col. I, 

without Hittite parallels still remain very difficult to handle. 
As for Palaian, no real possibility of interpreting this lan- 
guage has even been found as yet. 

( k } The In terpreration of Human 

The only Hurrian records heretofore known are mono- 
lingual texts, written in Babylonian cuneiform characters. 
Quite recently, however, a text consisting of a few lines 
written bilingually in Akkadian and Hurrian was found by 
the French in Ugarit (Ras Shamrah), and it is expected to 
be made public in the near future. But it is too short to be 
likely to help us much, and also the isolated Hurrian words 
appearing here and there in the Akkadian word-list are of not 
much practical use. Thus, on the whole we still remain de- 
pendent on the combinatory method which involves many 
more difficulties with respect to the interpretation of Hur- 
rian than in the case of Hittite, because the Humans used 
mainly a phonetic script and rarely employed ideograms. 
While in the Hittite texts the ideograms reveal the meanings 
of the words, even though they give no clue as to their pro- 
nunciation, the Hurrian records show the spoken form of a' 



most every word, but by doing so, they preclude the possi- 
bility of fathoming its meaning. 

This is why the religious texts written in the Human lan- 
guage, a great many of which are contained in the Hittite 
archives of Bogazkoy, and some specimens of which, in 
vowelless Ugaritic script, appear also in the temple library of 
Ugarit, are still as good as unintelligible, except for the mean- 
ings of a few words which Carl Georg von Brandenstein has 
deduced by the combinatory method. 

A more favorable aspect is presented by the "Mitanni 
Letter" of King TuSratta (cf. p. 46), discovered in the ar- 
chives of El Amarna, Egypt, in 1888. Tusratta sent a number 
of lengthy letters to Egypt, written in Akkadian and contain- 
ing a great many repetitions, and he sent also a very long 
communication, containing more than 400 lines, in his native 
Hurrian tongue. Now then, in any text written in cuneiform 
characters, the proper names, names of deities, geographic 
names, etc. are always clearly identified as such by their de- 
terminatives, even though the language of the record may 
be absolutely unknown to the reader. Since the names oc- 
curring in the Mitanni Letter are the same ones as appear in 
the Akkadian letters of TuSratta, it may well be assumed that 
also the Human letter deals with the same topics as the Ak- 
kadian epistles. Thus, soon tolerably intelligible word com- 
binations can be identified, above all in connection with the 
names, e.g., r Ni-im-mu-u-ri-i-as* kur Mi-zi-ir-ri-e-we-m-es - iw- 
ri-is", meaning "Nimmuria {name of the Pharaoh), king 
(hvrijf ) of Egypt (kur Mizirri) ," or I Ar-ta-ra-a-mas' am-ma-ti- 
jV-wu-us 1 , meaning "my grandfather, Artatama" (known as 
such from the Akkadian letters of TuSratta) . Let the follow- 
ing examples be cited here for entire short sentences recur- 
ring in Akkadian as well as in Hurrian: The Hurrian 
dingir 1 ^ e-e-en-na-su-us - na-ak-ki-te-en — the Akkadian 
iJam UES li-me^SSe-m-Sa = "may the gods will it"; the Hur- 


rian i-nu-u-me-e-ni-i-in "Si-mi-gi tar-su-an-ms 1 . . . ta-a-ti-a = 
the Akkadian ki-i-me-e a-mi-Iu-u-tum Mamas' i-ra-'a-am-su as 
"as mankind loves the sun" [Hurrian inu = Akkadian 
kl (-me) = "as"; Hurrian tarsuanni- = Akkadian amilutum — 
"mankind"; Hurrian tat- — Akkadian ra' amu = "love"]. 
The utilization of such opportunities enabled Jensen and 
Messerschmidt in the late 19th century, and recently also 
other investigators, to make considerable progress with the 
aid of the combinatory method in the understanding of the 
vocabulary and of the quite involved grammatical forms of 
the outlandish Hurrian language, and to translate at least 
individual passages of the Mitanni Letter more or less re- 
liably. Yet, although that text has been explored and analyzed 
for decades, the clearly or at least fairly intelligible parts still 
alternate with long passages which can be translated with 
great difficulty only or not at all. 

(1) The Interpretation of Urartaean 

The Urartaeans were younger kinfolk of the Hurriai 
They lived in what later became Armenia and left behind 
them about 180 inscriptions of various lengths (building 
dedications, votive inscriptions, war reports and individual 
annal passages), dating from the cjth-yth centuries B.C., 
written in the Neo-Assyrian cuneiform script, but mostly in 
the Urartaean language (which certain research linguists 
prefer to call Chaldaean ) , The relationship of Hurrian and 
Urartaean is, however, not close enough to permit us to ex- 
pect any result from the etymological method; the Urartaean 
language must be interpreted by the combinatory method, 
without outside references. The first assistance is given to us 
once again by the names (of persons, of deities, of geographic 
units) written in combination with certain determinatives, 
then by the short and clear phrasing of most of the inscrip- 





tions, as well as by the not too sparing use of ideograms. In 
studying inscriptions of very stereotyped contents, one often 
finds an ideogram in one and phonetic symbols in the cor- 
responding spot in another. Thus many a linguistic fact has 
been deduced from the monolingual inscriptions by the com- 
binatory method alone. There are also two stelae known 
which are inscribed bilingually, in Urartaean and Assyrian, 
viz.: the KeliSin Stela, in the Kelism Pass on the Iraqi- 
Iranian border, and the nearby Stela of Topzauae. But only 
the first-mentioned one has been reported on satisfactorily; 
it has yielded a number of lexical equivalences as well as a 
few grammatical facts. But no complete, scientifically in- 
contestable representation of the Stela of Topzauae, much 
more difficult to read, has been made available as yet. 

(m) The Interpretation of Eaily Ehmite 

A few words are all that can be said about the inscriptions 
left by the early Elamite kings in Babylonian cuneiform 
script, dating probably from the 13th and 12th centuries b.c. 
They must be interpreted chiefly without any outside ref- 
erence, for there is only one among them to which there 
exists a very brief Akkadian parallel text. The Early Elamite 
cuneiform writing contains determinatives, but few ideo- 
grams. An important lexical help, however, is provided by 
theNeo-Elamite translations of the inscriptions of the Achae- 
menides which reveal the meanings of a number of Elamite 
words, although still not of sufficiently many to clarify com- 
pletely the far richer vocabulary and frequently differing 
grammatical structure of the more ancient language. Elamite 
was long a stepchild of the research into cuneiform literature, 
and only in recent years did Hinz make good beginnings with 
the combinatory interpretation of Early Elamite inscriptions, 

( n) The Decipherment of Ugarific 

Early Elamite closes the sizable list of the languages writ- 
ten in Babylonian cuneiform characters, and all there still 
remains to be discussed in this connection is the alphabetic 
writing used by the inhabitants of Ugarit, on the northern 
Syrian coast, in the 15th and 14th centuries b.c. This script 
was written on clay tablets and resembles the cuneiform 
writing in appearance at least. It was discovered by French 
excavators in Ras Shamrah as recently as 1929 and aroused 
great interest as a new factor come to light in the otherwise 
thoroughly explored territory of Syria. 

The differences between this newly discovered script and 
the already known Babylonian cuneiform writing was im- 
mediately recognized, because this script consisted of a mere 
30 characters, all very simple in shape, and there were no 
determinatives. These features suggested promptly that this 
was no syllabic writing like the Babylonian cuneiform script, 
but an alphabetic system like the Early Persian writing. The 
distinct separation of most individual words by vertical 
strokes seemed to be a help to decipherment, whereas the 
absence of bilingual texts had to be regarded as an impedi- 
ment. On the whole, however, the prospects for the decipher- 
ment of an alphabetic script are always favorable, because the 
small number of individual characters restricts the possible 
interpretations to a much narrower range than the one to 
be considered in the case of a syllabic writing which em- 
ploys a hundred or more different symbols. 

The separation of the words was a great help in the study 
of the morphology, and since the latter seemed to indicate 
that the language was a Semitic one, Hans Bauer, the Semi- 
tist of Halle, felt justified, in April 1930, in tentatively as- 
signing Semitic consonantal values to the characters con- 
stituting the undeciphered words. This attempt at reading 


the words as if they belonged to a Semitic language was just 
one of various possibilities, for the newly discovered 
script in itself might have suggested quite different languages 
as well. But Bauer was lucky: The language was actually 
Semitic, and he had interpreted 17 characters correctly. In 
the meantime, also the Frenchmen Dhorme and Virollcaud 
attacked the problem, and by the following year the value of 
each and every character of the Ugaritic alphabet was deter- 
mined correctly, without any bilingual text whatsoever and 
without any outside reference at all, solely on the ground of 
the assumption that the language in question was of a Semitic 
structure. Let it be mentioned that also the inscriptions ap- 
pearing on several bronze axes played a part in Bauer's de- 
cipherment. These inscriptions appeared in a shorter form, 
consisting of only six symbols which Bauer assumed to be 
simply the name of the owner, and in a longer version which 
contained four other symbols before those six. Bauer sus- 
pected that these four symbols represented the word for 
"axe," which in Hebrew is garzen, written by the four con- 
sonants grzn, whereas the corresponding word in the cog- 
nate Ugaritic language is, as it was established later, hrsn 
(cf. Fig. 38). The Ugaritic consonantal alphabet is shown 
in Fig. 30. 

The accuracy of the consonantal values deduced by Bauer, 
Dhorme and Virolleaud was soon corroborated by various 
facts. The first and most immediate one of these confirming 
facts was that the soon published long texts, even though 
still obscure in many particulars, turned out to contain gen- 
erally well intelligible mythological tales in the Ugaritic lan- 
guage, still an unknown tongue and yet related to the Semitic 
languages. Moreover, the assignment of those values to the 
various characters yielded also a number of Hurrian names 
of deities which had not been sought in the texts and yet, 
once identified, formed a closed circle. And finally, later dis- 



coveries of Akkadian and Ugaritic lists of names of cities, 
etc. furnished a further confirmation of the correctness of 
the consonantal values deduced. These lists are not bilingual, 
they do not contain identical lists in two languages. They are 
monolingual lists, some in Babylonian cuneiform script, 
others in Ugaritic cuneiform characters, enumerating the 
inhabited places of the land of Ugarit in a strongly varying 

Fig. 38. Ugaritic axes with in- 
scriptions. (From Friedrich, Ras 
Schamra, Fig. 4. ) 

order. Thus, even if the name of a town appears in a certain 
place on a tablet bearing inscriptions in Akkadian, it can- 
not be expected to occur in the corresponding place on an 
Ugaritic tablet. They would hardly have offered a point of 
departure to the first decipherers, but they help us today 
better to understand the phonetic structure of Ugaritic by 
equating the Babylonian ^h-pi with the Ugaritic Hzp, the 



Babylonian -j*** with the Ugaritic Bq't, the Babylonian 
y?* with &e Ugaritic Jkn'm, etc. (sic* is an ideogram 
for "good/' and also the Ugaritic rim means "good") . 

I would not want to withhold a proof of the correct read- 
ing and translation of the figures from my readers. A similar 
list enumerates towns and villages and their deliveries of 
wine, indicating the numbers of the jugs delivered, and the 
numbers are written phonetically in Ugaritic. The sum total 
of these numbers is 148, and the list actually ends with the 
expression, written in Akkadian and using numerals, "1 me- 
at 48 dug cestui" = "148 jugs of wine." Thus, we may feel 
absolute confidence in the rapidly accomplished decipher- 
ment of the Ugaritic script, too. 

3. The Hittite Hieroglyphic Writing 
The late 19th century regarded the Hittite hieroglyphic 
writing as the third great problem in the field of ancient Ori- 
ental scripb, ranking as such with the Egyptian hieroglyphics 
and the cuneiform writing. A different view is being' taken 
nowadays. The plethora of texts written in cuneiform script 
has revealed to us so much about the races and language of 
eastern Asia Minor and northern Syria that the not very 
numerous Hittite hieroglyphic inscriptions with their not 
too significant contents appear somewhat pallid next to 
them; as a matter of fact, they are fust a late, weak offshoot of 
Hittite civilization. But, nevertheless, these inscriptions per- 
mit us to form an ever clearer idea of still another Indo- 
European language beside cuneiform Hittite, Luwian and 
Falaian, and moreover, the problem of the decipherment of 
hieroglyphic Hittite still remains an interesting one. For in 
1930, after six decades of futile efforts, finally a practicable 
method was found to achieve a combinatory decipherment 
without a bilingual text, and the later discovery of a large bi- 
lingual text confirmed and expanded the results obtained 


previously. Thus, in the case of the Hittite hieroglyphs there 
exists actually the possibility of checking up on the ability of 
the decipherer, 

(a) General Facts 

Monuments bearing inscriptions in Hittite hieroglyphics 
have been known since the middle of the 19th century ah., 
especially those found in eastern Asia Minor, on and about 
the soil of the ancient Hittite empire, as well as in the ad- 
jacent part of northern Syria, above all in the important city 
of Karkhemish at the bend of the Euphrates. The writing 
used in these inscriptions is primarily a monumental writ- 
ing, but it appears also on seals. Monuments bearing inscrip- 
tions in hieroglyphics and seals with cuneiform and hiero- 
glyphic characters are known to us from as early as the era 
of the Hittite empire (between 1400 and 1300 b.c.) . On the 
other hand, the major part of the North Syrian inscriptions 
are of more recent origin; they date from the ioth-8th cen- 
turies e.g. With the gradual absorption of Syria in the As- 

Fig. 59. Hieroglyphic Hittite inscriptions from Karkhemish. 
(From Friedrich, Entziffenuigsgescfrichte der fiethirischen Hiero- 
giyphenschj-ift, Fig, 2.) 



Fig. 40. Seals with cuneiform-hieroglyphic Hittite inscriptions. 
(From Friedrich, Entzifferungsgeschichie der Jiethitischen Hiero- 
glypftenschrifr, Fig. 5.) 

(a) Tarkummuwa of Mira. (b) Indilumma. (c) Kputahsu of Kizwatna. 
(d) Tabarna of fjatti. (e) Amuwanda of tfatti, (() So-called Ziti seal, (g) 
Urtji-Teiup of fjatti. (h) Suppiluliuma of Hattf. (i) IJattusili of IJatti. (j) 
tfattuSili and his wife, Puduhepa. (fc) Queen Pudufcepa of tfatti. 

Syrian empire, the hieroglyphic writing disappeared about 
700 B.C. A hieroglyphic Hittite inscription, from Karkhe- 
mish, appears in Fig. 39, whereas Fig. 40 shows several cunei- 
form-hieroglyphic seals. 

A special comment is necessary with respect to the foils 
oi lead found in Assur, probably brought there from some- 
where else, which are inscribed with a very cursive hiero- 
glyphic writing. Contemporary research considers them 
letters; the fact that they were found, rolled up, in the 
foundations of a house is probably attributable to a mistake 
of the builder who erroneously regarded them as having 
magic powers. Also the cursive form of writing suggests 
ordinary, every-day contents. It seems therefore that this 
writing was used on monuments in an elaborate, pictorial 
form (occasionally, however, also with cursive forms) and 
in daily life in a strongly cursive variety, analogously to the 
situation in Egypt where both the hieroglyphic writing of 
the monuments and the cursive hieratic and demotic scripts 
of ordinary daily life remained in use side by side. 

The designation "Hittite" hieroglyphics was introduced 
by the British research philologist Sayce, shortly after 1870. 
Egypt and Babylonia-Assyria were practically the only two 
nations of the ancient Orient known to the philologists of 
those days as two culturally and scripturally clearly distinct 
entities, and the newly discovered script on monuments of a 
likewise distinct art appeared as the legacy of a third civilized 
race of the Orient of remote antiquity. Both Egyptian and 
cuneiform records had made reference to a land and people 
named Hatti (or Khatti) in northern Syria, and the Hittites 
are mentioned occasionally in the Old Testament, too. This 
was the reason why Sayce regarded these monuments as 
works of the Hittites, and he considered them at first a Se- 
mitic race. The situation was clarified only by the discovery 
of the Hittite archive of Bogazkoy which revealed the Hit- 



tite language to be an Indo-European tongue, even though 
written in cuneiform characters. There arose then, however, 
the question as to the closer relationship of cuneiioim Hittite 
and hieroglyphic Hittite, which still has not been answered 
completely, The two languages are indubitably closely re- 
lated but by no means identical. In fact, hieroglyphic Hittite 
shows features closely related to Luwian. But when answer- 
ing this question, we must bear in mind that all the lengthier 
hieroglyphic inscriptions date from the last eras of hiero- 
glyphic writing, and that our only relics of the Hittite em- 
pire and cuneiform Hittite still consist merely of a few 
brief and linguistically barren inscriptions. 

(b) The Basic Principles of the Hieroglyphic Script 
and the Possibility of a Decipherment 

As we know today, the Hittite hieroglyphic writing also 
consists of the same three elements as the Egyptian writing 
and the Babylonian cuneiform script, viz.: word-signs (ideo- 
grams), phonetic symbols, and determinatives, some of the 
latter prefixed and some suffixed to the word which they 
qualify. The important fact is that the phonetic symbols, 
unlike the Egyptian symbols, but analogous to the cunei- 
form characters, represent clearly distinct syllables, indicat- 
ing the vowels. The individual words are often (but unfor- 
tunately, not systematically) separated from each other by 
the separation mark IC. The characteristic features of the 
word-signs consist in their carefully drawn pictorial shapes, 
their relatively rarer occurrence, and their position at the be- 
ginning of the words. Simpler and often cursive symbols 
occur very frequently, chiefly in the second half of a word 
after such pictograms; they may be regarded as the phonetic 
(syllabic) elements. Since in a pictographic script the word- 
signs are mostly directly understandable as pictures, they may 
well be expected to furnish clues as to the meanings of the 


words and the grammatical (nominal or verbal) functions of 
the endings. Let it be pointed out, though, that in the Hittite 
hieroglyphics the pictorial character of the symbols often 
became indistinct as a result of the conventionalization of 
their shapes. Who would expect, for instance, the symbols 
shown in Fig. 41 to stand for "house," "sun" and "god," 


Fie. 41. Hittite hieroglyphic symbols for 
<aS "House," "Sun" and "God." 

respectively? As in the cuneiform script, also in the Hittite 
hieroglyphic writing, the determinative of names of deities 
is identical with the pictorially unclear ideogram of "god." 
Ten persons are referred to by name on the family relief of 
Karkhemish, and each name is introduced by a small, oblique 
stroke, the determinative for persons which resembles the 
vertical wedge appearing before the names of male persons 
in the Babylonian cuneiform writing and seems to have been 
created in imitation of the latter. 

The direction of the writing is made evident by two clues: 
First, the picture of a person pointing at himself, appearing 
at the beginning of the inscriptions and meaning "I (am)" 
(cf. Fig. 42), and secondly, the unfilled portion of the last 

U& *ffi Fig. 42. The Egyptian and hieroglyphic Hittite picto- 
si W grams for the pronoun "I." 

line of many an inscription (cf. Fig 39) . It is evident also 
that, as in the Egyptian writing, the heads of all human ani- 
mal figures are turned as if they were looking toward the 
beginning of the line, the hands are stretched in that direc- 
tion, and also most of the feet appear to be walking that way. 
Moreover, the writing alternates in direction from line to 
line, in other words its direction is boustrophedon (a Greek 
expression which means literally "as the ox turns" and is gen- 


erally used as a technical term to designate this regularly 
alternating direction of writing, as an allusion to the direc- 
tion in which teams of oxen walk when ploughing a field) . 

