Copyright © 1956 by Westminster
Theological Seminary, cited with permission.
THE HA-BI-RU--KIN OR FOE
OF
SECOND
ARTICLE
MEREDITH G. KLINE
C.
An Ethno-Professional Interpretation.
It has appeared that the currently dominant
identifications
of the ha-BI-ru as a social class of one sort or another are
inadequate. They fail to discover a common
denominator for
all the ha-BI-ru (and the ha-BI-ru alone) that will satisfy all
the known documents. The investigation must turn to
other
possibilities. Was ethnic unity the
peculiar stamp of the
ha-BI-ru? Was their hallmark the practice of a particular
profession?
1. Ethnic
Unity. Examination of the morphological data
led to the conclusion that the variety of forms
found for the
word ha-BI-ru is most readily explained in terms of variations
of the proper name for an ethnic group.113
Other features point in this same direction:
There are indications of family relationships
among the
ha-BI-ru114 and of self-contained
communities or tribal organi-
zation in the ha-BI-ru
pattern of life.115
The word ha-BI-ru is used in contrast to particular ethnic
terms and, therefore, as at least the equivalent of
an ethnic
term itself. Repeatedly in Hittite rituals and
treaties the
ha-BI-ru are paired with the Lulahhu
(the people of Lullu).
In
one ritual116 this pair appears in a list of social
classes,
113 See supra, WTJ XIX, pp. 9-11.
114 See ibid. p. 21, n. 98 and cf. JEN
V, 452, 453, 456, 465; SMN 2145
for mention of ha-BI-ru women with their children or alone.
115 E. g., at
1956,
pp. 264-265) aptly compares the ha-BI-ru among whom Idri-mi
found political asylum to the tribe in Retenu in which Sinuhe passed his
years of exile.
116 No. 91 in Bottero, op. cit.
170
HA-BI-RU 171
suggesting that "the Lulahhu
and the ha-BI-ru"
had become
a cliche among the
Hittites for the social category of foreign-
ers.117 Such usage, however,
would be only local and secondary
in the case of the ha-BI-ru as it obviously must be in the
case
of the Lulahhu. As a
matter of fact, once it has been estab-
lished that the ha-BI-ru
cannot successfully be identified as
a social class, all evidence that they were
regarded in particular
areas as one specific group of foreigners,118 becomes so much
support for the interpretation of them as a
specific ethnic
entity.
Certain Egyptian texts also mention ha-BI-ru in
lists con-
taining ethnic elements. In the
lists 3,600 ‘pr
(i. e., ha-BI-ru) among those he took captive
on his second Asiatic campaign. They are preceded
by 127119
princes of Rtnu (Syria-Palestine) and 179
brothers of princes.
They
are followed by 15,200 .S3s.w
(Bedouin of the desert
region adjoining
used in the sense of the settled population of
Syria-Palestine)
and 15,070 Ngs (people of Nuhassi). The
intermediate posi-
tion of the ha-BI-ru in
sequence and numerically between the
aristocracy and the ethnic terms would make it
precarious to
determine from this text alone whether the ha-BI-ru were
a
social class or ethnic group. Similar ambiguity is
present in the
testamentary enactment left by Ramses III in which he cites
the properties accumulated by the temples of
opolis, and
politan section the serfs of
the temple are listed as follows:
"warriors, sons of (foreign) princes, maryannu, 'pr.w, and the
settlers who are in this place: 2,093 persons”.120
What is clear
is that the ha-BI-ru were in the eyes of the Egyptians an
easily identifiable group distinct from the Bedouin
and the
general population of Syria-Palestine--a fact
incompatible
117 Perhaps
more specifically, foreign servants. They are located in this
list on the border of the upper and lower strata of
society. In the somewhat
similar list, KUB XXXV, 45, 11, 2 ff., they are closely associated with the
slaves.
118 See supra, WTJ XIX, pp. 18
ff.
119 Or 217 or 144.
120 Papyrus Harris I, 31, 8.
ton, 1950), p. 261, n. 7) regards all these serfs
as foreign. Posener (in
Bottero, op.
cit., p. 170) considers the "settlers" Egyptians.