Particular difficulties were presented by the definition of 
syllabic phonetic values. The first students of the script had 
practically no bilingual texts available to enable them to 
identify names as such; the two cuneiformly-and-hieroglyphi- 
cally inscribed seals of Tarkummuwa and Indilumma (Figs. 
40-a and 43 ) contained so many obscurities even in the cunei- 
form portions that they did more to mislead than to clarify 
matters. The researchers thus were forced to look for other 

Now then, it could be safely assumed on the analogy of 
other inscriptions of the ancient Orient that the royal au- 
thors began also these records by stating their names and 
titles and designating themselves as the king of such and 
such land or city. Sayce had already identified the ideo- 

Taikv-muwa king Me+ra-4 land 
"Tarkummuwa, king of the land of Mera' 

^1 mkVM 

Fig. 43. Text of the Tarkummuwa seal. 

grams for "king" and "land" on the basis of the bilingual 
Tarkummuwa seal (Fig. 43), and these helped now to ana- 
lyze also the initial portions of other inscriptions into "I 

Nfrar nit i 

(a) Kar-ka-me '" 

(c) Ku+i(a)-ku-nia OITI 

(d) f| ? en 

(b) Tu-wa-nu-wa 01 "" 
(d) A-ma-tu 1 -"" 1 

Fig. 44. Hieroglyphic Hittite writing of the geo- 
graphic names Karkhemish, Tuwanuwa, Gurgurna 
and Ham at. 



/am/ X (a name written phonetically) , king of the land <Jh 
0T of the city [jj /of/ Y." And in fact, in the many inscrip- 
tions found in Karkhemish, for instance, all those places 
where names of lands or cities were expected to appear 
showed always the sequence of symbols shown in Fig. 44, 
followed by the determinative 1 ("city") or M ("land"). 
This group of symbols could therefore be suspected of being 
the name of Karkhemish. The corresponding places in the 
inscriptions from Tyana (Tu-wa-nu-wa in cuneiform Hit- 
tite), Mar'aS (Gur-gu-ma in cuneiform Hittite) and Hamat 
were always occupied by the respective groups of symbols 
shown in Fig. 44. 1 have deliberately selected mutually con- 
firmatory examples: Tuwanuwa contains the syllable wa 
twice, Gurgurna contains gu twice (in the first instance with 
an r) , the syllable tu appears in Tu-wa-nu-wa and in A-ma4u t 
ma in A-ma-tu and Gur-gu-ma. This method permits the 
identification of geographic names and, from those, of syllab- 
ic signs with a convincing assurance, without any bilingual 
text such as Champollion had, and without the knowledge of 
any list of monarchs, such as Grotefend had available. 

Names of the monarchs of the "hieroglyphic Hittites" 
were not as well known to the decipherers as had been the 
names of Persian rulers to Grotefend, but the Assyrian kings 
had recorded the names of a few North Syrian kings in their 
reports in cuneiform script on their military campaigns 
against North Syria. If the exact era of such a king was known 
and a hieroglyphic inscription from his city could be de- 
termined archaeologically to originate from the same era, 
there existed under certain circumstances a certain degree of 
probability of deducing also the correct reading of the names 
of "hieroglyphic Hittite" monarchs. This was how a Mu-wa- 
ta-Ii of Gurgurna, an Ur-hi-If-nu of Hamat and a Wa-r-pa- 
la-wa of Tuwanuwa were re-discovered in the inscriptions 
(Fig. 45). 


(a) Mu-wa-ta-li^) U+r(a)-bi-Ji-t,a a (c) Wa-Yr(a)-pa-la 


Fig. 45. Hieroglyphic Hittite writing of the names of the 
monarchs Muwatali, Urhtlinu and Warpalawa. 

(c) The Progress of the Decipherm ent 

With my discussion of the possibilities of a decipherment, 
I began, unawares as it were, to report on the course of the 
decipherment itself and have already mentioned some of the 
results of the research. Of course, these results were not ob- 
tained as easily as they seem to be to some one who is in the 
position to survey the whole matter after the conclusion of 
the progress and development. The progress of the research 
was much more complicated and controversial, and the 
earlier investigators in particular did not proceed along the 
lines of sufficiently clear and sharp logical considerations. On 
the other hand, though, they had to work with far fewer in- 
scriptions than the later analysts and our own contempo- 
raries. Thus the interpretation of those very inscriptions was 
a very painstaking process which involved many a blind alley; 
we may say, in fact, that for sixty long years, from r8 7 o to 
1930, everything in this domain was vague and uncertain. A 
report on all those uncertainties seems to be uncalled for 
today, when we finally stand on a firm foundation. Those 
who wish to know the state of affairs just before the outbreak 
of World War II are referred to the author's Eirtzifferungsge- 
schichte dex hethitischen Hiaogfyphensdmft* In the pres- 
ent book, I shall list merely some of the most important 
positive results. 

We owe a debt of gratitude to Sayce, the first investigator, 

tel 'P 5 * 01 ? of the Decipherment of the Hittite Hieroglyphic Writing." 
(Stuttgart, 1039; special issue No. 3 of Welt als Geschichte.) 


for the recognition of the meanings of the ideograms for 
"kinV "city/ 1 "land" and "god," as well as of the symbol 
s~\ as the ending of the nominative case (s) and of >$j- 
as the ending of the accusative (n) . Observing that all desig- 
nations of deities began with the ideogram for "god," he 
concluded correctly also that this ideogram was used as a 
determinative before names of deities as well. These correct 
interpretations, however, are buried under a vast number of 
fanciful and wrong decipherments and translations. Menant 
was the first man to recognize correctly (about 1890) that 
the picture of the person pointing at himself which appears 
at the beginning of the inscriptions has the same meaning as 
the Egyptian hieroglyphic depicting a person pointing at 
himself, i.e., I (cf. Fig. 42}, and that the first symbols of the 
inscriptions are to be translated, analogously with many other 
ancient Oriental inscriptions, as "I /am/ X." 

The most comprehensive, but at the same time also 
most difficult and most controversial attempt at a decipher- 
ment and interpretation in the early era of this research was 
undertaken by Jensen in 1 894. He identified the name of the 
city of Karkhemish correctly, but his subsequent determina- 
tion of the phonetic values of many symbols— some of which 
he regarded, quite unsystematically, as representing individ- 
ual consonants and vowels, others as open and closed sylla- 
bles, still others as symbols for more involved groups of 
sounds or ideograms— was so arbitrary that his findings were 
disputed even then, not even to mention his opinion that 
the hieroglyphic inscriptions contained the Indo-European 
Armenian language in the same form in which it was handed 
down to us in the literature dating from the Christian era. 
Thus, the work of Jensen actually aroused distrust in the 
decipherments, and today, when the decipherment finally 
rests on a firm foundation, we may disregard without com- 


The Kings and Queens oi the New Empire 


year, B.C 


Hieroglyph af the 
Ktng. with reference 

lffi. Mi! 


^ N 

■■ f Qata^ 

MM ffi Htjagaj 




W'.*A «*>;** 

I ' 

TuThatiia. ffl. 
IJ "X 

{TLLtboUja ±J_ t 
ncl King) 



Hie:aglyph of Tava.. 

nanna., uilh reference 

(40 destroyed) 

(See below) 

< 7»'Su-k£fc. i tii™<at> 




S -<*■;.. 


^ 13-3? 




52 - 5? PJ 

(See below) 


^r nu vandal. 



1. Daau.hepa. 

(I. nme) 
2 ? 

I. Danuhepa 

[]. nme) 

{b,? ^acat J 


!£ iw ? 


Fig. 46. Hieroglyphics representing the names of great Hit- 
tite lungs and queens. (From Giiterbock, Siegel aus Bogazkay 
I, p. 01.) 

ment his first attempt in 1 894, as well as his later opinion that 
the inscriptions consisted of mere conglomerations of ideo- 
graphical^ written royal titles, not only without any histori- 
cal data but even without indicating the name of any person 
or geographic entity. Jensen's attempt at decipherment was, 


unfortunately, a useless waste of a great deal 
also Thompson, Cowley and Carl Frank, wh 
decipherment just "before and shortly after World War I 
failed to achieve any convincing result, 

Meriggi, Gelb, Forcer and Bossert were the ones who 
brought the whole undertaking onto more solid soil, about 
1030. The Italian Meriggi succeeded 'in identifying an ideo- 
gram meaning "son," and thus a genealogy which proved 
useful in the reading of the names of the monarchs. One of 
the words identified by Gelb was. the verb meaning "to 
make," the correctly transliterated form of which, aia-, be- 
came a factor of importance in the correct classification of 
the language of the hieroglyphic tests as cognate with 
Luwian. Forrer's recognition of a formula of imprecation 
was fundamental for the analysis of the sentence structure. 
Bossert discovered the proper reading of the royal name 
WarpaJawa and of the name Kupapa, the name of an often 
mentioned goddess. All the investigators named here operate 
with geographic names and names of persons, as those dis- 
cussed above. It is difficult to appraise the independent con- 
tribution of Hrozny, already known to the reader as the suc- 
cessful first investigator of cuneiform Hittite, who also 
began to work on hieroglyphic Hittite in 1932. 

In the years 1933-1937, also the number of cuneiform 
Hittite seals from Bogazkoy increased gratifyingly, and the 
investigators learned from them the hierogljphic Hittite 
written forms of the names of most great Hittite kings (Fig. 
46), although most of them in ideographic form, for 
Mu-ta-Ji = Mu(wa)ttalli was the only one written syllab- 
ically. But at least the names of the queens Puduhepa and 
Tanuhepa were written likewise with syllabic signs and 
yielded the phonetic values of a few more syllabic symbols, 
and they were helpful also in establishing the reading of the 



name of the chief goddess of the Hittite mountain shrine of 
Yazilikaya, &Ha-ba-tn = the cuneiform ^Hi-bat (Fig. 47) . 

<3E> CD & CH Fig. 47. Hieroglyphic Hittite version of the 
A Ha-ba-tx name of the goddess Hebat. 

Thus, to sum it all up, at the time when World War II 
broke out, the reading of a great many syllabic signs seemed 
to be certain or very probable, and we had a fair idea about 
the inflexion of hieroglyphic Hittite, too, on the basis of the 
lengthier inscriptions of later dates. The lexical research, 
though, was still encumbered with many uncertainties, above 
all also because of the not all too great number and slight 
variations in text of the hieroglyphic Hittite inscriptions. 
Thus even the translations already supplied by Meriggi and 
Hrozny" for most of the inscriptions had to be regarded with 
reservations, and in Hrozny s translations in particular some 
more question marks had to be added to the many already 
appearing in them. The Indo-European character of the 
language was only very vaguely recognizable. 

The fall of 1947 then brought a big, electrifying surprise: 
Bossert found several long inscriptions on the hill called 
Karatepe in eastern Cilicia, which were partly in hieroglyphic 
Hittite and partly in Phoenician and dated from the late part 
of the 8th century b.c. They soon turned out to be bilingual 
records. As of this date, the almost intact Phoenician inscrip- 
tions, totalling about sixty lines, have been published com- 
pletely, but the just fragmentaiily preserved hieroglyphic 
Hittite texts only partly, and even so not in their original 
form but as they were schematically copied by Bossert, so 
that there remain a few uncertainties concerning the se- 
quence of the symbols. It is therefore not yet possible to 
reach any conclusive judgment, but the extraordinary signifi- 








yqtljf. Co 3 


106 «S 1M 1« 10 * 

yQ<\ yw$ hj w*<9 

18 W 

At" h' 


f ;'• fa? f 


%y>? fit y»w q T/i* />*-? 


108 I m I 206 


xvtfi*, yw li*. ?wr 

w mm 

1 m* m > 3Y 


m 1 51 i I aij * I at J ao k J , zw M* 

/>3* fy c w yi7r ?w<\x to a> 



Fig 48. Sentences XIX-XXII and XXXVIII-XL of the bilingual 
text of Karatepe. (Bossert, Archiv Orienta'lru XVIII, 3, pp. 24-25, 
and Bossert, Jahrbnch fiir klemasiatische Forschung I, p. 272.) 

cance of the bilingual texts is obvious even so. We see to o 
great satisfaction that the determination by the combinatory 
method of the readings and grammatical forms, as worked 
out by the above-named investigators in the 1930's, was 
correct in every essential point, and also a great many of the 



Hieroglyphic Hittite: 
XIX. fort ba+r(a)-ni-$i-pa-wi "arm li-mi-tA-2 build? tu-mi-ha (rest de- 
stroyed) XX. ideocr. a"-tu-wa-a+rfa)-i-wa-ra "head -ti-i kwa?-ta-n(uj "i-ta 
a-s.-te ?-u-s.-?-?-i XXI. kwa?-wa kwa?-i "DOWN-nfuJ nu-ti ta?-ta "mu-ka-s.-s.- 
n(u) HousE-na-a XXII. "a'-mu-pa-wa'-ma-ri a-sw-da-wa-f-ra-l. "FOOT-pa- 
ti-i-n(u) "DowN-nfuJ "tu-fja" 

« PP^ 111 ' " i ' Wa ' F0RT "' STONE? ^-mi-fja XXXIX. wa-tu-ta a-Si-i-da-wa- 
ta-a-n{uJciTY "a-ta-ma-I-n( a ; tu-ha XL. kwa?-pa-wa-mu SEizE-n(u) "<«weath- 

ERCOD-hu-1-S. d DEER-HEAD-I-i.-ha* "J.-fa I-da FORT-S3 STONE tU-mi-n(u) 


.JS **■?£*■$ ' zt b ' fcJ # 1 gblm b-mqrani XX. b-'i in 'Ira fm 
hl j2$? XXL * W * M *■ hht M XXII. W-'afc 'ztwd It-nm thi p'm-i 
XXXVIII w-bn *nk h-grt z XXXIX. wit 'nfc Ira 'ztwdj XL. k b'l 
W-rip sprm Hh-n I-bnt 

Translation of the Phoenician Text: 

"XIX. And I built strong forts at the frontiers on the spots XX. where 
there were evil men, gang-leaders, XXI. none of whom had been subject to 
the House of Mps (dynasty of Asitawadda), XXII. but I, Asitawadda, laid 
them under my feet. 

XXXVIII. And I built this city, XXXIX. and I gave (to it) the name 
Asitawaddija (?), XL. for Baal (in hieroglyphic Hittite, "the Weather-god") 
and the Reset of the birds (in hieroglyphic Hittite, "the Deer-god"?) sent 
me out to build (it)." 

Fig. 48. Sentences XIX-XXII and XXXVIN-XL of the bi- 
lingual text of Karatepe. (Bossert, Archiv Oikntalni XVIII, 3, pp. 
24-25, and Bossert, Jahibuch fur kleinasiariscne Forschune I, p. 

2J2.) ° r 

meanings of words ascertained on the ground of the bilingual 
texts agree with those accepted prior to the knowledge of 
such texts. A part of the bilingual text of Karatepe is shown 
in Fig. 48. Thus, in the case of hieroglyphic Hittite, too, we 
have the unusual experience that the bilingual text is avail- 
able not at the start of the research, but more or less at its 
close, as a welcome corroborating evidence. This is why it is 
incorrect to liken this bilingual find to the Rosetta Stone, 
as one occasionally hears it done. 


In the preceding paragraph, I said advisedly that this 
bilingual text was found "more or less at the close" of the 
research, because the investigation of this not quite so simple 
script is still very far from its real conclusion. Namely, the 
inscriptions of Karatepe demonstrate absolutely clearly that 
the syllabic signs of the Hittite hieroglyphic writing show 
very many variants which still must be determined individ- 
ually before the reading of the inscriptions can be called ab- 
solutely certain. And naturally only a minor portion of the 
many ideograms of this script appear in the Karatepe records 
corroborated as to meaning by the Phoenician translation. 
Also the vocabulary made certain by Karatepe is not all too 
large. It is important from the point of view of comparative 
linguistics that the Indo-European character of the language 
of the hieroglyphic texts and its close relationship to Luwian 
is more clearly recognizable now. Thus Karatepe is not only 
a welcome final confirmation of prewar research, but in 
many respects also a new beginning which warrants the hope 
that we may soon be nearer to a complete clarification of this 
problem, of equal interest to grammatologists and to lin- 





The research of the 19th and zoth centuries achieved its 
three great and, in a certain sense, classical feats of decipher- 
ment on the Egyptian writing, on the cuneiform script and 
_on the Hittite hieroglyphics, and it derived from them the 
most valuable conclusions concerning the linguistic and cul- 
tural history of the ancient Orient. The decipherment of no 
other writing and the translation of no other language can 
quite match the value and significance of these three accom- 
plishments. Yet, also those other decipherments and trans- 
lations have many a feature or element worthy of universal 
attention and interest, and for this reason I shall outline here, 
more briefly, the most important ones. 

In order to present a more comprehensive view, I shall 
classify the following reports in three groups, the first of 
which will comprise those instances where, analogously with 
the heretofore discussed three outstanding decipherments, 
the task involved both the decipherment of an unknown 
writing and the interpretation of an unknown language, the 
second will include the cases where only an unknown writing 
had to be deciphered, but the language written by it was a 
known one, and finally the third group will comprise the in- 
terpretations of unknown languages written in some known 
script (as in the case of cuneiform Hittite) . But let it be 
stated right here and now that simple and clear-cut as the 
basis of the classification of a language into the first one of 


8 6 


Y 1 











v v Y Y 





























Fig. 49, The Lycian and Lydian al- 
phabets. (From Friedrich, Klein asi- 
atische Sprachdenkma'Ier, p. 157-) 

these three groups may seem to be, it is by no means so 
obvious and unambiguous. Namely, this particular group in- 
cludes also Lycian and Lydian, two languages written in dis- 
tinctly alphabetic scripts which are closely related to the 



Greek alphabet (cf. Fig. 49), so that in these two. instances 
we may speak of "decipherment" of scripts with qualifica- 
tions only. Thus, the unlocking of the secrets of the inscrip- 
tions found in these two languages might be regarded also as 
purely linguistic interpretations, which would belong in 
our third group by definition. And the other two decipher- 
ments mentioned in our first group involve, objectively re- 
garded, almost exclusively just a decipherment of the script 
where the translation of the language itself plays no part at 
all or a part of a mere secondary importance, so that these 
two cases might be assigned to our second group as well. 

1. The Decipherment of Other Unknown 
Scripts and Languages 

(a) The Translation of the Lycian Language 
The language of Lycia, situated on the southwestern coast 
of Asia Minor, has been preserved in 150 inscriptions as well 
as in brief legends on coins, although the latter consist often 
of abbreviations and in general are of no help in the trans- 
lation of the language. A few inscriptions have been known 
since the early part of the 19th century, but most of them 
were published only as a result of Austrian expeditions in 
1884 and 1889. We are indebted to the Austrians also for 
the perfect publication by E. Kalinka of the Lycian inscrip- 
tions under the title Tituli Lyciae lingua Lycia conscripti 
(Tituli Asiae Minoris, Vol. I, Vienna, 1901). The bulk of 
the inscriptions are epitaphs, dating from the 5th and 4th 
centuries b.c, which differ from each other but little as to 
contents. Outstanding among the few inscriptions of other 
types is the lengthy text on the Xanthos Stela with its his- 
torical contents, although the language used in it is of such 
an ancient and so strongly dialectal form that our under- 
standing of it is still very imperfect. Another fact to be con- 



sidered is that determinatives, which are such excellent clues 
to the meanings of texts in cuneiform characters, are non- 
existent in alphabetic scripts. The epitaphs are an easier prob- 
lem because many of them are written bilingually in Lycian 
and Greek. The Greek text is, however, often a more 
free rendition of the meaning of the Lycian versio: 

I reproduce here two of these bilingual inscriptions 
ing the Lycian texts in the customary Latin transliteration. 
Only inscription No. 117 of Kalinka's publication contains 
literally identical Greek and Lycian texts (Fig. 50) 

111 J_''.tUlll 

3re or less 
ins, show- 

W nm me Nee* wwae^ w.4® 

2?n M. °*r? />x i\a n.n n n o~t\ /\ •=*> ^ a c- <Tti 

(1) tbeija ; er&wa^tja : me ft : (z) prHnawatt : siderija ;parm[fft]' 
(?) l«b] '■ tideimi [h)rppi : etli ebbi se (4) iadi : ehbi : se tidttmi : 
pHbii- (5) leje :To |iv^t«t t6S' tic- (6) o^wcto Ertipios Dap- 
(a£vo- (7) vxo; m-M eauTui xal ttji yuv[a]- (3) ixl xixl u:a>t 

Fig. 50. Lycian-Greek inscription No. 117 
(From Kalinia, Tituli Lyciae lingua Lycia con- 

The meaning of this inscription, according to the Gree 
text, is: "This monument was made by Sidarios, the son of 
Parmenon, for himself and his wiie and his son Pybiales." 
The word-for-word translation of the Lycian text is: "This 
monument, now he who built (it), (is) Siderija, son 



Paimena, for the own self and the own wiie and the son 

This yields the following translations of words: ebe = this; 
erawazi/a = monument (?); prnnawate — he built; tideimi 
= son; hrppi — for; etli for atli, dative singular of at/a- = 
self; ehbi- = own; se = and; lada — wife. The "me ti" intro- 
ducing a sentence after a prefixed object contains the particle 
me, the meaning of which is more or less "now," and a 
pronoun, ti, which ought to be regarded as a relative one, 
meaning "who," rather than a demonstrative one, denoting 
"this," respectively "him, it." This me ti (rne ti) and the 
alternative expression me ne (me ne) are not translated in 
the Greek text. 
Let us consider now inscription No. 25a (Fig. 51 ) : 
Translation of theGreek text: "Porpax, (son) of Thrypsis, 
nephew of Pyhbates, the Than i.e., a man from the city of 
Tlos, (erected the statues) himself and his wife Tiseusem- 
bra, from Pinara, daughter of Ortakias, niece of Prianobas, 
for Apollo." Translation of the Lycian text: "These statues 
(?), now J he who/ dedicated (them) (is) Kssbeze, son oi 
Ciupsse and nephew of Purihimete, the Tloan, the own self 
and the own wife Ticeucepre, the woman from Pille = 
Pinara in Greek, daughter of Urtaqija and niece of Pri/'en- 

This text adds only two word translations to those already 
known from the preceding inscription, viz. : tuhes = nephew 
or niece; chatra = daughter. The wordatru (accusative singu- 
lar) is a phonetic variant of atlu — the self. These two inscrip- 
tions will suffice to show that the word translations that can 
be gained from the bilingual texts are by no means too numer- 
ous. But the differences in the spelling of the names, chiefly 
in the second inscription, demonstrate furthermore that the 
Greeks were obviously not able accurately to represent the 
alien Lycian sounds and combinations of sounds with their 



©# *T Will t2T:k P@ 

PoPnAE ©FTf l«^ HYP! 


(1) ebeis : iuctdris : m[. . .] (2) tutvete : kssbeze : crupfssehj 
(3) tideimi : se purihime[teb] (4) tubes : tl&ntta : atm : ebb[i] 
(j) se ladu : ehbi : tueucSpri" (6) pilleiini : urtaqijakn : cbatru (7) se 
prijenubebH : tubesn (8) IIopTraJi 0p6|io? II opt- (9) [3<£tou<; dcSeX- 
<pi8o3<; (10) TXctieii; socut&v x.a[l] (11) ttjy yuvxtxa Tiasu- (12) 
^fijjpav ix ITivfitptov (13) 'OpTaxfa $uvaT£p<a) Ilpt- (14) 
avo^a aSeX<ptS7|V (15) 'ATudXXtovt. 

Fig. 51. Lycian-Greek inscription No. 25a, (Kalinka, TituJi 
Lyciae lingua Lycia conscript!. ) 

own graphic resources (in one case, of course, the name in 
question is entirely different in the two languages— Kssbeze 
in Lycian, but Porpax in Greek). And the reader may well 



imagine that the research into the Lycian language en- 
counters not only lexical and grammatical problems, but also 
a whole series of problems of a phonological character. 

This branch of philological research is now over a hundred 
years old, but it has not accomplished too much in all that 
time. This poor record is attributable, insofar as the earlier 
period is concerned, not only to the use of an etymological 
method based without sufficient criticism on Indo-European 
linguistic principles, but also to a lack of sufficient linguistic 
material. It was only after the Austrians had discovered new 
inscriptions that the Frenchmen Six and Imbert and the 
Britisher Arlcwright embarked on a perfect, systematic re- 
search in the 1880's. All three of them attempted to 
clarify the difficulties of the Lycian script and phonology. 
Six studied the Lycian coins in particular, which bore also 
Greek and Iranian names. Arlcwright made a special study 
of the rendition of Greek names in the Lycian inscriptions 
and the representation of Lycian names in the Greek texts; 
the observed inaccuracies yielded all sorts of conclusions con- 
cerning spelling and pronunciation. 

The study of Lycian obtained a second impetus from the 
work of the Scandinavian savants Bugge, Torp, Vilhelm 
Thomsen and Holger Pedersen, about the turn of the cen- 
tury. Laudatory mention is due above all to Thomsen's 
Etudes fyciennes (Copenhagen, 1899), a strongly combi- 
natory analysis of the particles me ne (mS ne) and me ti 
(me ti) which were used in Lycian to introduce sentences. 
The second standstill of the flourishing study of Lycian, soon 
after 1900, was caused by the lack of new, illustrative linguis- 
tic material, for not many new facts can be learned from the 
monotonous epitaphs. 

Also the new departures made since 1928, after the prog- 
ress in Hittitology, by Meriggi and by the old and yet still 


vigorously active Holger Peders 

permanent enlivenment. But tl 

clarified the question of the lin 

The opinions of linguists differed strongly even in the 20th 

century whether Lycian was to be considered a radically 


Ited in any 

east definitely 

uistic affinities of Lycian 

changed Indo-European languag 

possibly Caucasian tongue. In 

decades, the scales have tipped v 

Indo-European character of Lyci 

convincing arguments in his Lyi 

closer affinity of Lycian and the Indo-European languages o 

ancient Asia Minor, and specifi 

known of those languages. 

ome completely alien 

J the past two 
. m favor of the 
Pedersen set forth 

Hjtiitisch* for 



painstaking y gainei 

(b ) The Translation oi tl\\ 

Whereas more than a hundr< 
have not netted more than a mere few 
results, the study of Lydian has grown almost effortlessly ou 
of nothing since World War I. A look into P. Kretschmer's; 
EinJeittmg in die Geschicftre der griechischen Sprache* * will 
show what vague and insufficient notions people still had 
about this language in the last years of the 19th century. How- 
ever, the scant number and monotonous nature of th 
linguistic records preclude very far reaching results in th 
case, too. 

The language of Lydia, situate 
Asia Minor, has been preserved in over 50 inscriptions, foun 
chiefly in the course of American excavations at Sardes, th 
Lydian capital, in 1910-1913. Unfortunately, also in thi 

' "Lycian and Hittite."— Det Kg!. Dan^HIHwUMwhiej Sehiab, 
fflol. Meddelelser XXX 4, 1945. 

** "Introduction to the History of thyjEffcHd U&fcliaMs T-rf^ttii 
1896. — Cf.pp. 384-391. 



case, most of the inscriptions are epitaphs dating from the 
4th century B.C., and considerably similar in content, and 
only a few votive inscriptions and other texts deviate from 
the general pattern. The inscriptions were published in Vol. 
VI, Part II, of the encyclopedic work Sardis (Leiden, 1924) . 
A quick insight into the principles of this formerly com- 
pletely unknown script and language was made possible by 
three bilingual texts. Two of these, two quite short inscrip- 
tions in Lydian and Greek (Fig. 52), though, would prob- 
ably have been of very little help, but the third bilingual text, 
written in Lydian and Aramaic (Fig. 53), presented eight 
well legible lines in each version and thus promised a good 

THITU fltAJtUA-t U11A1 

1MU *tlAT *TJ 
Tl T A t 1A<?AT<US 


Fig. 52. The two Lydian-and-Greei bilingual texts. (From 
Sardis, Vol. VI: Lydian Inscriptions, Part II, Nos. 20 and 40.) 

result. Aramaic, a Semitic language, was the language used 
by the various races of the Persian empire in their mutual 
dealings, and as such it was known in Lydia as well. 

Let the two bilingual texts be quoted here word for word, 
for the sake of clarity (quoting the Lydian and Aramaic 
words transliterated in Latin characters). The first Lydian- 
and-Greek inscription (Sardis No. 20) reads: 

nannai btrkivalis artimaX Not woo; AtovixrixXecu; 'Ap-r^iSt. 

Translation: "Nannas, (son) of DionysikJes (dedicates 
this statue) to Artemis." 

The name Nannas is identical in both languages; the Greek 
dative Artemidi is translated in Lydian by the oblique case 


ArtimuA (nominative: Artimui). The name of the father, 
DionysikJes in the Greek sentence, a derivate from the name 
of the god Dionysos, is rendered in Lydian as Bakiva-, cer- 
tainly derived from the Lydian name of the same god, viz., 
Bakchos. Whereas the Greek uses the genitive case of the 
name, the Lydian text shows as adjective, Bakiva -li, meaning 
literally "the one of Bakiva." 

word for word, the Lydian version of the second inscription 
(Sardis No. 40) is more explicit than the Greek wording: 

Fig. 53. The bilingual text written in Lydian and Aramaic. 
(From Sardis, Vol. VI: Lydian Inscriptions, Part II, Plate I.) 

(i) em taiev asvtl (2) bartaras \ atit (3) 






Translation of the Greek text: "Paitaras (dedicates this 
pillar) to Athene." 

In the Lydian portion, the Greek nominative Partaras is 
equalled hy the Lydian form, of the same pronunciation, 
Bartarai, and the Greek dative case Athenaei by the Lydian 
oblique case Asvil (-1 = -a in ArtimuA.) . It has been con- 
cluded by the combinatory method that esv taSev (contain- 
ing the demonstrative pronoun es-, "this," known from the 
third bilingual text) must mean "this pillar (?)"; | atit, the 
phonetic value of the first character of which is uncertain, is 
presumably a verbal form, meaning, "erect (ed)" or "dedi- 
cate (d)" or something similar. Accordingly, the Lydian 
version means presumably, "This pillar (?) dedicates (or 
dedicated) Bartaras to Athene." 

Far more information can be derived from the longer text 
written bilingually in Lydian and Aramaic, which reads, 
transliterated in Latin characters, as follows: 

[LYDIAN]— [(i*) bor/X X artakjassaX palmXuX dav] (1) foJraX 
isIX bakiliX est mfud els'k {v&nas] (2} laprisak pelak kudkit tst esX 
van[aX] (3) bXtarvod akad manelid kumlffid silukalid akit n[3pis] 
(4) esX mruX buk esX v&naX buk esvav (5) lapirisav bukit kud ut esX 
vSnaX bXtarvo[d] (6) aktin it apis pelXk fensXJfid fakmX artimus 
(7) ibsimsis ariimuk kulumsis aaraX biraXM (8) kXidaX kofuXk piraX 
ptlXk bill v t bapettt. 

/translation/— /(1*) In the year 10 of Artaxerxes the 
king ... in the/ (1) month (?) — on the fifth (??)— Bakillis. 
This stele and this /cavern/ and the wall (??) and the plot 
of land (?) and where (?) at this cavern (3) (there is) the 
antechamber, now this (is) belonging to Mane, the Kumli- 
son, the Siluka. Now who /soever/ (4) this stele or this 
cavern or this (5) wall (??) or where (?) at this cavern (is) 
the antechamber, (6) now who ever damages (?) whatso- 

ever, now to him Artemis (7) the Ephesian and Artemis the 
Koloean courtyard and house, (8) earth and water, {real" 
property and whatsoever (is) his, shall scatter \" 

[aramaic]— (1) b V 1-mrMwn Int X *rtbM mlk] , 
%nb stwn' w-m'rP rdht' (3) 'trt* w-prbr %j '/ sprb ^nb prbrh *k 

(4) V mn J br &*!} srw % ( ? )' w ~ mn K! ' ! stwni K*& '^ W m l rt> 'a 
1-rdbP I-qbl %J prbr l-m l rf (6) ^nb 'br mn xjjbbl 'wjprk mnd'm *k 

(7) >rtmw %jkln> w-'pttj trb$h bjtb (%)qnjnb tjn w-tnjn w-mnd'mh 
jbdrwnh w-jptb. 

/translation/— (1) On the fifth of (the month of 
Markheswan, year 10 of Artaxerxes the king, (2) in Sard 
the fortress. This stele and the cavern, the wall (??),( 3 ) th 
plot of land (?) and the antechamber which at this sepulchral 
chamber (?) (is) the antechamber thereof, now (4) (this 
is) of Mane, the son of Kumli, the Siluka (?). And whoso- 
ever on this stele or (5) the cavern or on the wall (??) 
insofar as the antechamber at (6) this cavern (is), now 
whosoever destroys or mutilates whatsoever^ now (7) Ar- 
temis of Koloe and of Ephesus shall his courtyard, his house, 

(8) his real property, earth and water, and whatsoever (is 
his, to him scatter and to him break up." 

The beginning of the Lydian version of this tat is 
damaged, but it can be restored with certainty on the pattern 
of similar epitaphs. The Aramaic version contains a few 
words which are difficult problems even for Semitists, but 
it is cleai on the whole nevertheless. Thus we obtain a reliable 
translation of the Lydian inscription, too, which in turn 
proves valuable in the translation of monolingual Lydian 
epitaphs, worded mostly along the same pattern. And, a - 
though this cannot be elucidated here in further particulars 
we obtain a glimpse into a small section of the grammar anc 
lexicon of the Lydian language, a glimpse which does give 


some notion of the language as a whole. Many scholars have 
tried to clarify the text written bilingually in Lydian and 
Aramaic, but particular laudatory mention is due to the 
thorough treatment given to it by Kahle, the Semitist, and 
Sommer, the Indo-European and Hittite specialist. 

A certain difficulty was presented by the determination of 
the phonetic values of those Lydian characters which were 
not identical with the letters of the Greek alphabet. Even 
the names occurring in the bilingual inscription have not 
helped fully to establish the values of the still disputed 
characters, and there are a few rarely occurring symbols the 
phonetic values of which are still not known with any cer- 
tainty. More disturbing still was the uncertainty about the 
frequently appearing symbols + andT; the + was at first read 
as h, but now is read as p, whereas y was at first regarded to 
mean u, but is now read as A. The currently accepted reading 
of these characters is influenced by the Lydian word meaning 
"king," which appears in the inscriptions as ^y a I'TA+ (read 
from right to left) and has been recorded by Greek authors 
as palmys. We may therefore transliterate the Lydian word 
as paArnAu-, where \ the Greek lambda, designates a variant 
of I. Thus, classical Greek literature is co-instrumental in this 
case in the deciphering of the Lydian alphabet. 

The question of the affinities of the Lydian language has 
not been fully clarified as yet. Formerly, Herbig advocated 
the theory of a closer relationship with Etruscan, because 
according to Herodotus (1, 04}, the Etruscans were supposed 
to have immigrated to Italy from Lydia. Nowadays, more 
credence is given to the opinion of Meriggi, that Lydian, like 
Lycian, is an Indo-European language, although strongly 
changed by alien linguistic influences. This view is based 
on similarities of the morphological systems, notably of the 
inflection of verbs, on similarities of certain pronouns, eg., 
amu=I or to me (dative), pis — who, pid = what (cf. 


Oscan pis and pid, Latin quis and quid) , napis — whosoever, 
-ad = it (cf. Hittite -at = it), kud = where, as well as on 
certain individual words, such as bira- — house, the Hittite 
equivalent of which is pir— not to mention uncertain ety- 
mologies. In this case, too, the paucity of linguistic material 
prevents the formation of a comprehensive view. 

(c) On the Translation of the Language of Side 

We are told by the Greek historian Arrian (cf. Anabasis 
J, 26:4) that, the city of Side, in Pamphylia, on the southern 
coast of Asia Minor, had a language of its own in the days of 
classical Greece. Specimens of this language, in an unde- 
cipherable script, were known even in the 19th century a.d.; 
they were inscriptions on coins dating from the 4th and 5th 
centuries B.C., but they were so brief that they defied the 
most concentrated efforts of Waddington (1861), Fried- 
lander (1S77 and 1883) and Six (1897) who were unable 
to decipher them. 

A short inscription in Greek and Sidetic was found by the 
Italian excavators Paribeni and Romanelli in Side in 1914 
(Fig. 54). But also that text was too short to enable the 
investigators to decipher it, especially since also the Greek 
portion of it was not legible clearly and unambiguously. Only 


ncynuo hoa5 sHojm 

Fig. 54. The Artemon inscription, written 
in Greek and Sidetic. (From Bossert, Belleten 
14, Fig. 2.) 



in 1949, in the course of new excavations in Side, did Bosch 
find another, somewhat longer and weD legible inscription, 
written bilingually in Sidetic and Greek, which enabled 
Bossert to make a certain headway in the decipherment of 
the Sidetic script. (See Fig. 55.) 

fYt 'AfNK "WKJff* IA^a<A mMit* 
f O N/^JH N^EAr9W0€Sl J HAIL 

V * % .-** \ 

Fig. 55. The Apollonios inscription, written bilingually in Sidetic 
and Greek. (From Bossert, Belleten 14, Fig. 4.) 

I quote here first the Greek text of the second bilingual 

(0 PA]t»M.<£vioc 'AwoXXoSdpoo (2} [t]ou 'AtoXWhEou 
iv&J-Tjxe^ (j) [djx6va t^yS* &«utoD S-edt? Tt£<n. 

The translation of this Greek sentence is: "ApoJIomos, 
(son) of Apolhdoros, (the son) of Apollonios, erected this 
image of himseJf for all gods." 

In this instance, both the dedicator and his grandfather 
are called Apollonios, and also the name of the father, 
Apollodoros, is derived from the name of the god Apollo. 
These consistencies must recur in the Sidetic portion. And 
the Erst and third words of that text do actually appear to 


have the same identical stem, even though their flexional 
(declensional) endings differ, and at least the initial portion 
of the second word is similar. On the ground of this observa- 
tion, Bossert transliterated the Sidetic text tentatively as 
follows: p-u-l-u-n-i?-o?? p-u-r-d-u-r-s?? p-u-J-u-n-i?-o??-a-s.<? 

m-a-s?-a-r-a?-e?-o??-[ ]. 

The results thus obtained encouraged him to re-examine 
the older bilingual text once again, and he decided that it 
had to be read as follows: 

'A*iy^\ ] 'Ar.i&M >A$bpf 

pll-ttf-iti m-a-i-ithi11-*-i-*-i1t 

t-if-t *. tM ? HvhtW 6-*-m- 

The translation of the Greek part is: "Athena Arte- 

mon, (son) of Athenippos, (dedicates) thank-offering." 

The determination of the phonetic values of a number of 
characters represents the entire progress that we have been 
able to accomplish in the domain of the Sidetic inscriptions 
to this date. New results, above all with respect to the lin- 
guistic forms and linguistic affinities, may be expected only 
after the discovery of new inscriptions. 

(d) The Decipherment of the Numidian Script 

Another script and language, the translation of which is 

deserving of being mentioned here, are the products of quite 

another part of the Old World; they are the script and lan- 

_guage of ancient Numidia, which once occupied what is 

today Algeria and Tunisia, in North Africa. 

The Numidians, members of the Berber race, were at first 
economically, culturally, and also politically dependent on 
Carthage; during the second Punic war (218-201 b.c.), with 
Roman backing, they became an independent, unified nation 
under Masinissa, and in the 2nd century b.c. they built a 
mighty empire in North Africa which expanded at the ex- 




pense of the constantly more and more subdued Carthagin- 
ian state, until it became a tributary of Rome, 

The political and cultural independence of the Numidians 
manifested itself also in the. creation of an alphabetic, or 
rather purely consonantal, script which has been preserved to 
our days in more than a thousand inscriptions, most of them 
brief and often casually written epitaphs. Thugga, the mod- 
em Dougga in Tunisia, is the only place where also a few 
official building inscriptions in the Numidian language have 
been preserved, A variant of this script is still being used 
today by the desert tribes of the Tuareg, although with such 
individually and independently developed forms that the 
modern script cannot be used for the decipherment of the 
ancient writing. 

Fortunately, though, there are quite a few bilingual in- 
scriptions available and helpful in the decipherment of the 
Numidian alphabet (Fig. 56). The Numidians wrote their 
inscriptions not only in their own language but frequently 
in Punic or Latin as well, and by this date nine epitaphs 
written bilingually in Punic and Numidian and about fifteen 
in Latin and Numidian are known. Their value must, how- 
ever, not be overestimated because it is a peculiarity of both 
the monolingual and multilingual Numidian inscriptions 
that they are limited to the naming of names and are not 
concerned about stating other facts. Building inscriptions 
list conscientiously the names of all craftsmen who had any- 
thing at all to do with the construction. These many names 
appearing in the bilingual Numidian inscriptions are a wel- 
come aid to help us determine the phonetic values of the 
characters of the Numidian alphabet with absolute clarity 
at least. 