172
with the theory that the ha-BI-ru were an indistinct social
class.121 Of course in
their foreign status among the Hittites was a local and
temporary condition. It is clear, too, that their
presence in
from Syria- Palestine,123 which was
somehow distinct from
other such troops both general (e. g., the Hr.w) and elite (e. g.,
maryannu). One plausible
explanation of their distinctive-
ness would be that it was ethnic.124
From the Mesopotamian area too come examples of
ha-BI-ru used as the equivalent of an ethnic term. In
the
Mari
texts, for example, the ha-BI-ru are distinguished from
such ethnic groups as the Beni-laminu,
Beni-Simal, and "the
men of Talhaya".125 So again in the
Palestinian area the
121 G. Posener,
ibid., p.
175, observes that in the case of the term 'pr.w,
"Les
determinatifs les designent
simplement comme des
strangers; it ne
s'ajoute aucun
signe qui caracterise une classe sociale,
un genre de vie ou
une occupation, comme on en trouve, d'une fagon reguliere
ou sporadique,
apres des appellatifs
d'emprunt comme mri, mrjn, mskb, n'rn, kt
(n), etc."
According
to Albright, the foreign warrior determinative is used on the
smaller Beisan stele
of Seti I.
122 Cf. also the stele of Ramses IV in the Wadi Hammamat recording
the personnel of an expedition sent to procure
blocks of stone (Couyat
and Montet, Inscriptions hieroglyphiques
du Ouadi Hammamat, no. 12).
The
high priest of Amon heads the list followed by nine
civil and military
officers (Nos. 2-10), 412 subordinate officers
(Nos. 11-16, 18, 21, 22),
5,000
men of the army (No. 17), 800 'pr.w (No. 19), 2,000 slaves (No. 20),
130
quarrymen and stone-cutters (No. 25) and ten skilled artificers and
artists (Nos. 23, 24, 26, 27). Similarly, two
hieratic papyri from
dated to the reign of Ramses
II depict 'pr.w
drawing stone. (Papyrus
123 The Beisan
stele attests the presence of some ha-BI-ru in that area
near 1300 B. C. and the Papyrus Harris 500 account
of the taking of
Joppa
locates ha-BI-ru
there in the 15th century (though the manuscript
itself is 13th century).
124 If the 12th century proper
name, p3-'pr (see no. 191 in Bottero,
op. cit.) has anything to do with the ha-BI-ru, it
might be an indication of
their ethnic distinctiveness since names of the type
article plus substantive
are often ethnic (e. g., p3-hr); they are, however, also professional (e. g.,
p3-hm-ntr, "the
priest").
125 See supra, WTJ XIX, p. 14, n.
66. Cf. A 109. Contrary to Bottero
(op. cit., p. 188) ha-BI-ru is not shown to be an
appellative by the Mari
texts and others which designate certain towns or
countries as the place of
proximate origin or residence of the ha-BI-ru. The ha-BI-ru of
these
HA-BI-RU 173
ha-BI-ru, according to the Amarna
and other evidence, were
a well-defined group which could be contrasted
with ethnic
elements like the Sutu, native Palestinian
troops, and "men
of the
Another feature which comes as no surprise on
the assump-
tion that the ha-BI-ru were
an ethnic group is the mention of
the "gods of the ha-BI-ru" in the Hittite treaties.127
It would
not be as common for inter-ethnic professional
groups to have
guild deities128 and it is unlikely that a
general social class had
its own gods.129 Relevant here is the god dha-BI-ru found in
an Assyrian Gotteradressbuch130
and in Hittite ritual.131 Pos-
sibly the similarity of dha-BI-ru
and LUha-BI-ru
is accidental132
but otherwise there could be evidence here of the
tribal
character of the ha-BI-ru in the appearance of their
eponymous
tribal god.133
texts may also be understood as a distinct ethnic
element not indigenous
to, or only temporarily located in, these places.
126 Cf. e. g., EA 195:24 ff. ; 246:5 ff. ; 318:10 ff.
127 Gustavs
(ZAW, N. F. 3, 1926, pp. 25 ff.)
disposed of the opinion of
Jirku (Orientalistische Literaturzeitung, 1921, pp. 246 ff.; 1922, p. 38; and
Der Alte Orient, 1924, pp. 18 ff.) that the proper translation is "the gods
Ha-BI-ru". Jirku was compelled to
regard as a scribal error: ilaniMES sa
LUSA-GAZ (Keilschrifttexte aus Boghazkoi (hereafter KBo) I, 2, Rs.
27;
cf. I,
3, IV, 5). Nor could he explain the genitive found in all cases but
one (excluding, of course, the use of the
ideogram). The one exception is
a Hittite nominative: (KBo V, 3, I, 56) which Gustavs treated adjectivally.