As an example of bilingual texts written in Numidian and 
Punic, I reproduce here the big inscription discovered on a 



} Mpha) 


• — 




J 3 























— -^- 

III t 






































t s 

Fig. 56. The Numidian alphabet. (From Jensen, 
Die Schrift, Fig. 102.) 

temple of Masinissa in 1904 (Fig. 57). I indicate here also 
the transliteration of both texts, in Latin characters, as well 
as a translation, as follows: 

[funic]— (i) t mqdt % bn' W Tbgg l-Msnsn h-mmlkt bn G'jj 
h-mmlkt bn Zlhn b-tjt b-ft l ir S-[mlk] (2) MJsfflstl b-tt Sft h-mmlkt bn 
'fsn h-mmlkt rbt m't Snk bn BnJ w-$ftbn Ngm bn Tnkw (3) msskwj 
Mgn bn Jrttn bn Sdjln w-g^bj Mgn bn S/trb m't bn 'bd'imn b-m[ml]kt 
(4) gldgjml Zmr bn Msnfbn 'bd'Imn h-'dfr] hmlm h-'J? Mql' bn "ijn 
h-mmlkt bn Mgn h-mml[kt] (;) Mm 'I h-mlkt % *tjn bn *nkkn 
Pit w->rt bn Sft bn Snk 




[numidian]— (6) sk[n] . Tbgg . bnjfi? , Msnsn . gldt . w-GJf . gldt . 
w-Zllsn . ijt (7) sbsndb . gldt . sjsb . gld . Mkwsn (8) Sft . gldt . 
tv-Fftt . gldt , mwsnh . Snk . w-BnJ . w-Snk . d-Sft . w-M[gn?] 
(9) w-Tnkw , msjkw . Mgn . w-Jritn . w-Sdjln . g?b . Mgn . w-Sft , 
mw[snbj (10) w-Smn . gldt . gldgmjl . Zmr . w~Msttf . w-Smn . 
gldmsk . M[ql'?J (11) w-Sjti . gldt , w^Mgn . gldt . tnjn . Sjn , 
w-Nkkn . w-PtH . d-Rft} (12) <w~>Sft . w-Snk (funic) w-b-btfm 
H«> bn Jtnb'l bn Hnb'l w-Nftsn bn Sft 

/translation of the punic version/— (1) This temple 
/accusative case/ there built the citizens of Thugga for 
King Masinissa, son of King Gaja, son of the Suffete Zllsn, 
in the tenth year of the reign (2) of Micipsa, in the year of 
King Sft, son of King 'fim. Commanders of the one hundred 
/were/ Snk, son of Bnj, and Sft, son of Ngm /Magon?/, son 
of Tnkw. (3) Msskwj /an official title/ /was/ Magon, son 
of JrStn, son of Sdjln, And gzbj /another title/ /was/ 
Magon, son of Sft, of the commander of the one hundred, 
of the son of King 'Abdesmun. (4) Gldgjml /another title/ 
/was/ Zmr, son of Msnf, son of 'Abdesmun. Leader (?) of 
the fifty men /was/ Mql', son of King 'Sjn, son of King 
Magon. (5) In charge of this work /were/ 'Sjn, son of 'nkkn, 

son of PtS, and AriS, son of Sft, son of Snk (12) And 

the architects /were/ Hanno, son of Jatanbaal, son of Han- 
nibal, and Nftsn, son of Sft. 

Fig. 58 shows two epitaphs as examples of the scantier 
Latin- and Numidian inscriptions. These two specimens 
show at the same time also the customary direction of the 
Numidian script which ran from the bottom to the top (the 
only exceptions being the inscriptions of Thugga in which 
this convention yielded to the Punic custom of writing from 
right to left) . As it can be seen, the Latin versions of both 
inscriptions indicate also the ages of the deceased, 75 years 
in the first and 70 years in the second one, whereas these 

r^ eo o\ o « *t 

ii *■* *» 

c ** *- 

r y ? 

< ~ 1 

** •* £. 

r < c 

* < < 

<£ <cC £" 

^ «*. -r*. 

< ** V 

r* »• v 

V *£■ p 

<r o ju 

* r © 

£^ Q 

- 1- 5 

j-* <■* M 


9 $ 

F* F~ 





^«. J~- j*~ *■ <s 

s. t ^ 

•* T •£• **" 

- f ^^^ ^^ ^»^. 

► t » / 

■* x r c 

£ -^ 5T f 

■2 M £ 

/^ o j^ ■£ 

•^ ^ S **v 

5 ■**■ *o <s 

< -C ^. J^ 

ST 1 * *" 

ID <n c 







R > 

c < 

< Inl 

3i 11 

< L 

< ifi 






K X 

A Jt 

t. II 

na i 

11 Jt 

* ." 

£ in 
















^ , 











A M 

E r 






II i c- 

S ft ^ 

P 41- ^ 

c » r 

2 A > 

* * £ 

1 C © 

1 j)r 


«• s 










J— ■- 

C § 

m c 

jr o 

<j , 

1-1 <, 

» J? 

"3 a 
.S &■ 

* V3 




numbers seem to be missing from the Numidian portions. 
The words msw and mnfcd are Numidian titles. 











- II 




= C 

II 11- 

X I 

u u 

a)= CILVIII 17317. 

h h h 

d h b n 

k w j r 

n s [b] s 
m m n w 

b) = CILVIII j2ao and 17395. 

1 h 

t 1 b h d 

t m j w k 

1 x vr 





Fig. 58. Two bilingual (Latin and Numidian) inscriptions. 
(From Chabot, Recueil des inscriptions Libyques, Nos. 85 and 

It is comparatively easy to decipher the Numidian script 
on the strength of the bilingual inscriptions. Although these 
inscriptions, because of the very nature of the texts, give but 
quite incomplete information concerning the Numidian 
language, they justify certain conclusions in that respect as 
well. Thus, the Masinissa inscription permits us to recognize 
gld (as well as its derivative, gldt) as the Numidian word for 
"king." In every inscription, wherever the father of somebody 
is named, there occurs the short word -w, for "son"; a heavily 


damaged inscription, written bilingually in Numidian and 
Punic, not illustrated in this boolc, makes evident the mean- 
in^ of two designations of occupations, viz.: nbb-n n sqr' 
means "wood-cutters," and nbt-n n zl' means "iron-founders" 
(both plural). Rossler recognized the verbal form eskan 
("they built") in the first word of the Masinissa inscription, 
written as sk[n]. Even this meager linguistic material makes 
it evident that, insofar as the vowelless consonantal skeleton 
of the language justifies a conclusion, the Numidian language 
of antiquity is identical with the Berber spoken, as a second- 
ary language to Arabic, in North Africa today, or in other 
words, that the Berber language has practically not changed 
at all in the course of two thousand years. This is a valuable 
conclusion, reached despite the scantiness of the linguistic 
material available. 

2. The Decipherment of Other Unknown Scripts 
The actual translation of a language played a part of minor 
importance even in the decipherment of the Sidetic and 
Numidian records, but the two cases which we shall discuss 
next go even further and can truly be regarded as cases of 
pure decipherment, because it could be presupposed in each 
case that the language of the inscriptions was a known one. 
There is a case which I even omit altogether, although it 
would belong in this chapter, namely the decipherment by 
Vilhelm Thomsen of the Early Turkish (Tiirfci) runic writ- 
ing (appearing in inscriptions dating from the 8th century 
a.d., found in various parts of Siberia and Mongolia) because 
this script is not of the ancient Near East and thus lies both 
temporally and spatially beyond the range of the present 
book, and also because its decipherment offers hardly any- 
thing new technically. It resembles the decipherment of the 
Ugaritic writing (p. 84 et seq.), in that the deer 

lerer as- 


sumed beforehand that the inscriptions were in a certain 
specific language, and he determined the phonetic values of 
the unknown characters, without any bilingual text to help 
him, but assisted by the knowledge of the structure of the 
language, the way a cryptographer decodes a modern, arti- 
ficial secret writing, 

( a ) The Decipherment of the Cypriote Script 
The Greeks of Cyprus, whose ancestors had settled on the 
island about the turn of the and millennium b.c. to the first, 
did not use the same script as all the other Greeks in later 
historic times, but employed a peculiar syllabic writing, 
written from right to left, which knows syllables only of the 
consonant + vowel type and is very poorly suited to the 

< Mt l! 
















2(0) Z 





































± m 






v y 


2 /£ 
















Fie. 59. The Cypriote syllabary. (From 
Thumb, Handbuch der griecftischen Dia- 
fekte, p. VII.) 


Greek language. In other words, the Greeks of Cyprus had 
left the mother country on the continent before the intro- 
duction of the alphabetic writing and adopted in Cyprus 
the principles of writing of an entirely different race. 

Since Greek looks very strange when written with the 
characters of the Cypriote syllabary (Fig. 59), a few com- 
ments are indicated concerning the writing conventions of 
the Greeks of Cyprus. In writing syllables containing plo- 
sives, the user of this syllabary cannot distinguish lemis, 
fortis and aspirates, i.e., d, t and 6, or b, p and ph, or g, k and 
X, but can write only t, p and k, in other words ta for da, ta 
and 0a, pi for bi, pi and phi, etc. This script does not distin- 
guish, as does the Greek writing in the cases of e and o, be- 
tween long and short vowels, so that, e.g., o-ne-te-ke is written 
for ovidriKt (onctheke) = "he erected" (Attic: dn^w). 
Nasals are not indicated when preceding a consonant, e.g.: 
pa-ta for iravTa (panta) — "everything." Consonants occur- 
ring at the end of a syllable are written with an unpronounced 
auxiliary vowel, mostly e: ka-re for yap (gar) = "for, thus"; 
te-o-i-se for tJtoTs (theois) = "to the gods"; to-ko-ro-ne for 
tA(v) x«p°;< (to/n/ choion) = "the room" (accusative). The 
occurrence of two or more adjacent consonants at the be- 
ginning or in the middle of a syllable is likewise prevented by 
the interpolation of an unpronounced vowel between the 
consonants; when the group of consonants would occur in 
the initial position, the first syllable symbol contains the 
same vowel as the second one, e. g., Sa-ta-si-ka-ra-te-se for 
Zroo-utpiiTijs (Stasikrares, a name), but in a medial position 
the vowel contents of the two syllable symbols are identical 
only if the group can occur in an initial position, too, e.g., 
A'po-ro-ti-te-f for 'A^poSlrat (Aphrodirai) = "of Aphrodite," 
but otherwise the interpolated vowel will be identical with 
the vowel immediately preceding the consonant group, e.g., 



a-ra-ku-ro for &pyhpo (argyro) = "of the silver." Since Cypri- 
ote Greek shows many a deviation from standard Greek lin- 
guistically, too, it is easily understandable that in many in- 
stances no certain reading can be established at all. Thus, a 
word written as a-to-ro-po-se can be read as aw^m* (an- 
thropos) = "man/' but also as &t P otos (atropos) = "immu- 
table," as &T P ovos (atrophos) = "not (well) fed," and finally 
also as &8optto$ (adorpos) = "not dined." 

Inscriptions, coins and medals from Cyprus bearing this 
script began to be known about 1850. It could not be sus- 
pected at first that they were written in the Greek language 
but in an alien script. The scholars of those days did not yet 
know the several bilingual records, in Cypriote and Phoeni- 
cian as well as in Cypriote and standard Greek, which we 
know today, so that the new domain of research remained a 
playground of fantastic hypotheses until 1870 or so. 

R. H. Lang, a Britisher, published the first Phoenician-and- 
Cypnote bilingual insciiption in 1872; that was the inscrip- 
tion which was published by the Semitists in the Corpus 
Insciiptionum Semiricarum I 89, and by the Hellenists in 
Collitz-BechteFs Ssmmlung der griechischen Dkkkt-In- 
schrifren (Gottingen, 1883-1915) as No. 59 (Schwyzer, 
Dialectorum Graecarum exemp/a epigraphica potiora, Leip- 
zig, 1923, No. 680), and which is reproduced in this book as 
Fig. 60. It is a votive inscription addressed by a Phoenician 
nobleman by the name of Baalrom to Apollo of Amyklai in 
the fourth year of the reign of the Phoenician king Milkjaton 
of Idalion and Kition, i.e., in 388 B.C. The Phoenician portion 
is strongly damaged, but it can be completed easily accord- 
ing to similarly worded inscriptions originating from the 
time of the same monarch. The names in the inscription thus 
furnished the fundamental clues for the decipherment, and 
the use of a point for separation mark between words facili- 


tated the task. Since the word "king" occurred in the in- 
scription twice, Lang already suspected that a certain group 
of symbols represented the name and title of King Milkjaton. 

George Smith ingeniously carried the decipherment a long 
step forward almost at the same time. He, too, concentrated 
at first on the names Milkjaton, Idalion and Kition. The great 
number of characters (about 55) suggested to Smith from 
the very start that the Cypriote script was not alphabetic but 
syllabic. This opinion was corroborated in his mind by the 
fact that the geographic names Idalion and Kition did not 
end, as could be expected in view of the Greek form, in the 
same vowel, -i; thus, the -i in the ending of both geographic 
names had to be contained already in the syllabic symbol 
(the Cypriote forms are, reading from right to left, Ti ~ Ke- 
ti-, and £pf = E-da-Ii-). On the other hand, he encoun- 
tered the I of Milkjaton (more precisely, that is, the syllabic 
sign £ Jlif of Mi-Ji-ki-j'a-to-ne) in the word Idalion. 

Also Smith recognized the word which Lang had already 
assumed to mean "king." It occurs in two places, with a dif- 
ferent ending in each, but Smith reasoned that the reason 
was that it was used in two different declensional cases, first 
in the genitive and then— as Smith assumed, erroneously— 
in the nominative. The question as to what language other 
than the Cypriote changed the penultimate phoneme of the 
word meaning "king" in the course of its declension, led 
Smith to the Greek word for "king," basiletis (genitive singu- 
lar: basileos) and he concluded— rather superficially, and yet, 
as we know today, correctly— that the Cypriote inscriptions 
were written in the Greek language. This conclusion decided 
the direction of the further work of decipherment. 

With the aid of the names and of the word basileus, Smith 
determined the phonetic values of eighteen syllable signs 
with a certain degree of probability. With their aid he tried 
then to make headway in reading the brief inscriptions on 




(i) [b-jmmxl-hh y] b-lnt *rb l 4 l-mlk . Mlkjtn [mlk] 

(2) [Ktj w^ijlsml] \ yjt* w-jtrf . 'drtn , B'lr[m] 

(3) [bn *bdmlk l-H}j l-Rsp Mid . k im l ql-j brk 

"^ 12S the xth ^ ot the month of yl in the year 4 of King 
MillcjatOD, /king/ T s 

"{2) /of Kition and Idalion,/ This (is) /the image/ which gave 

and erected our Lord B alroro, 
"(3) /the son of 'Abdimilk, for/ his /god/ ReJef of Mkl, for he 

listened to his voice. Hebless(ed?) (him.)" 


(a) In the syllabic script of the original: 

<i) [i to-i I te?-ta?-r a ?-te?-i? \ pe-tt-i] \ pa-si-le-vo-st I mi-it-ki- 

ja-te-no-se J ke-ti-o-nt \ ka-ttl-ta-U-o-ne \ pa-si-le-u- 
(z) I -ot-tot-stf J ta-ne e-pa-koj-me-na-ne \ to pe-p a -me-ro-tit I 

ne-vo-so-la-ta-se ] to-na-ti-ri-ja-ta-ne \ to-U ka-ie-st-ta-se I e 

va-na-xe \ 

(}) [Pa? -a? -la? -roi -moist? \] A-pi-ti-m-H-ko-ne \ to A-po- 

lo-m I to A-mu-ko?-h-i j a-po-i vo-i I ta-se e-u-kt,-la?st 
(4J [ej-pe-tu-ke i tu-ka-i j a-^-ta-i J 


(b) Modem Greek transliteration: 

(t) fl(v) rot I TETixpTSt? J /tret] I PaaiXg/oq [ MiXxtji-B-ovo ; 

KEt(6v I kkt (?) "ESaXfov I p«oiAeii- 
{2) [o(v)tog? I t«v i7raYo3tievav to tce([x)ttoi(a£pov [ ve/oaxatas 

tov I a(v)Spija(v)Tav | t6(v)Ss_KotTeaTaCTe | 6 /iiva? I 

(3) [BaaXpojioq? | ] 6 'AJ18i[mXxov | Tot 'A7v6X(X)ovi | 'ApjxXoi 

&p' 01 /ot I tSc. su);oXaq 

(4) [4]xetux s- Uy) "^X " I AC**2t- 
"(1) /In the fourth year when/ King Milkjaton over Kition and 

Edalion rule — I I 

"(3} /-d/, on the last day of the five-day period of the /intercalary/ 

days, did this statue erect the Prince 
"(3) /Ba'alrom,/ the (son?) of 'Abdimilk, for Apollo of Amyklai, 

after he for himself the desire 
"(4) had accomplished; in good luck I" 


Fig, 60. Phoenician-and-Cypriote bilingual inscription 
Idalion. (From R. H. Lang, Transactions of the Society of Bibli 
cal Archaeology I, 1872, Table following p. 128.) 

the medals which presumably contained nothing but names 
—Greek names, as Smith now assumed. He identified in 
them, for instance, the male names Euagoras, Euelthon, 
Stasioikos, Pythagoras (rather Philokypros!), Stasiagoras 
(rather Stasikypros/) etc. 

The failure of Smith to continue his decipherment, which 
he had begun quite correctly on the whole, must be attributa- 
ble to his scant knowledge of the Greek language. Thus 
Birch, the Egyptologist, published the much -needed explicit 
proof to the effect that the Cypriote language, according to 
the evidence of its linguistic structure, could not be an 
Semitic or Egyptian tongue, but only a variant of Greek. Th 
Greek of the inscriptions, as Birch read them, seemed still t 
be remarkably barbaric and corrupt. This impression was dm 
partly to the incorrect way in which several syllabic signs were 
still being read— a syllabic writing represents always a more 
difficult problem for the decipherer than an alphabetic script, 
just because of the greater number of individual signs in a 


syllabary-and partly also to the ignorance of the decipherers 
as regards the peculiarities, now being gradually clarified of 
the quite odd Cypriote Greek dialect. 

Johannes Brandis, the German numismatist who regret- 
tably died at such an early age, identified the word meaning 
"and/' and by this important discovery be made a better 
analysis also of the monolingual inscriptions possible. Ac- 
cording to the Greek lexicographer Hesychos, the Cypriote 
word for "and" was kas, instead of the kai of standard Greek, 
and Brandis found this kas in the inscriptions. He deter- 
mined also the phonetic values of further syllabic signs. Since 
the Greek character of the language was already an estab- 
lished fact, the decipherers were no longer restricted to using 
solely the few and scant bilingual inscriptions in their work, 
and they were free to determine the phonetic values of the 
still questionable syllabic signs from the tolerably intelligible 
sentences and phrases found in monolingual inscriptions, on 
the basis of their knowledge of the Greek language. 

The long monolingual bronze inscription of Idalion, pub- 
lished by Moriz Schmidt, a Hellenist of Jena, in 1874, proved 
to be particularly important. Schmidt succeeded with the 
aid of this text in establishing the phonetic values of a great 
number of further syllabic signs, by the combinatory method, 
and also in definitely determining the basic features and 
principles of the Cypriote writing, viz., that it consists solely 
of syllabic signs which may represent a single vowel or a con- 
sonant followed by a vowel, but no other combination. 

Deecke and Siegismund brought the decipherment essen- 
tially to its conclusion. Their important accomplishment 
consisted in the elimination of the last difficulties in the 
reading of the Cypriote script, by establishing the phonetic 
values of syllabic signs denoting syllables beginning with the 
phonemes /// or /w/. The subsequent linguistic study of 
details need not concern us here any further. 


The somewhat perfunctory conclusion of George Smith 
that the Cypriote inscriptions were written in Greek proved 
to be no less correct than the likewise hasty conclusion of 
Hans Bauer concerning the West Semitic character of the 
Tj saI itic language. Moreover, also those scholars are being 
proven right who claimed that this script, so unsuited to the 
Greek language, had been borrowed from another race. A 
few inscriptions from Cyprus which have been known since 
1010, are written in the same Cypriote script, but in a non- 
Greek language which is being regarded, with full justifica- 
tion, as the still unknown language of the non-Indo-Euro- 
pean autochthonous population of Cyprus. We know even 
an inscription written bilingually in that unknown language 
and in the Attic Greek of the 4th century B.C. This inscrip- 
tion still appears to be untranslatable, despite an attempt by 
Bork, because of the shortness of the Greek portion in par- 
ticular. But the names of persons can be identified, even 
though showing certain discrepancies, in the two versions, 
and they prove that the phonetic values of the syllabic sym- 
bols are the same in the alien language as in the Cypriote- 
Greek inscriptions. 

(b) On the Decipherment oi the Proto-By bh'c Script 

I still have to discuss here the decipherment of the ancient 
inscriptions from Byblos, a city on the northern Phoenician 
coast. They are written in a peculiar script, partly resembling 
pictograms and partly cursive in shape; the French Dhorme, 
one of the scholars who deciphered the Ugaritic writing (cf. 
p. 84) attempted its decipherment in 1946, but his efforts 
have still not produced any clear and definite results. 

Byblos was one of the oldest cultural centers of Phoenicia; 
its close relationship with Egypt originated as far back as 
the beginning of the 3rd millennium b.c. and was still main- 
tained in the 2nd millennium b.c. The oldest known inscrip- 



tions of Byblos written in the Phoenician alphabetic script 
date from about 1000 B.C., but written records had been 
made in that city, obviously, prior to that date, too. Egyptian 
inscriptions appear on sarcophagi from the early 2nd mil- 



3*np w ^ y 

£3 2M ^^ 









CTty? ai 



Fie. 61. Stone inscription in the Proto-Byblic script. (From 
Dunand, Byb/ia Grammata, Fig. 26.) 

lennium; according to the evidence found in the archives 
of EI-Amarna, Egypt (cf. p. 3), Byblos used the Babylonian 
cuneiform script and the Akkadian language for international 
communications. In addition to these alien, and perhaps 
just occasionally used, systems of writing, the special script 
to be discussed here can be observed on two well preserved 
bronze tablets, a damaged stone tablet and three fragments 
of stone tablets, as well as four bronze spatulae, all of which 
are to be regarded on archaeological grounds as dating from 
the early part of the 2nd, if not the late part of the 3rd mil- 


lennium B.C., and which for that very reason I would like to 
name Proto-Byblic, i.e., Early Byblos script. All the inscrip- 
tions were published by M. Dunand in his book, Bybh'a 
Grammata (Beyrouth, 1945), on pages 71-135. (Cf. 
61 and 62.) 

ft \\y^ l^Ta^o ;> ^%^ fx* 
*, «* =t ¥ n a!+i 3? * 4 « * v ^-*r^ 


4- ^ f PI 


Fig. 62. Proto-Byblic bronze tablet 
Grammata, Fig. 28.) 


nand, Byb 

The prospects of a successful decipherment are not ver 
favorable, for the following reasons: There exist no bilingu; 
texts or similar references. The number of the texts is n 
high. Only one of the two undamaged bronze tablets runs a 
long as 41 lines, the other one is only 15 lines long, the spatu 



lae bear very short texts, and even the biggest one of the 
damaged stone tablets bears only 10 lines of writing. Only on 
one of the four spatulae are the words separated by vertical 
strokes, in the other inscriptions the words are not separated 
Moreover, the number of symbols is quite high; 1 14 different 
symbols have been distinguished to this date. This circum- 
stance very soon warranted the conclusion that this was a 
syllabic, and not an alphabetic, script. And as I have stated 
repeatedly in the present book, a syllabic system of writing 
always confronts the decipherer with greater difficulties than 
an alphabetic script, even when the circumstances are more 
favorable than in this particular case. 

Nevertheless, Dhorme proceeded confidently to decipher 
the script soon after the publication of the Byblm Grammata. 
He started out from the premise that the language of the in- 
scriptions must be a Semitic tongue-Phoenician, to be ex- 
act. This hypothesis is by all means a very plausible one, for 
we cannot find one single non-Semitic element in the thor- 
oughly known history of Byblos. 

For the decipherment proper, Dhorme proceeded from 
the end of Tablet c shown in Fig. 62, where the seven times 
repeated sign seemed to represent the numeral 7. He con- 
cluded that a date was indicated there, and he read the entire 
15th line tentatively as b-snt 7, "in the year 7," "in the 7th 
year." The reason for which Dhorme transliterates vowelless 
Semitic words in the style of the later Semitic alphabetic 
writing, is not that he considers the Proto-Byblic script a 
consonantal system of writing. In view of the high number 
of characters, he is convinced that this is a syllabic writing, 
with specific signs for the syllables ba, bi, bu, fe, If, Su (pos- 
sibly also for ab, ib, ub, etc. ) . But he wanted to be content at 
first with obtaining the consonantal "skeleton" which is al- 
ways the foundation of all understanding in the Semitic lan- 
guages. Since in following this method, he is bound to come 


across several signs for the same consonant sound, he trans- 
literates for the time being mi, ma, ma, m, n 2 , ns, ru, etc., 
hoping to be able some time in the future to replace these no- 
tations by ma, mi, mu, na, ni, nu, an, in f un, or similar syl- 

Using the letter values b, S, n and t, obtained from the 
group of signs presumed to represent a date, Dhorme found 
the group ni (the delimitation of which as a separate word, 
however, in either direction is by no means certain! ) in the 
first line of the same inscription, and he believed that it repre- 
sented the word nM, "metal, copper," since the text ap- 
peared on a copper tablet. With the aid of the knowledge of 
the sign for h, he was able to read the word mzbh, "altar," in 
lines 6 and 10, and the knowledge of the m in turn enabled 
him to read, in line 14, before the designation of the year, 
the designation of a month, viz., btmzt, "in (the month of) 
Tarn muz," containing a second sign for z, which he trans- 
literates as zi. The designations of the month and year can 
logically be" expected to be preceded by the designation of 
the day of the month, written by a numeral. The s.5 is now 
completed to MS, "six(th)," and the signs following it 
turned out to be jm-m, "day" (with two new signs for ml), 
so that the full date reads; b-M$ jm-m b-tmzi b-Snt 7 = "on 
the sixth day in (the month of) Tammuz in the year 7/' 

Untiring work and a constant revision of his findings en- 
abled Dhorme finally to obtain a result which he could pre- 
sent to the Academy of Paris on August 2, 1946, He pointed 
out that he had searched neither for the name of Byblos nor 
for the words for "god" or "king," nor for any narrative of 
conquests or pious deeds, and that the report of an engraver 
on what he and his co-workers did in decorating the temple 
had revealed itself to him unbidden. Also the long Tablet d 
revealed a similar content to Dhorme. 

Dhorme thus obtained a transliteration of the Proto-Byb- 



lie inscriptions, or rather of the consonantal "skeletons" of 
the words. The missing vowels must still be determined to 
make the decipherment complete. That should not be a very 
difficult task, for we are passably familiar with the diachronic 
picture of the Phoenician vowel system, in word-stems and 
flexional endings alike. Thus, e.g., a substantive the last con- 
sonant of which is m, would have to end in mu in the nomi- 
native singular, in mi in the genitive singular, and in ma in 
the accusative singular, etc. For the time being, I am still 
not sure to what degree or extent it is possible to substitute, 
as mentioned above, more accurate syllabic values for 
Dhorme's transliterations, and for this very reason I shall not 
express any judgment concerning the correctness of his de- 
cipherment as yet. Let it remain an open question also 
whether or not Dhorme's separation of the individual words 
is correct. At any rate, Dhorme can always support his views 
by pointing out the fact that his transliterations yield texts 
that make sense (their graphical, lexical and grammatical 
features may, of course, still be in need of rectification) as 
well as the fact that in particular the transliteration of the 
date of Tablet c which served as his point of departure is 
corroborated reciprocally by various arguments. Dhorme's 
decipherment of the Proto-Byblic script is certainly more 
likely to gain universal recognition than the attempt of 
Grimme to decipher the Sinaitic writing, which for that very 
reason I shall discuss later only, among the undeciphered 
scripts (pp. i5C)etseq.). 

5. The Translation of Other Unknown 


For consideration of space, only a few typical examples 
will be discussed in this section, in order to cast light on the 
problem as such. 


(a) On the Translation of Etruscan 
The facts that Etruscanhas been in the focus of more in- 
tense interest than many another language discussed in this 
book, and that more effort has been expended on its decipher- 
ment than on that of the other languages, are indubitably the 
result of its geographic location in the center of Italy, home- 
land of an ancient civilization. That land, of such impor- 
tance in the development of European civilization, seems to 
have been so strongly under the influence of the Etruscans in 
the early epoch of its history that a study of the Etruscan 
language always appeared to be desirable and attractive. The 
results are, however, by no means commensurate with the 
efforts devoted to this task; despite the existence of an im- 
mense literature dealing with the translation of Etruscan, 
this author still finds it difficult to state even today whether 
or not the translation of the language may now be said to 
have been successfully accomplished. Considerations of 
space make it absolutely impossible to discuss with any 
semblance of thoroughness all that has been undertaken in 
order to unlock the secrets of Etruscan, and the most I can 
do in this book is to give an outline of a few principal features. 
The Etruscan inscriptions known to exist number more 
than 8,000, the oldest ones of which date back to prior to 600 
b.c. Most of them are, however, quite brief, some no more 
than mere fragments, and chiefly epitaphs, consisting usu- 
ally simply of the name of the deceased and perhaps a brief 
indication of his age or the public offices held by him. Only 
a few sarcophagi, such as the sarcophagus of Pulena and that 
of Alethna, bear longer inscriptions, and only a few inscrip- 
tions can be called really long, such as the tile from S. Maria 
di Capua, belonging to the 5th century, with about 300 
words on it, and the more recent Cippus Perusinus (Perugia 
cippus), with about 120 words. The texts different from the 
epitaphs are strongly in the minority as for contents, too; let 

i 3 8 


there be mentioned the two leaden tablets, hearing impreca- 
tions of Volterra and CampigHa Maritima, furthermore the 
bronze Jiver oi Piacenza which once was an instrument of 
fortune-telling, and also two dice on which the numerals 
from one to six are written out in words. In addition to the 
"epigraphic" monuments, there are the linen wrappings in- 
scribed with a religious text which were found on an Et- 
ruscan mummy in the museum of Zagreb in 1892 and have 
been known to scholars as the "Zagreb mummy wrappings." 
The more than 1500 words appearing on them constitute 
the longest Etruscan text known. The small number of 
longer and thus linguistically more informative Etruscan 
texts is one of the reasons for the meagerness of the results 
heretofore achieved by the attempts at the translation of 
this language. 

The texts themselves offer the linguists very little to go by. 
The known bilingual inscriptions consist of just a few quite 
short and uninformative epitaphs in Latin and Etruscan. 
For instance, Inscription 378 of the Corpus Inscriptionum 
Etruscanim (= XI 1855 of the Corpus Inscriptionum Lati- 
narum) consists of the Etruscan text "V. Cazi C. clan" and 
of the Latin phrase "C. Cassius C. f . Saturnmus/' the latter 
meaning "G(aius) Cassius, G(aius') s(on), Saturnmus"; 
hence, the Etruscan word clan in the Etruscan portion, in 
which the name Saturnmus is missing, means "son," and this 
meaning is confirmed by other inscriptions. Also isolated 
translations of Etruscan words occurring in Latin literature, 
e.g., aisar for "god " furnish only minor help. 
^ The Etruscan glosses accompanying pictorial representa- 
tions of subjects from the Greek mythology, however, do 
show the Etruscan forms of a number of mythological names, 
some in an Etruscanized Greek form, as e.g., Apulu (Apol- 
lo), NeSuns (Neptune), Hercle (Hercules), Axmemrun 
(Agamemnon), Alcsentre and other variants (Alexander), 


Pecse (Pegasus), Sersipnei (Persephone), etc., others in 
Etruscan "translation," as Tim'a or Tins* (Zeus. Jupiter), 
Tuian (Venus), Fufhms (Dionysos), Tmms (Mercury)^ 
etc. While the first one of these two groups in particular 
supplies many data concerning Etruscan phonology, these 
names are hardly of any use to one who tries to gain an in- 
sight into the language as a whole. 

Thus the only practicable way of translating the inscrip- 
tions remains, in the main, to proceed from the inscriptions 
themselves. Now then, the co-existent use of two distinct 
methods becomes even more strongly evident in the field of 
Etruscan than in the translation of other languages, viz.: the 
combinatory method, which seeks to clarify the meaning of 
the texts on the basis of clues contained in those texts them- 
selves, and the etymological method which tries to accom- 
plish the purpose on the strength of phonetic similarities 
between the language studied and some known language. 
The possibilities of learning the secrets of Etruscan by the 
combinatory method are very scant, though. It is under- 
standable therefore why the etymological method has again 
and again been advocated by various scholars in the study of 
Etruscan in particular, and why attempts have been made at 
interpreting this language on the strength of phonetic simi- 
larities to the most different languages, such as Basque and 
Caucasian, Proto-Germanic, Greek and Proto-Indo-Euro- 
pean in general, and even Sumerian, to name just a few of 
the languages selected for this comparison. Above all, indi- 
vidual research linguists would again and again consider 
Etruscan an Italic language and most closely cognate with 
Latin and Osco-Umbrian, and they would try to translate it 
on the ground of its phonetic similarities to these languages, 
obviously misled by a number of unquestionably Italic loan- 
words in Etruscan. All these attempts at a translation, often 
published in bombastic terms, are entirely worthless and 



tend only to make laymen skeptical about meritorious inter- 
pretations of other languages as well. 

In the study of Etruscan, the scantiness of the clues makes 
it especially difficult to make any headway by the combina- 
tory method, which can grope its way forward step by slow 
step only. It compares, for instance, the bronze liver of 
Piacenza with similar fortune-telling "livers" of clay used by 
the Babylonians and the Hittites, and on the strength of this 
comparison it determines what names of deities appear on 
the Etruscan liver, which include also usil, "sun," and tiv, 
"moon," two appellatives the knowledge of which helps in 
the clarification of other connections. Thus, the epitaphs 
often indicate not only the name of the deceased, but also 
his age in numerals. Therefore, if we find the expression avih 
x tivrs y, we may safely assume that tiv meant not only 
"moon," but also "month," and consequently, avil must 
obviously have meant "year." 

Inscriptions appearing on implements and beginning with 
the word mi, followed by names, e.g., mi Qancvilus Fulnial, 
have pointed at the conclusion, based on the analogy of Italic 
inscriptions on such objects, that mi must have meant "this" 
(possibly also "I") and is to be translated as this (is)— re- 
spectively I (am)— Tanaquil Fulnia's. The circumstance 
that a woman's name followed by puia frequently appears 
after names of men in the epitaphs, suggests that puia must 
mean "wife." Therefore, Vel Sefae puia-c is translated as 
"Vel Sethre and (-c) wife," and ©anxil Ruvfi puia Arnflal 
Aleflnas as "Tanaquil Rufia, wife of Arnth Alethna." The 
words Jupuce or svalce in the statement concerning the age 
of the deceased are considered as verbal forms meaning "(he 
or she) lived" or "(he or she) died," and avils LX lupuce is 
translated consequently as "he lived 60 years" or "he died 
(with) 60 years," svalce avil LXVI as "he lived 66 years," 
etc. Accordingly, also the frequently occurring word amce 


(showing the same ending, -ce) is regarded as a verbal form 
and translated as "(he or she) was"; thus, RamOa Matulnei 
sex Marces Marumal puia-m amce SeOres Ceisinies is taken 
to mean "Ramtha Matulnei was (amce) the daughter (sex) 
of Marcus Matulna, but (-m) the wife of Sethre Ceisinie," 
etc. These few examples demonstrate at least how the com- 
binatory method is able to extract not only lexical meanings 
but also grammatical rules from the inscriptions. 

An entire literature has been written about the six num- 
bers, 0u, zal, ci, sa, max, hud, appearing on the dice, since they 
indubitably represent the spoken forms of the numerals from 
1 to 6. The only difficult, still not finally solved question is 
their proper sequence. They were probably arranged simi- 
larly to the system used on most ancient dice, so that two 
opposite numbers always totaled seven, but which of the 
numbers should be taken to mean "one"? It so happens that 
also the indications of the age of the deceased in the epi- 
taphs contain numbers written out in letters, namely those 
mentioned above as well as three more simple numbers, 
cezp, sem£ and nux<£ (which must mean "seven," "eight", 
and "nine") as well as their derivatives, i.e., tens as zaOxum 
(from zal), cialx (from ci), muvalx (from max?), fealx 
(from sa), cezpalx (from cezp), sem^alx (from sem<l>). The 
frequency of the occurrence of the tens seems to warrant 
conclusions as to their numerical values: We may expect 60 
and 70 to occur often, but 80 and especially 90 but seldom. 
However, since the scholars are, as stated before, still not in 
agreement as to the various details, a mention of this in- 
teresting problem must suffice at this place. 

Naturally, the "Zagreb mummy wrappings," the longest 
relic of the Etruscan language, attracted most of the investi- 
gators, especially since in 1932 infrared irradiation revealed 
and permitted to be read also the completely faded portions 
which had been assumed to be totally illegible. The fre- 



quent occurrence of the names of deities indicates that the 
text is of a religious character, but it differs strongly from 
the inscriptions both in vocabulary and in phraseology and 
confronts the combinatory method of research with insur- 
mountable obstacles. Yet, Olzscha succeeded in advancing 
these studies considerably in 1934-1936, He found that the 
wrappings had originally constituted a single scroll, several 
meters in length, and the text had been written on it in 
columns from right to left. The gaps in the preserved text 
could be partly filled in because long passages are repeated 
in the text. This grouping of the text was an essential pre- 
liminary to Olzscha's conclusions. He established, further- 
more, that these analogous portions of the text could be 
divided into a number of verse-like units oi identical 
structure, the arrangement of which recurs regularly in 
the individual sections. This finding opens up a compre- 
hensive vista of a long, coherent text, whereas formerly only 
arbitrarily detached pieces could be viewed. The appearance 
of the name of a deity introducing each verse makes one 
think of prayers. The prayers are separated from each other 
by short sections which, on the strength of the vocables 
occurring in them, Olzscha considers ritual directions con- 
cerning sacrifices. Directions for sacrifices and prayers are 
found to alternate similarly on the Iguvfne Tables, the long- 
est document extant in the Umbrian language, and similar 
Roman prayers are contained finally in Cato's treatise de 
re rustica. Although the Umbrian and Roman prayers do 
not exactly blend with the Etruscan ones so as to form what 
we could consider a "bilingual record," they still constitute 
close parallels to them and can be used, with proper caution, 
in the translation of the Etruscan text of the mummy wrap- 
pings. At any rate, they encouraged Olzscha to undertake a 
new, complete translation of this peculiar literary relic of 
the Etruscan language. I do not propose to claim that he 


found the right answer in each and every particular, 
work seems to have opened up a new inroad into this hereto 
fore so unyielding language. 

Let us hope that future research will be able to so 
the enigma of the linguistic affinities of this language which 
today still stands isolated. Recent discoveries disprove Her- 
big's theory of a close affinity between Etruscan and Lydian.* 
In the opinion of this author, though, also Olzscha is mis- 
taken in considering Etruscan a cognate language 

(b) On the Translation of Other Languages of 
Ancient Italy 

After the preceding discussion of Etruscan, a short g 
at the translation of ancient Italic languages in general does 
not seem to be out of place. 

The credit for the translation of Osco- Umbrian, the c 
relative of Latin among the Indo-European languages 
ancient Italy, is due chiefly to the linguistic science of the 
19th century. This is a domain where the etymological 
method appears to be not only permissible, but in fact the 
only one that can be expected to produce a result. The reason 
is that Osco-Umbrian is almost as close to Latin as Dutch is 
to German; not only have the two languages very many 
words and grammatical features in common, but with a si: 
larity of all phases of communal and private lifejalip I 
style and phraseology of the inscriptions parallel each other 
completely. Let this be demonstrated by the following speci- 
men of an Oscan inscription from Pompeii which I accom- 
pany here also by its Latin translation in order to make 
especially evident the close parallel of the two languages: 

* Cf. p. 109. 

i 4 4 


[oscan]— (1) V. Aadirans V. eftiuvam paam (2) vereiiai 
Pumpaiianai tn'staa- (3) mentud deded eisak eitiuvad (4) 
V. Viinikiis Mi. kvaisstui Pump- (5) aiians tnibum efcak 
tumben- (6) nieis tanginud upsannam (7) deded isidum 

[latin translation]— (1) V. Adiranus V. (filius) pecu- 
niam quam (2) iuventuti Pompeianae testa- {3) mento 
dedit, ea pecunia (4) V, Vinicius Mr. (filius) quaestor 
Pom- (5) peianus domum hanc conven- (6) tus sententia 
faciendam (7) dedit, idem probavit, 

[enclish translation]— (1) Which money V{ibius) 
Adiranus (son of) V(ibius) , (2) to the youth of Pompeii by 
testa- (3) ment did give, from that money (4) did V(ibius) 
Vinicius, (son of) M(ara), quaestor of Pompeii, (5) this 
house according to the meet- (6) ing's decision to be built 
(7) caused; he approved (it). 

Whereas Latin and Osco-Umbrian are close linguistic 
relatives and their speakers shared the same culture, the 
situation is quite different with respect to the language of 
the Veneti of northeastern Italy. Venetic is not just an Italic 
dialect, but perhaps a separate branch of Indo-European 
which does have features in common with Italic, but also 
with Celtic, Germanic and Illyrian. A clear insight into the 
Venetic language is impeded by the fact that the entire 
linguistic legacy of the Veneti consists of a number of brief 
inscriptions on utensils and tools which follow more or less 
the same pattern, viz., "I (am) X's (tool)" or "X presented 
me to Y." 

In the course of the translation of the Venetic inscription 
exo Voltfxenefi ve$o$, the meaning of exo is determined, of 
course, etymologically on the basis of its phonetic similarity 
to the Latin ego, etc. ("I"); this conclusion, however, is in- 
fluenced not only by that phonetic similarity but also by the 


objective, combinatory consideration that inscriptions on 
tools found also in cother regions of Italy and the Old World 
usually begin with a 'l" and are phrased as if the tool itself 
were speaking. The same objective consideration applies 
when the Venetic inscription mexo Vhuxiia zonasto 
Rehtiiah is translated etymologically as "me (mexo) pre- 
sented (zonasto) Vhuxia to Rehtia." The words exo and 
mexo remind the translator of the German ich and mich (I 
and me), rather than of the Latin ego and me; the word 
zonasto is a Greek-style s-aorist of the verbal stem zona- = 
Latin dona-re, "to present." In these cases, therefore, the 
translation is not obtained by the method of phonetic 
analogy alone, but lby the etymological method, supported 
by objective, combinatory considerations. 

Whereas this sort of etymological interpretation still 
moves more or less on firm soil, the following example leads 
one easily onto the territory of hypotheses. Ancient Calabria 
in the extreme sou.theast of Italy (between Brindisi and 
Taranto— not identical with modern Calabria opposite 
Sicily)— was inhabited once by the Messapians who spoke an 
Indo-European language, cognate perhaps with Illyrian, pre- 
served in a number of brief inscriptions as well as two 
longer ones. One of them, found in the city of Basta, begins 
with the words tfototria marta pido vastei basta veinan aran, 
which Krahe as "Tutoria Marta /a woman's name/ 
handed over /= sohd?; pido/ to the city (vastei) /of/ Basta 
her (veinan) field t(aran)." He considers vastei the dative 
singular of a word cognate with the Greek astu, "city"; he 
looks at pido as a -verbal form, taken to be a root aorist, 
*(e)pi-do-t, "gave ower," from the Indo-European root *do- 
in the Greek di-do-oii, "I give"; he regards ara- as a substan- 
tive belonging to thie Latin arare ("to plough") and mean- 
ing "field" (like the Latvian ara for "field"), whereas he sees 
veina- as a possessive pronoun, "his," derived from the Indo- 



European *sueino- and identified with the Gothic seina- 
and the German sein. The impression gained here by an un- 
biased observer is that the translation of the words is based 
on the phonetic similarities alone, without sufficient objec- 
tive clues, and that this translation may but not necessarily 
must be right. For instance, von Blumenthal translated the 
same phrase, breaking it into words somewhat differently, 
tfotoria marta pidova steibasta veinan aran, as "The dead 
(marta) Tutoria bequeathed her field by a testament" /con- 
sidering pidova. the instrumental of *(e)pi-dova-, "surren- 
der," and steibasta an s-aorist of a verb analogous with the 
Latin stipuJari, "to have something stipulated." This ex- 
ample demonstrates how the etymological method can lead 
to different results by different interpretations, and how both 
results and interpretations can still claim a certain degree of 
plausibility. But only one of them can be right at best, and 
all of them may be wrong. This uncertainty is all the more 
unpleasant because Krahe makes use of this very interpreta- 
tion for a foundation on which to build such important com- 
parative-linguistic conclusions as the analogy of the forma- 
tion of the pronouns my, your, his in Messapian and the 
Germanic languages. It should be very understandable that 
this author cannot conceal his scruples about such a far- 
reaching etymological interpretation. 

(c) On the Translation of Phrygian 

In contrast to the linguistic disunity of ancient Italy where 
a number of closely related Italic and more remotely cognate 
Indo-European languages as well as the non-Indo-European 
Etruscan were spoken, Greece was a closed linguistic unit, 
for even though its inhabitants spoke several distinct dialects, 
those were still dialects of one and the same language. We 
find no parallel in the Greek speech area, above all, to the 
relationship between Latin and Osco-Umbrian. But there is 


at least one language which investigators have considere 
more closely related to Greek and to a certain extent also 
interpreted etymologically on the basis of Greek, viz.: the 
language of the Phrygians who lived in the interior of Asia 
Minor. The Phrygian language has been preserved in about 
twenty-five Old Phrygian inscriptions, written about the 7th- 
6th centuries B.C., with a slightly modified variant of the 
Greek alphabet, and in nearly 100 Neo-Phrygian inscriptions, 
originating from the era of the Roman Empire and written 
in standard Greek characters. But Neo-Phrygian was only 
occasionally used as the language of an entire inscription; 
generally, the inscription is written in Greek, with an im- 
precatory formula added to it in Neo-Phrygian. 

This imprecatory formula is also the only Phrygian text 
which we understand with any degreee of certainty. In this 
connection, we might even speak of a kind of bilingual rec- 
ord, inasmuch as the average wording of the Neo-Phrygian 
imprecatory formula, ios ni semoun knoumanei kakoun 
addaket etittetikmenos eitou corresponds to the occasionally 
encountered Greek formula, rh &i raiinji daMtiw mKbv woa- 
vwfiati Karripa.pi.vQ$ fjrw, i.e., "but whoever will inflict evil to 
this sepulchral chamber, shall be accursed." Accord- 
ingly, the Phrygian words can be interpreted as follows: 
ios — a relative pronoun (Indo-European *io-s, Greek hos); 
ni = an emphatic particle added to that pronoun; semou(n) 
= dative singular of the masculine and neuter form of the 
demonstrative pronoun meaning "this" (Indo-European 

h- and *ko-, Slavic si- with the masculine-neuter dative 
singular semu, etc.); knouman = "grave" or "monument on 
a grave"; kakoun — "evil" (loan-word from the Greek fcakds, 

evil," or originally cognate with this word, unknown in 
other Indo-European languages?); ad-daket = "he inflicts" 
(ad = Latin ad, "to"; daket = "he places, puts " cf. Greek 
ti-the-mi, "I lay, place," aorist e'-the-ka); eti-ttetikmenos — 



"accursed" (passive perfect participle ending in -menos as in 
Greek, in which connection again it is an open question 
whether this identical ending is to he ascrfbed to linguistic 
borrowing or to original cognateness) ; eitou = "he shall he" 
(Greet esto) or "he shah go" (Greet ltd)? 

The interpretation of the imprecatory formula is feasible 
also without the etymological method. Also the Old Phrygian 
inscription ates [ artiaevais | afcenanolavos [ midai | Javaltaei I 
vanaktei j edaes, written clearly and with the words separated 
from each other, appears to be more or less clearly under- 
standable as "Ates, (son) of Arkiaevis (?) placed 

(it) for Midas Lavaltas, the prince, " but the etymological 
method is still not completely eliminated in the translation 
of vanaktei, "to the prince" /Greet (v)anax, "prince"/ and 
of the verbal form edaes, "he placed" /s-aorist of the root 
*dhe- in the Greet fi-the-mi, "I place"; Hittite dais", "he 

The combinatory method is all too often compelled to 
admit its own incompetency when it is applied to longer in- 
scriptions, more ancient or more recent ones alike, espe- 
cially when the words are run together without any indica- 
tion of where one word ends and another one begins, which 
is the method usually observed in the Neo-Phrygian texts, and 
encountered in some Old Phrygian inscriptions as well. In 
such cases, the etymological interpretation based on phonetic 
and morphological analogies with other Indo-European Ian- 1 
guages has become more or less the standard procedure. The 
investigators who apply it can cite precedents, viz., the trans- 
lation by the etymological method of Old Persian on the 
basis of the cognate Sanskrit language, or that of the Osco- 
Umbrian inscriptions with the aid of the knowledge of 
Latin. As to details, though, opinions diverge no less widely I 
than in the instance of the Messapian inscription cited as an 
example on page 145. For instance, the word-group attiadei- 


ton appearing in the Neo-Phrygian phrase etittetikmenos 
attiadeiron (in which the word erittetikmenos is known, as 
explained above, to mean "accursed") is broken up by some 
linguists as Atti ad-eitou, "he shall hurry to (the god) Attis" 
(ad-eitou = Latin ad-ito = "he shall hurry to"), but by 
others as at Tiad eitou, "he shall go to Zeus" (at = Latin 
3 d = "to"; Tiad for *Tian-de = "to Zeus"), so that the 
former translate the phrase as "accursed, he shall hurry to 
Attis," hut the latter as "accursed, he shall go to Zeus." As a 
second example, I mention the Neo-Phrygian word group 
otuvon'etei which R. Meister breaks into the two words 
otuvoi vetei and translates as "in the eighth year," assuming 
that vetei = the Greek (v)6ios = "year," and otuvoi = the 
Latin octavus = "eighth" (with a change of the -cr- by as- 
similation to -t(t)-, analogously to the Italian otto < Latin 
octo, "eight") . But O. Haas breaks the same group into 
three words, Otu voi vetei, and he translates them as "Otys 
to his (voi) relatives (Greek etes)." 

R. Meister, in particular, went to extremes in translating 
Phrygian on the basis of Greek analogies. He evidently re- 
garded the Phrygian language as a strongly altered dialect of 
Greet. Let his method be illustrated on two sentences from 
the Old Phrygian inscriptions appearing on the tomb of 
Arezastis: zostututa?i? | a?e?mnoz? | akenanolavos, trans- 
lated as "who is begotten from the blood of Atenanolas," 
viz.: zos = a relative pronoun ( = Greet hos); tututai (?) 
for *t£tuttai from the Greek teucho, "I create"; aemnoz 
(?) = genitive for ablative singular from the Greet halma, 
"blood." The second sentence is the artist's note (?) appear- 
ing at the end: ataniz en | kurzanezon f tane lertoz, "Atanis 
chiseUed this in the Gordian's (territory) " viz.: lertoz = 
"he chiselled," on the analogy of the Greek *{e)lertose; 
tane = "this" (Greek ra-de) ; Kurzanezon = genitive plural 



of an adjective of citizenship derived from the name of the 
city of Gordion. 

But also the other investigators have not yet succeeded in 
achieving any unambiguous and convincing explanation of 
the Phrygian inscriptions. While O. Haas does not follow 
one-sidedly the method of translating on the basis of Greek 
analogies, even he proceeds very arbitrarily in the separation 
of words and in his etymological interpretations, and in most 
cases he fails to produce proof of the latter. Thus, diounsin 
is interpreted, without any objective proof, as "the living 
one" and traced back to the Indo-European *guMont-si-n, 
but at the same time also augoi is translated as "living" and 
explained as a derivative of *aiugof {from Old Indie ayu-, 
"life"). The word argousi is called a loan form of the Greek 
archousi, "(at) the archons," and the word isgeiket a loan 
form of the Greek eischeke, "he has received." This ought to 
suffice to demonstrate that the translation of Phrygian is 
still very much in its infancy, and that the general public 
will still be wise to regard it with a little skepticism. 




Now that we have become acquainted with quite a number 
of decipherments, we are in the position to derive from them 
a series of basic principles relative to the decipherment of 
unknown scripts and languages in general. These basic con- 
siderations could have been presented at the beginning of 
this book as well, but if the uninitiated layman bad met them 
there, they might easily have impressed him as somewhat 
dry and abstract, whereas given here, they represent a practi- 
cal summary of the facts elucidated on the preceding pages, 
and thus they might not be entirely without value for future 
decipherments either. 

To begin with, I must state once again the fact, self-evi- 
dent and trite as it may be, that the decipherment of any un- 
known script or language presupposes the availability of 
some clue or reference; nothing can be deciphered out of 
nothing. In those cases where one has absolutely no possi- 
bility available to link the unknown to something known, the 
amateur can give free rein to his imagination, but no real or 
lasting result can be accomplished. 

Furthermore, we must distinguish three different types of 
decipherment, which at the same time represent three dif- 
ferent degrees of difficulty. The task at hand may comprise 
the translation of an unknown language written in a known 
script, as in the cases of cuneiform Hittite and Etruscan, or 
the language may be a known one but written in an unknown 




script, like Cypriote Greek, and finally, there are cases in- 
volving an unknown language written in a likewise unknown 
script The last variant of the problem is, of course, the most 
difficult one. This is why, for instance, the translation of 
cuneiform Hittite cannot be considered being on the same 
level with the decipherment of the Egyptian hieroglyphics 
or of cuneiform writing in general. 

When some unknown script is to be deciphered, a number 
of preliminary questions of fundamental importance can be 
clarified in most cases even before the real work is begun. The 
direction of the writing can be recognized in most instances 
by the position of the blank end of the last line of the in- 
scription (cf. the Hittite hieroglyphic inscription shown in 
Fig. jo, and the inscription written bilingually in Phoenician 
and Cypriote, shown in Fig. 60) . The decipherer can decide 
whether or not the words are separated by systematically re- 
curring strokes, periods, colons, or other similar signs, and 
he can draw conclusions as to the difficulty of the decipher- 
ment accordingly. 

Above all, the number of the written symbols usually war- 
rants a conclusion as to whether the script is alphabetic, a 
pure syllabary {as in the Cypriote) or a mixture of ideo- 
graphic word-signs and syllabic signs (like the cuneiform 
writing or the Hittite hieroglyphic script). A script consist- 
ing of less than thirty signs will presumably turn out to be 
alphabetic; the probability of its decipherment is higher 
than that of a more complicated system. Scripts containing 
fifty, a hundred, or even several hundred different symbols 
may justifiably be regarded beforehand as more or less com- 
plicated syllabic systems of writing, perhaps employing also 
word-signs, and their decipherment can be expected to in- 
volve more considerable difficulties. It was decided rather 
accurately, even before the decipherment of one single cunei- 
form character, that the trilingual inscriptions of the early 


Persian kings consisted of three parallel versions, one (the 
Old Persian text} written in an alphabetic script, a second 
one (the Neo-Elamite version) couched in syllabic symbols, 
and a third (the Akkadian) one presumably written with 
ideographic word-signs. 

The decipherment proper is facilitated most by a bilingual 
text ("bilinguis"), i.e., an inscription in which the text 
written in the unknown language or script is followed or 
preceded by its translation in some known language or script 
The preceding discussions of the various decipherments 
demonstrate that such bilingual texts (in fact, even some 
trilingual texts, i.e., trilingms) are fortunately available often 
enough. Neither the Egyptian hieroglyphics nor the Baby- 
lonian cuneiform writing could have been deciphered with- 
out a bilingual text, and the Hittite hieroglyphic writing is 
the sole instance where the unknown script, and partly also 
the language, was deciphered without the aid of a bilingual 
text and the decipherment was later confirmed by a subse- 
quently discovered bilingual text. 

Regardless of which of the three types of decipherment is 
involved, the decipherer will first be on the lookout for 
names of persons, cities, countries, etc. in the known portion 
of the bilingual text and then try to find their equivalents in 
the unknown part. A prerequisite of this procedure is that 
the names actually be identical or similar in the two versions, 
and in most instances they are, too. The rare case that some 
name is entirely different in the two versions (cf. Italian 
Ragusa = Croatian DubTovnik) occurs also in the languages 
of the ancient Orient, for instance, the Urartaean city Ardini 
was called Musasir in Akkadian. The decipherer finds cor- 
roborative evidence of the correctness of his way of reading 
the written symbols in particular in the favorable case when 
the same written symbol occurs more than once in the same 
name or in more than one name; I mentioned the hieTO- 



glyphic Hittite names Tuwanuwa, Cuiguma and Amatu (cf 
pp. 92 et seq.)i. Names are the most important, often the 
only means of gaining the first foothold in the field of the 
reading of an unknown script; but they are important also 
when the problem at hand is purely one of linguistic inter- 
pretation, because they are instrumental in the grouping of 
the words which in turn is valuable with respect to the trans- 
lation of words and the determination of their grammatical 
functions. Besides the names, the titles are important, as in 
phrases like, e.g., "X, king of the land of Y," etc. 

In the absence of bilingual texts, the decipherers loot for 
other media of help. I mentioned above (p. 53) how 
Grotefend had made good use of his knowledge of the names 
of the Early Persian kings as recorded by Herodotus to open 
up a way to the understanding of the Early Persian royal 
inscriptions, and what an important part was played in that 
connection by the fact that Hystaspes, the father of Darius, 
had been no king. And the mention of the names of North 
Syrian cities and their rulers in the war reports of Assyrian 
kings was helpful in the first groping attempts at the reading 
of Hittite hieroglyphic inscriptions. Legends on tools, sus- 
pected to mean "(Ax of) Y" or "This is the tool of Y," were 
valuable in the decipherment of the Ugaritic script as well 
as in the translation of Etruscan. The clues which permit one 
to make headway in the unravelling of the secrets of ex- 
tinct languages and scripts are quite varied and cannot be 
classified under rigid, inflexible rules. The detection and ex- 
ploitation of these possibilities depends chiefly on the indi- 
vidual skill of the decipherer, and every project of decipher- 
ment may present new surprises in this respect. It may, 
however, happen also that no point of attack can be found 
at all and all the efforts of the decipherer are in vain. 

When the problem at hand is to transfete an unknown lan- 
guage written in the cuneiform characters deciphered about 



a century ago, the investigator is in a particularly favorable 
position because cuneiform writing itself paves the way to 
its understanding by making use of a combination of differ- 
ent scriptural elements, i.e., ideograms, syllabic symbols and 
determinatives. If an Assyriologist who knows only Akkadian 
looks at a cuneiform Hittite or Urartaean test, he will im- 
mediately recognize a number of familiar elements, chiefly in 
the ideograms and determinatives. The determinatives will 
enable him, even without knowing the language of the in- 
scription, immediately to classify certain words as names of 
men or women, of gods or goddesses, geographical names, 
names of professions, etc.; the ideograms will enable him to 
recognize inflectional (declensional or conj'ugational, as the 
case may be) endings, from which he can draw conclusions 
to help him both in the translation of an individual inscrip- 
tion and in the recognition of general grammatical facts. The 
great extent of the aid rendered by the ideograms and de- 
terminatives in cuneiform writing can be properly appreci- 
ated only by one who has personally attempted to contribute 
to the understanding of a language written in cuneiform 
characters as well as of another language written in a differ- 
ent script, say Lydian or Etruscan. In the language written in 
cuneiform characters, a number of linguistic facts become 
immediately evident; on the other hand, uncertainty con- 
tinues to prevail even about the most simple questions, such 
as whether a given word is a proper name or a common noun. 
Also the Lycian text on the Stela of Xanthos (cf. p. 104) 
would be easier to translate if the script had included de- 
terminatives, etc., to facilitate its understanding by the 

One must always try to interpret an unknown language, 
whether written in cuneiform script or by some other system 
of writing, by the combinatory method, i.e., on the ground of 
objective clues and conclusions. This is, of course, often a 



difficult undertaking when the unknown language is not 
written with cuneiform characters, especially when the num- 
ber of available inscriptions and records of the language is 
small, or when the texts are particularly brief and uninforma- 
tive. It is therefore psychologically comprehensible, and yet 
methodologically wrong, that, e.g., Etruscologists were all 
too willing to be guided by phonetical coincidences and at- 
tempted to translate unknown words etymologically accord- 
ing to the meaning of similarly sounding words in known 
languages. I can never warn my readers too often or too em- 
phatically against this procedure, for its basic principle is as 
if one wanted to translate the Latin Jaus, "praise," on the 
ground of its assonance with the German Laus or the Eng- 
lish louse, or the Central American Mayan word catz, "poul- 
try," on the analogy of the English cats, or the modern 
Greek nay, "yes," on the analogy of the English dialectal 
nay (meaning no) . The etymological method is permissible 
to a certain extent when applied to closely related languages, 
but even in those cases it must go hand in hand with objec- 
tive considerations. And phonetic similarity plays tricks even 
in the cases of closely related languages: The German Gift 
means "poison," and the German verb bekommen does not 
mean "to become," but "to receive." In any case, the ex- 
plorer of an unknown language must never declare flatly that 
"Words that sound alike or similar in two languages, mean 
the same thing in both," but he must endeavor to present 
objective arguments at least to support any such conclusion 
based on a phonetic similarity. 

It is, of course, especially difficult, and in many cases prac- 
tically impossible, to decipher an unknown script without the 
aid of a bilingual text containing names. Many a failure to 
accomplish the decipherment of some writing is directly 
attributable to the lack of bilingual inscriptions or other ref- 
erences. But the possibility of deciphering a script without 


bilingual texts, too, is demonstrated by the example of the 
Hittite hieroglyphic writing. Thus, it might be more prudent 
to say when discussing a script which has defied all attempts 
to decipher it, that science has not yet been fortunate enough 
to discover a suitable point of departure for the decipher- 
ment. This principle is another conclusion warranted by a 
study of the earlier and later stages of the decipherment of 
the Hittite hieroglyphics. 

Combinatory methods are the proper procedure in the 
decipherment of scripts, too, but this is another instance 
where the proper procedure has not always been duly fol- 
lowed. As the amateur is all too prone to declare in the 
translation ot a language that "Whatever sounds alike or 
similar, also means the same thing," in the decipherment of 
a script people are likely to commit the methodological error 
of stating that "Whatever looks similar in two scripts, means 
the same thing." Thus, Hrozny* committed a fundamental 
error in his attempts to decipher the Cretan-Minoan writing 
(cf. p. 165) as well as when trying to decipher the Indus Val- 
ley script (cf. p. 170), viz., he tried to determine the mean- 
ings or values of the unknown characters on the ground of 
indiscriminately applied analogies with symbols of similar 
(and occasionally not even similar) appearance of the Hittite 
hieroglyphic script, but also of the South Arabian and other 
scripts. This is the same amateurish method which is applied 
by a layman completely ignorant of Russian writing who 
looks for the first time at a Russian text and sees in it a num- 
ber of familiar characters, fully identical with the correspond- 
ing letters of the Roman alphabet, such as A, K, M, O, etc., 
and is therefore convinced that also the Russian P (read as 
our R) stands for the Roman P, the Russian C (read as our 
S) for the Roman C, the Russian H (read as our N) for the 
Roman H, etc. This error is pardonable in a layman, but it 
should not be committed by a scholarly investigator. The 

i 5 8 


fundamental question whether the Indus Valley script is an 
alphabetic, syllabic or ideographic writing, is not raised by 
Hrozny at all. Future decipherers ought by all means to guard 
against such basic errors. 


I have stated {p. 157) that many a failure to decipher a 
given script is attributable to the simple fact that the investi- 
gators have not yet discovered the suitable point of depar- 
ture. In conclusion of the present book, I shall present a few 
examples of such unsuccessful attempts to decipher un- 
known scripts, but these are merely examples, without any 
claim to completeness. Occasionally I shall be in the position 
also to point out a few facts which may explain why it has 
been impossible to accomplish the decipherment as yet. 

1. The Sinaitic Script 

I include the so-called Sinaitic script, even at the risk of 
exposing myself to contradiction, among those writings the 
decipherment of which still has not been accomplished. This 
script has been preserved in a few brief inscriptions, dis- 
covered by Flinders Petrie, an archeologist, in the ancient 
copper and malachite mines of Mount Sinai, particularly 
near the ruins of a temple of the Egyptian goddess Hathor, in 
the winter of 1904-1905; Petrie determined, on archeologi- 
cal grounds, that the inscriptions had been made about 1 500 
b.c. The very carelessly and disuniformly written signs look 
partly like Egyptian hieroglyphics, but there are not more 
than thirty-two distinguishable individual signs, so that it is 
quite logical to suspect that they are the symbols of an alpha- 
betic system. We are, however, still so uncertain about the 




shapes of the individual symbols that it is often doubtful 
whether two signs of very similar appearance are to be re- 
garded as two different signs or as variants of one and the 
same character. After all, we do know an alphabetic script 
consisting of characters closely resembling those of the Egyp- 
tian writing, viz., the Meroitic script (cf. pp. 26 et seq.), 
although it is true that the Meroitic characters originated 
much later, in the Roman era. 

The British Egyptologist Gardiner worked on these in- 
scriptions during World War I. Since it was impossible to 
read them on the basis of Egyptian ideographic and phonetic 
signs, and because a few signs resembled early Semitic let- 
ters, Gardiner assumed that this script might represent a 
preliminary stage to Semitic alphabetic writing, more exactly 
a link between Egyptian hieroglyphics and the Semitic al- 
phabetic system. In other words, he fell into the error men- 
tioned above (cf. p. 159) of letting himself be guided by 
similarities in shape and appearance, and despite the muti- 
lated state of the inscriptions, he felt justified in reading the 
group of symbols shown in Fig. 63 as the Semitic b% "lady," 
and interpreting it as a Semitic designation of the Egyptian 


a(\} + 


Frc. 63. Alleged Ba'aJat in Si- 
naitic script. (From Jensen, Die 
Schrift, Fig. 183.) 

goddess Hathor. Gardiner's theory was that the creators of 
this script had proceeded as follows: They took over the 
Egyptian symbol |— | (pr = house) in the shape of [7"],but 
since the Semitic word for "house" was baith (beth), they 








































































































.^ - 

























t s 



\ J 






■1 s 

^ ft 























c| £ 
























•0 + 




























1 s 












'1 ? 











o« + 

'.I s 






















































e*- ^ 
































a? lis 


















ie signs of the Sinaitic 
Die Schrift, Fig. 187. 

script. (From J 
1 I 





used it as a phonetic symbol of the letter b (the Semitic 
beth); likewise, they used the Egyptian ^ (/rt = eye) to 
represent the Semitic 'ayin = eye, as well as the letter ' (the 
Semitic 'ayin), etc. 

Gardiner's hypothesis was welcomed enthusiastically by 
many scholars, because the missing link between the Egyp- 
tian and Semitic scripts had been the target of a long search. 
Also many a book on the history of our alphabet discussed 
the Sinaitic script, rashly and without scruples, right at the 
beginning, as an early stage of that history and development. 
This view was, however, opposed by renowned investigators, 
among them Hans Bauer, whose name I mentioned in con- 
nection with the Ugaritic records, and who devoted a great 
deal of effort also to the study of the origin of the Semitic 
alphabetic system of writing. But the theory of Gardiner was 
advocated all the more decidedly by H. Grimme. He not 
only compiled the complete table of signs shown in Fig. 64, 
based on newly discovered inscriptions and better photo- 
graphs of those already known, but he also proceeded to 
transliterate the inscriptions (originally written without 
separation of the individual words) into Semitic characters 
and thus to translate them. 

This author not only regards these translations as pre- 
mature, but also the decipherment of the signs as fully un- 
founded. We do not even know what symbols represent 
variants of the same letter, and thus we are not at all sure of 
the very number of the characters, nor are the phonetic 
values assigned to the symbols based on any safe foundation. 
A firmer soil will conceivably be reached eventually, but for 
the time being it seems to be advisable to be skeptical about 
the decipherment as such, as well as about its application in 
drawing grammatological conclusions of far-reaching im- 

2. The Cretan-Minoan Script 

No success has been accomplished as yet in the decipher- 
ment of the different systems of writing discovered on the 
island of Crete as witnesses to its pre-Greek culture.* The 
overall view of the development of writing in Crete is com- 
plicated in that we know a Cretan pictographic writing used 
on stone seals in a primitive and in a more advanced stage 
(about 2000-1600 B.C.), consisting of about 140 signs (cf. 
Fig. 65) , and two linear scripts, preserved on clay tablets and 
seals, designated Linear Class A and Linear Class B, respec- 
tively. Linear Class A, consisting of 85 signs {cf. Fig. 66), is 

Fig. 65. Cretan pictographic writing. 
(From Jensen, Die Schiift, Fig. 75.) 

Fie. 66. Cretan "Linear Class A. : 
(From Jensen, Die Schrift, Fig. 79.) 

* For the latest developments, see Appendix, p. 17;. 



not restricted to any specific part of the island, whereas 
Linear Class B (cf. Fig. 67), the symbols of which number 
73, is found only in one part of Crete, viz., in Knossos, but 
inscriptions in Linear Class B have been found also in Pyl os 
on the Greek mainland. We know vessels found in Thebes' 


Fro. 67. Two inscriptions, from Pylos, in Cretan- 
Minpan Linear Script B. {From Peruzzi, Aportaciones 
a la interpretation <fc Jos textos minoicos, p. 80.) 

Orchomenos, Eleusis and Tiryns, inscribed with symbols of 
special forms which may well be called Linear Class C, Since 
pictographic symbols are of a minor importance in attempts 
at decipherment, the term "Cretan script" is generally used 
to designate Linear Classes A or B, unless expressly stated 

The decipherment of these inscriptions is still in its initial 
stage. I shall mention here by name only two especially 
capable scholars specializing in this field: The indefatigable 
Johannes Sundwall of Finland, and the American Alice 
Kober who unfortunately died so very young. They painstak- 
ingly collected the different signs, investigated their occur- 


rence individually or in groups, and then they made careful 
attempts to determine their character, whether ideograms or 
syllabic signs, but still without trying to read them, for the 
time is still not ripe for taking that important step. 

The more impetuous investigators are not satisfied by 
this progress. In particular, Hrozny, who had successfully 
laid the foundations for cuneiform Hittite research and then 
worked, even though with questionable success, on the hiero- 
glyphic Hittite records, felt qualified for reading the Cretan 
script, too. Unfortunately, he resorted to a method which 
1 decried as amateurish (cf. p. 157), viz., he compared the 
Cretan symbols to Hittite hieroglyphics of similar appear- 
ance and simply assigned the Hittite phonetic values to the 
Cretan characters. In addition to the Hittite hieroglyphics, 
he made use, to a lesser extent, also of the Egyptian hiero- 
glyphics as well as of the South Arabian and other scripts to 
determine the phonetic values of the Cretan symbols. Un- 
fortunately, all these efforts, undertaken at such great ex- 
penditure on reproducing Cretan symbols in print, etc., are 
mere flights of fancy which are devoid of all objective founda- 
tion. Not only all the readings of geographic names and 
names of deities are to be regarded as fictitious, but so are 
also all the conclusions concerning linguistic features com- 
mon to Cretan and hieroglyphic Hittite. 

Sittig' s "decipherment of the oldest syllabic writing of 
Europe" is based on more serious and objective considera- 
tions than the work of Hrozny. Sittig investigates the struc- 
ture of the language statistically and compares, above all, 
the initial and final sounds of Cretan words with those of 
the non-Greek Cypriote language, for the two languages and 
scripts were suspected in the past to be related. Sittig reasons 
that if a Cretan symbol is used structurally exactly like a 
graphically similar symbol of the Cypriote script (i.e., if both 
appear, for instance, only at the beginning of a word or never 



as the final symbol of a word, etc.) , te is justified to regard 
the two symbols as representing the same sound, too. Thus, 
after laborious preliminary studies, he finally ends up once 
again with the comparison of the external shapes of signs, 
which is a very deceptive foundation. In his attempts at read- 
ing entire texts, he does not use Cretan tablets, but inscrip- 
tions from Greece, in which he tries to identify the Greek 
language. Whatever he thus reads out of-or into— the in- 
scriptions, is partly Etruscan (Tyrrhenian) and partly Greek, 
but the latter language does by no means always appear in 
the form to be expected according to the teachings of dia- 
chronic linguistics. But aside from these facts, the salient 
point is in my opinion that it is wrong and misleading simply 
to take the attitude that graphically similar Cretan and 
Cypriote signs represent the same sound, disregarding 
weighty statistical data. This author is therefore justified in 
calling also Sittig's attempt at a decipherment unsuccessful. 
Among the scriptorial relics of Crete mention must be 
made, finally, of the Phaistos Disk, a clay tablet found in the 
fortress of Phaistos in 1908. It is covered with pictographic 
symbols which are arranged in distinctly separated groups 
and are unlike any symbols known from Crete or anywhere 
else. A frequently occurring character is the likeness of a 
human head with a plumed head-dress which reminds the 
observer of Asia Minor, so that it has been conjectured that 
this isolated relic (Fig. 68) was imported from southwestern 
Asia Minor. Also this isolated and relatively brief text can- 
not be deciphered without points of reference. Nevertheless, 
the decipherment has been attempted several times in this 
case, too, but always unsuccessfully, even though some at- 
tempts-like ihe most recent one by E. Schertel-donned the 
scientific trappings of a mathematical method, operating 
with frequency curves for the various symbols. Schertel 
claims that the inscription on the disc is in an Indo-European 


Fig. 68. The Phaistos Disk. (From DEringei 

language closely cognate with Latin, 
syllabic and alphabetic writing, and that the text is a dou 
hymn to Zeus and the Minotaur. These numerous positive 
Bndings are in striking (and for that very reason suspicious) 
contrast to the meager results gained from the Cretan texts 
by Johannes Sundwall and Alice Kober, At any rate, how- 
ever, those meager results are the rnore trustworthy ones. 

3. The CaeianJ I 

st J 

The script and language of ancient 
southwestern Asia Minor, between Lydia and Lycia) 
another enigma that still continues to baffle investiga 
The known Carian inscriptions now number almost 
hundred, but most of them are quite short. For a long time, 
only very few texts found in Caria itself were known, and the 
bulk of the known Carian records consisted of short scrawls 
left in Egypt by Carian mercenaries, often very unelearly 
scribbled and also not always faithfully reproduced 
publications. More recently, however, the numbei 
scriptions discovered in Caria proper has been incre 
fact, three of them contain Carian and Greek texts a 



expected at first to turn out to be the bilingual inscriptions 
needed for the decipherment of Carian. But closer study 
destroyed this hope; in these inscriptions (cf. Fig. 69) either 
the Carian or the Greek text or both are so badly damaged 

Fig. 69. Inscription in Greek and Cari- 
an. (From Robert Hellenics VIII, Table 

that it has still been impossible to identify the Carian equiva- 
lents of the names appearing in the Greek texts. Thus, even 
if these bilingual inscriptions did consist of Carian and Greek 
versions of the same text, they still remain useless for the 
decipherment of Carian for the time being. 

Nor are there any other clues available now to permit any 
progress in the decipherment. The uncertainty is increased 
by the circumstance that there is often nothing to indicate 
where a word ends and another word begins, even in the 
only longer inscription extant, that of Kaunos (Fig. 70) . We 
are not even sure about the nature of the Carian script as 
yet. Bork held that it was mixture of an alphabetic (Greek) 


/o i-l H v • F CMTFT7 
rMrAXn4-+ArP + EForMkS') 
fXAFHA + OXMDtP-AP'fro^^^X] 

^Tfci Y+onopr p-ra- mtax pmm< 

fi^rP ^AfrPVOYOFX^TPHAWifrf 



- - y 4y £££E±P t *++ Mf* 

Fig. 70. Carian inscription of Kaunos. (From 
Bossert, /ahrbuch fur kleinasiarische Forschung, 

'■ P '. !! " > 

and a syllabic (Cypriote?) writing, but nowadays Bossert is 
more inclined to call it a purely alphabetic system, not much 
more different from the Greek script than are the Lycian 
Lydian alphabets. 

So Carian still leaves an unknown quantity in Asia Minor, 
an area so rich in discoveries and positive results in the field 
of decipherments. Let us hope that it will no remain 
known forever. 

4. The Indus Valley S 

Another system of writing about which we are still com- 
pletely in the dark is the script appearing on a large number 
of seals and small copper plates found in Harappa (in the 
Punjab region of India ) and lately in particular in Mohanjo- 
Daro in the Indus Delta. A few isolated specimens of this 
Proto-Indic script, customarily referred to as Indus Valley 
script or simply as Indus script, reached even Babylon, obvi- 



ously in the course of trading, and Babylonian archeology 
succeeded in dating these inscriptions as originating about 
the middle of the third century b.c. 

There seems to exist no possibility of deciphering this 
script either. The extreme brevity of the many inscriptions 
is a further obstacle. One of the hypotheses formulated is 
that these inscribed objects were brief seals or stamps used 
for administrative purposes (cf. Fig. 71). Nor is there any 
agreement concerning the number of the symbols; some 
scholars count as many as 400 different symbols, others not 
more than 150. Most investigators are of the opinion that 
this script is a mixture of ideograms and phonetic charac- 
ters, perhaps syllabic signs. 

The only serious work preparatory to a decipherment has 
been done by Meriggi; while he tries to interpret the alleged 
ideograms pictographically, he does not postulate any pho- 
netic values. A more daring course was taken by Hrozny who 
tried to apply the same method to the Indus Valley script as 
to the Cretan writing i.e., to decipher it on the ground of 
graphic similarities to the Hittite hieroglyphic writing. In 
this manner he claimed to deduce a great many phonetic 
values, which, however, as well as the conclusions he drew 
from them concerning a kinship of the Froto-Indians with 
the historically far younger "Hieroglyphic Hittites," are to 
be regarded as mere flights of fancy. 

And one must regard likewise as a flight of fancy the at- 
tempt of Hevesy to establish links between this ProtoTndic 
script and the writing discovered on Easter Island, on the 
extreme eastern fringe of the Polynesian archipelago, basing 
his reasoning on the indubitably striking similarities in the 
shapes of characters of the very ancient Indus Valley script 
and those of the Easter Island writing, known merely from 
the last centuries, or rather from our very own era. 
The Easter Island writing has been preserved on a number 








ti$U-;. \Vm& \ ■ \ ;fef ;. 

Fig. 71. Ptoto-Indic seals. {From Diringer, 
Tie Alphabet, Figs. 41 and 43.) 

of wooden tablets, discovered on that island in and after 
1870. Unfortunately, by that time none of the natives was 
able to read them, so that here we have an instance of an 
"undeciphered script of our own era." The pictographic sym- 
bols number about 500. They are so arranged on the tablets 
that the symbols in the alternate rows are always in inverted 
positions with respect to the symbols in both immediately 
adjacent rows, so that the reader had to turn the tablet up- 
side down every time he finished reading a line (cf . Fig. 72 ) , 
As a matter of fact, it is very doubtful even to what extent 
the use of the words "reader" and "reading" is justified here, 
ie., to what extent, if at all, these tablets are written records, 

1 7 2 


ideographically or phonetically written, of historical events, 
religious hymns, etc. It is possible that they were simply 


Fig. 72. The writing of Easter 
Island. (From Jensen, Die Schrift, 
Fig. 275.) 

mnemonic devices in which each individual symbol was to 
remind the reader of an entire phrase, verse or section of a 
song, incantation, etc., like those employed by other primi- 
tive races. In any case, there is very little reason for hoping 
that we shall ever be able to reveal the meaning of these 
tablets of Easter Island. 

To try to compare this almost totally unknown quantity 
of our own era with the Indus Valley script, separated from 
it by such a vast distance in space and above all in time, is in 
my opinion altogether too rash an attempt. The mere out- 
ward resemblance of the symbols, undeniable as it is (Fig. 
73) , is still no clue to their meanings or values, nor to the lan- 
guage and the contents of the texts written with them. He 
who does not believe in supernatural connections had better 
ascribe the outward similarity of the two scripts to mere co- 

Our survey thus concludes, seemingly unsatisfactorily, 
with a series of unanswered questions and fanciful conjec- 
tures. In my opinion, however, it would not have been right 





































%>& % w (j u (K rx 

Fig. 73. The resemblance of the sym- 
bols of the Indus Valley scrirjt and of 
the Easter Island writing. (From Jen- 
sen, Die Schrift, Fig. 277.) 

for me to present only the great accomplishments of de- 
cipherment and to disregard the still unsolved problems. It is 
always good to see not only the seemingly momentous 
achievements, but also the limitations of knowledge. After 
all, there still remains the hope that one day also these 
limitations will be overcome and the solution of the seem- 
ingly insoluble problems will be achieved. 


While the original German edition of this book was on the 
press, a very important accomplishment was made in the 
study of the writing of ancient Crete. Unless all appearances 
are deceptive, Michael Ventris, a young British architect, 
succeeded in deciphering the Cretan script known as Linear 
C/ass B. (cf. p. 164). We have learned, to our great sur- 
prise, that the language of these texts is an archaic form of 
the Achaean dialect of the Greek language, which thus was 
used not only in Pylos and Mycenae in Greece proper in the 
1 3th century b.c, but also in Kncssos, on the island of Crete, 
as far hack as about 1400 b.c. Although the details of the 
study are still very far from being positively settled and the 
syllabary shown in Fig. 74, based on the latest accepted find- 
ings, is still likely to be changed further in the course of the 
next few years, the essential facts and general principles ap- 
pear to be quite clearly established. 

It can therefore be stated now that Linear Class B is com- 
posed of approximately 88 phonetic symbols, the phonetic 
value of almost all of which has been exactly established, and 
each and every one has been found to be a syllabic symbol 
representing an open syllable, i.e., a consonant followed by 
a vowel; furthermore, there are also a number of strongly 
pictographic word-signs, for "man," "woman," "horse," "war 
chariot," "tripod," etc., which may be regarded to a certain 
extent also as determinatives. The progress of the decipher- 
ment is impeded by the lack of bilingual texts, but is facili- 
tated by the consistent separation of the individual words by 



a 1* 

pa $ 
da \- 
ta C 

fat © 





na f 

h % 

me ty- 
ne *P 
re Y 

22 % 

Table of Pharetic Symbols 
i ¥ «r Cf 

M«r A 

da f 


/»' ft 

mi V 
ni T 

si A 

du irt 

rtr f 

Str H| 

nu 1=1 
m ¥ 
m E 

fa f 





More Frequent Still Undeciphered Symbols 

st, (:-| 82 *{ S5(«iV ^ 

/,S' 7 * ^ a L e of the P hone «c symbols of the Cretan Linear Class B script 
(Menggi m Ghtte 34, 1954, p. 17.) D ^"P 1, 

Ventris followed the example set by his predecessors, 
Sundwall, Kober and Sittig, in choosing the structure of 
the written words as the starting point for his decipherment, 
a method in general use also for breaking diplomatic codes 
and decoding secret writings. His conclusion that this was a 
syllabic script was reached on the ground of the number of 
the phonetic symbols, and of the analogy of the Cypriote 
syllabic writing. His immediate aim then was to arrange all 
the syllabic signs in a chess-board-like chart, in which every 
syllable beginning with the same consonant would appear 
in the same horizontal tow, and all those ending in the same 
vowel in the same vertical column. 

A very important task preliminary and preparatory to es- 
tablishing definite syllabic values consisted in the exact de- 
termination of the various inflectional forms of the individual 
words; Alice Kober had done a great deal of painstaking de- 
tail work in that respect. She had recognized the fact that 
certain words, substantives according to the pictographic 
word-signs appearing next to them, appeared in three differ- 
ent case forms, but with their stems graphically unaltered. 
Likewise, the appended word-signs for "man" or "woman" 
made it possible to distinguish certain masculine words from 
the corresponding feminine ones by the different endings 
added to the unchanged stems. This is as though we distin- 
guished the Italian bam-bi-NO ("boy") from bam-bi-NA 
("girl") or tut-TO ( "all"— masculine singular) from tut-TA 
("all"— feminine singular) by different syllabic endings 
added to the same identical stem. In such a case, no and na 
would appear in the same horizontal row of the chess-board- 
like chart, to and ta would be in another horizontal row, 
whereas no and to would be found in one vertical column, 
and na and ta in another vertical column. The arranging of 
the syllable signs in the chart with maximum accuracy prior 
to the determination of their phonetic values was one of the 





most difficult tasks, but also one of those most important 
for the further progress of the research. 

In establishing the phonetic values of the syllabic signs of 
Linear Class B, Ventris proceeded differently from the in- 
vestigators before him, in that he completely disregarded the 
resemblance of the Cretan symbols to the signs of the well 
known Cypriote syllabary. He rather concentrated on recog- 
nizing names which would be good subjects for phonetic 
analysis. Male names could be recognized by the ideogram 
for "man" appended to them. Ventris assumed that certain 
groups of signs which occurred frequently suffixed to names 
of persons, were designations of origin and had to contain 
names of localities on Crete. Quite a few such place names 
wers known, and judging by the number of the syllabic signs, 
a shorter, very frequently occurring group could be the name 
Knossos, a longer group the name Amnisos, etc. Making the 
best use of the chess-board-like chart, after wearisome ex- 
periments, Ventris then arrived at the chart of syllabic pho- 
netic values shown, in its present state, in Fig. 74. The find- 
ing that in addition to the Cretan geographic names also 
Greek names of persons, as well as Greek names of deities, 
and even Greek words and inflectional forms were identified 
in the texts, was at first a surprise to the decipherer himself. 
Of course, the Greek words are rendered in these texts in 
an oddly awkward manner, which goes far beyond the cus- 
tom current in the Cypriote syllabic writing and appears to 
be an outright mutilation of the actual spoken forms. Since 
every sign stands for a syllable consisting of a consonant plus 
a vowel, and of no other pattern, difficulties arose first of all 
in the representation of two or more adjacent consonants. 
These difficulties were obviated either by the graphic interpo- 
lation of mute vowels (writing ko-no-so for Knossos, a-mi- 
ni-so for Amnisos) or by writing only one of two adjacent 
consonants (e.g., pa-i-to for Phaistos). In syllables ending 

in 1, m, n, r or s, the script did not indicate these final sounds 
at all, nor did it indicate the i of any i-diphthong. In other 
words, the users of this script wrote po-me for poime'n 
("shepherd"), i-yo-te for iontes ("those who go"), ka-te-u 
for khalkeus ("blacksmith"), and the nominative singular 
kdrvos ("young man"), the accusative singular korvon as 
well as the nominative plural fcorvoi fused in writing in the 
single form ko-vo. Thus, a given written word or phrase can 
be read and of course also translated in very different ways. 
Consequently, a considerable degree of uncertainty will have 
to be taken into account with reference to all future trans- 
lations of texts written in the Linear Class B script 

It seems, however, to be indicated to present a few argu- 
ments to demonstrate that the decipherment progress Is on 
the right path and the uncertainties are to be attributed 
primarily to the imperfection of the script. The Greek 
tripus ( "tripod" ) occurs in one text first in the singular form, 
ri-ri-po — tripus, and is accompanied by the picture of a 
tripod, as a word-sign or determinative, and the figure 1, 
whereas in another place it appears in the dual, ti-ri-po-de = 
tripode = "two tripods," accompanied by the same picto- 
gram and the numeral 2, The same text contains also the 
word depas ("goblet") with various qualifiers: In one in- 
stance, we see di-pa me-vi-yo ti-ri-yo-ve = dipas mevion tri- 
oVes = "a smaller, three-eared (three-handled) goblet," with 
the picture of a three-handled goblet and the numeral 1; we 
find also di-pa me-vi-yo qe-to-ro-ve = dipas mevion qetr-dves 
= "a smaller, four-eared goblet," with the picture of a four- 
handled goblet and the numeral 1; there occurs the phrase 
di-pa-e me-zo-e ti-ri-o-ve-e — dfpae mezoe triSvee — "two 
larger, three-eared goblets," with the picture of a three- 
handled goblet and the numeral 2; and finally, there is the 
particularly pretty di-pa me-vi-yo a-no-ve — dipas mevion an- 
6ves = "a smaller, earless goblet," with the picture of a gob- 


let without handle and the numeral 1. In my opinion, all 

these findings are striking arguments for the correctness of 
the decipherment of Ventris. 

Whereas Linear Class B is thus revealed as Greek, Linear 
Class A seems to be non-Greek and, like the Cretan picto- 
graphic writing, defies all attempts at decipherment for the 
time being. 


Abu, 3 2 

Akcrblad, 19 

Akkadian cuneiform writing (Babylo: 
Assyrian), 3, 4, 39-44, 45. 48+ 51* 5 a . 59, 
60-68, 70, 71, 72, 73, 74, 79, So, Si, &2 t 63* 
85, S6, 90, 01, 132, 153, 155 

Alexander (he Great* 2, 4, 18, 30, 33 

Alexandrian*, Clemens, 5 

AmaJJS, 4 

Amcnemhct ITI. a 

Amenophis III, 3, 24 

Arnrfir)])liis IV, 3 

Amosis, 3 

Arabic writing, 63, 67, 133 

Aramaic language, 49, 110-114 

Arkw right, 108 

Armenian language, 95 

Arrian, 115 

Assarhaddon. 32 

Assurbanipal, 33 

Ave J tic language, 56, 58, 53 

Babyl onion- Assyrian Language, see Akkadian 

Basque language, rjg 

Bauer, Hans, 83^84, l 3h l6 * 

Be hi stun Inscription, 58-59, 65 

Berber language, 123 

Bilalama, 33 

Birch, 25, tig 

B lumen thai, van, 146 

Bork. 130, 168 

Bosch. 116 

fiossett, 97, 98-99, 116-117, 1^9 

Botta, 60, 64 

Brandenstein, Carl Gcorg von, 80 

Brandis, Johannes, 130 

Brugscn, H tin rich, 35 

Budge, 26 

Bugge, 108 

Burnouf, 57 

Catnbysei, 35 

Carian script, 167-169 

Caucasian languages, 109, 139 

Caylus, Count. 51 

Celtic language, 144 

Chabai, 26 

Champollion, Jean Francois, 17, 19-25, 53, 63* 

Chardin, 51 
Charlemagne, 30 
Chefren, 3 
Cheops 2 

Chinese writing, 8, 16, 29, 35, 70 
Collitz Bechtel, 126 
Coptic language, 11, 21, 22 
Cowley, 07 
Cretan-Minoan script, 34, 157, 163-167, 170, 

178, 180 
Cypriote^ reek writing, 124—131, 165-166, 177, 

Cyrus, 33, 55, 63 

Darius, 44, 48, 54, 56, 58, 62, 63, 154 
Deeeke, 130 

Delit2Seh, Fried rich, 67 
Delia Valle, Pietro, 51 
Dhorme, 84, 131 , 134-136 
Dull and, M., 133 
Dutch language, 143 
Early Egyptian, 18, 24, 26 
Early Persian writing, 44, 48, 49, 50-^5 , 6d> 

83, 154 

Early Turkish runic writing, 123 

Easter Island writing, 170-173 

Egyptian demotic writing, 7, 18 

Egyptian hieratic writing, 5-6 

Egyptian hieroglyphic writing, 1, xi, 1-29, 34, 

35. 39> 44. 5*i 86, 90, 01, 0.5, 102, 132, 152, 

153. 159* 160, 162, 165 
Rlamitc language, 35, 44, 52, 82^83 
English words, 12 136 
Epiphanes, Ptolemy, 18 
Erman, Adolph, 24, 36 
ErmaU'fiankc, 4 
Etruscan language, 114, 137-143, 146, 131, 154, 

*55* 1^6, 166 
Falkenstein, 69 
Finnish language, 70 
Forrer, 97 
Frank, Cart, 97 
Friedlander, 115 
Gardiner, 26, 160-162 
Gelb, 97 

German language, 143, 144, 145J 146, 156 
Gothic language, 146 
Greek alphabet, n, 21, 37, 104, 147 
Greek language, k, 18-24, 75. 9*, 105-117, 124- 

131, J39, 145, 146-150, 166, 167, 169, 174- 

Greek language, modern, 156 
Griffith, 27 

Grimme, H., 136, 163 
Grotefend, Georg Fricdrich, 53-59, 62, 63, 93, 

Gudea of Lagash, 30 
Haas, 0-, 149, 150 
Hammurabi, 30, 3i t 33, 68 
HattuiiU HI. 4* 47 
Hebrew language, 1, 63, 67 
Hceren, 57 
Herbig, 114, 143 
Herodotuf. 51, 55, 114, 154 
Hcsychos, 130 
Hevcsy, 170 

Hincks, Edward, 25, 58, 64, 66 
Hinz. 82 
Hittite cuneiform writing, 35, 38, 39, 45 _ 47j 

60-79, &6, 90, 93, 97, 102, 109, 115, 151-152, 

155p 165 
Hittite hieroglyphic writing, xt, 47, 70, 71, 86- 

ioi t 102, 153, 153-154, 157, 170 
Homer, x 
HorapolloQ, 17 
Hoscah, 32 
Hrozny, Fried rich, 70, 74, 75, 76, 97, 98, 157- 

158, ifi5 d 170 
Hungarian Language, 70 
Human Language, 35, 45, 46, 70-81, 84 
Hystaspei, 56, 63, 154 
Illy rig n Language, 144, 145 
Imbcrt. io3 
Indilumma, 92 
Indo-European Languages, Is, 44, 45, 46, 47, 

70, 74, 73, 78, 86, 90, 95, lot, 108, 109, 114, 

143. 144, T46. i47n 148- 150. 166 
Indus Valley script, 157, 158, 169-173 
Italic Language!, 139, 140, 143, 144, 146, 149 
Japanese writing, ifi, 70 
Jensen, 81, 95-97 
Kahle, 114 
Kalinka, E,, 104-105 
Kannpfer, Engelben, 51 




Kelisin Stela, fia 

KLrcbcfp Alhanaxiu*, 17 

Kober, AJice, 164, xoftiff 

Krsbc, J4S, 146 

KrctsciinKr,. P t , icg 

Lands be rger, Beono, 63 

Lang, ft. H. h 136-117 

Lajseo, ChTiitlan, 57 

Latin alphabet, 16, 70, 105, na p 115, ng 

Latin Language, *, 33, 75j ioj^ Illh JI5> „^ 

iao p 1.23, ij&, ij 9p 143-149. *54 i^7 
Latvian Language, 145 
L^ysrd, 60 
Ltpi-iLs, Richard, 25 
Lflttr*asccTn, hidor, 63 
Lnwian Language, 45, 78-751, R5, go 07. ioj 
lfcu.ii language, lOj-tag, 114, i 55fc 165 
Indian Language, iqj, 1*9-1 i;, 143, 455, 169 
fclarjethp, 2 

Marintisa. 117, 119, tan, iai, 123 
Majpero, 26 
Masai. Language. 156 
■Meissner, 32 
-leister, FE.p 149 
Hcaznt, 95 

McOCJ, 3 

Mer^ei, 97, 9*, 10&, f 14, iyo 
Meroitic script, 26-28, 160 
Mcssapian Language, 14.5-14-6, I4B 
Meiserschmidt. -fii 

Middle Assyrian coiMifofca writlrjgj 35- 
Middle Babylonian cuneiform writing 35* 63 
MUkjaton, 126,127 
MJtanni Letter, 80-81 
Miinter, Fried rich, 52-54 
Mursili II h 47 
MyoerLnuSp 2 
Napoleon, 17 

Ncbuchadnrixar II, 32, 63, 65 
Necho, 4 

Nea- Assyrian cuneiform writing a^, 37-3^ 42- 
_4J. *5, 47s Bi 
N en- Babylon Eat cuneiforni writing 1 ,. 35^ 37-3S, 

42-43. *6 
Neo-Egyptian writing, 18, 26 
Nco-Elamrte cuneiform writing,. 59-60, Sip (53 
Neo- Phrygian Language h 147-149, 
New High German, 75 
Niebuhr, Carrtcn, 51-52 
NofrLs, 5a 

Nomidian Language,. 117-135 
Old Assyrian (Cap&adoeiaii) cuneiform wric- 

Old Babylonian, cuneiform writing, 35, 38, 

Old Iodic Language, jec Sanskrit 

Otd Persian Language, 45, 4ft, 50-54, 60, 61. 

J*8, 153 
Old Phrygian Language^ I47-149 
Olzscha, 143, 143 
Oppert, lutes, 50, 66 
Oscn-Umfcrian Language, 135, 142, 143-14^ 

J 4*. ljS 

Palaian Language, 4=5, 78-79, 85 

Papyrus Ebcrs, r6 

Paribeni, 115 

Pedersen, HolgeCp tog, 109 

Petsb, 33 

Fettle, Flinders, 155 

Pbotniekn language, ofl-ioij 126, 138, 120^ 

132, T34 f 136, 152 
Phrygian Language, 146-150 

Proto-Bytlie writing-, 131-156 
Proto-Gtrntanie language, 135 
Proto-Hardan language, 45, 47, 76, 57 

Fjammeuchus I, 4 

Punic language, nfl-123 

Ramses t , 3 

Ramses II, 3-4, 24, 47 

Ramse* III, 4 

Kask. 57 

Rawliauon, Henry, sS p 6i a 65, 66 

Roman alphabet, 157 

Roman names, [2, 24 

Romaneili, 115 

Kosetta Swnep i&-35j 100 

Row Let. 123 

Rouge, dc, 25 

Russian aJphabet, 157 

Sitmanauav I, 3: 

falnanttsttr ni, 32 

SarrJ i-Adad I, 3 1 

Sanherib, 32 

Sanskrit taaguage, 5S, 50, 14B 

Sargign | K 30 

Sar^on ll h 32 

Sayce, Sg, 4a, 94-55 

Scbertdp E., 166-E67 

Sctunidi, Moriz, 130 

Se ? i °f hv&nStt, =» 3. 43, *4r 45» *8. 49, 63. 
67, 83, fl4, fie, nop in, 160, i6l 

Seiastrb II B 2 

Shesltonk, 4 

Sidetic language, 11 5-1 17, 123 

EicgLsmoiid, J30 

Sinairjc writing-, 136, 150-162 

Sitog, 165-166, [77 

Six. io3, 115^ 

Slavic Ungnages, 147 

Smith,. George,. 117-131 

Sodent Wolfram vna, 68 

Sommcr, F., 76 f 114 

South Arabian: writing, 157, 165 

Sunarian Eaaguage, 35, 38, 39, 43, 43, 44, 6B- 
6g h 71-72. 76. 139 

SundwaJE, Jobaitries, 1^4, 167^ 177 

Suppitulituna, 46, 47, 4B 

Taharka, 4 

Tatbot, Foi, 66 

Tarkummusra, -ga 

Th&msfn h Vilheim, i&S, 123 

ThampsoQ, 47 

Thimeaa-Daiigin, 69 

Tlriitmojis 1, 3 

Thatrrcroa III, j 

Tigrath-Pilescr I, 31 

Tigladi-Pileser itl, 32 

T<£>zauae Stela r 82 

Torp, 10S 

Tukulti-S'inurti I, 31 

Turkish language, 70 

Tustatta, 46, 48, So 

Tychsen, Ohv Gerhsrd, 52 

Ugariric knguagc, 48, *u, 8o d S^-SS, 123, 131 

Dram-can. language, 35, 3% 4:5, 47, Ac-Si, 143, 

Ura^tii, 32 

Veoette !anguage h 1*4^45 
Venrris, MJchaei, 175-1 Bo 
ViraEieaud, Sj 
Wadttingtaii, 115 
Wdssbach, P. H., 6S 
Wincfcelmann, 11 
Winkler, Hugo, 6g 
VehUi 33 

Xantbca Stela h 104, 155 
Jkira, ^, s 55 , ^^ 5 g h 0I( 5 3j ^ 
Young, Thomas, 30-12 
Zoega, 17, 20