(Cf. Goetze in Bottero, op. cit., p. 81). Might this reflect the
fact that what
appeared like a nominative elsewhere, i. e., ha-BI-ru,
was a shortened
gentilic? Gustavs
also proved groundless Jirku's view that the ilani was
a plural of majesty.
128 In
129 Greenberg (op. cit., p. 87, n. 9) argues that the
summary type formula
used to designate the gods of the ha-BI-ru
points to an agglomeration of
gods from diverse sources, not to a single pantheon
of an ethnically unified
group. That this is gratuitous is apparent from the
use of the same sum-
mary formula for the gods of
the ethnically unified Lulahhu.
130 KAV 42, II, 9. It is part of a corpus known as the
"Description of
the city of
131 Collection of tablets found
at Boghazkoi (hereafter Bo) 5239:7 and
6868:2.
132 W. von Soden
(in Bottero, op.
cit., p. 135) says of the Neo-Assyrian
dha-BI-ru that it represents the Akkadian ha'iru, hawiru, "spouse".
133 So Jirku,
op. cit. Of uncertain relation to dha-BI-ru
and LUha-BI-ru
are the personal names ha-BI-ra-am (of Old Akkadian
texts), ha-BI-re/ri
174
There are also instances of peace treaty and
covenant
oaths governing the relation of ha-BI-ru groups to kings.134
These
are compatible with an ethnic but not with a social class
interpretation.
The ethnic view is not without problems. Often
urged
against it is the onomastic
evidence, for ha-BI-ru
names range
inside and outside the Semitic sphere.135
Caution, however, is
required in drawing ethnic conclusions from onomastic data.
A
migratory group will adopt names current in their new land,
for imitation of the higher social strata is a
common human
foible.136 According to an ethnic
interpretation of the ha-BI-ru
they will everywhere have assimilated their names to
the
indigenous population except, as far as the
evidence goes, at
Nuzu where they are apparently recently arrived from
a
Semitic
area and even there the process of assimilation to
Hurrian names may be seen to have begun.
and ha-[BI]-ir-di-il-la (from
til-la (from Nuzu), and the Egyptian personal names containing the
element ‘pr.
Gustavs (ZAW,
N. F. 17, 1940, pp. 158, 159) judged ha-BI-
ir-til-la to
be "H. is lord" and thus further evidence of dha-BI-ru. If that
were correct, the fact that -tilla is a common element in Hurrian names
would suggest Hurrian
associations for dha-BI-ru (cf. supra, WTJ
XIX,
p. 4, n. 17). Moreover, most of the Nuzians who bear the names ha-BI-ra
and ha-BI-ir-til-la appear to have Hurrian
relatives. And along with
dha-BI-ru in the Assyrian Gotteradressbuch
are mentioned the Hurrian
deities Seris and Hurris (cf. Albright, BASOR
81, 1941, p. 20. n. 20).
Problematic,
however, for Gustavs' interpretation are the facts
that in
every other case the word compounded with -tilla is verbal
or adjectival
and tilla is itself a Hurrian deity or
surrogate for one.
134 Cf. supra, WTJ XIX, p. 17 and n. 84; and P. A.
Pohl, Orientalia
25,
1956, p. 429. See below for further
treatment of these texts as evidence of
the ha-BI-ru professional character.
135 "The analyzable Old
Babylonian names are Akkadian; those from
Alalalb are, with few exceptions, non-Semitic;
one of the two from
is non-Semitic; from
At
Nuzi H. names, mostly Akkadian,
differ in a marked degree from those
of the local (in this case, Hurrian)
population . . .". So Greenberg sum-
marizes. op. cit., p. 87.
136 While granting that this is
a "proven tendency", Greenberg, ibid.,
n.
9, says that the edge of the above argument has "been dulled by frequent
use". It may be the beginning of scholarship
to realize that an accumula-
tion of authorities does not
validate a view but it is a bit novel to judge
that popularity invalidates one.
HA-BI-RU 175
The wide dispersal of the ha-BI-ru throughout the Fertile
Crescent
and adjacent areas which has earned for them in
modern studies the epithet "ubiquitous" has
also been thought
a difficulty for the theory of ethnic unity. But
it is reasonable
to envisage this ubiquity of the ha-BI-ru as
the sequel of an
ethnic wave that dashed across the
even the earliest extant mention of ha-BI-ru in
Babylonia.137
If
so the question arises whether their ultimate origins lay in
the desert enclosed by the Crescent or in the
tracts beyond.
In opposing the ethnic view Greenberg appeals to
what he
believes to be evidence in the Amarna letters of accretions to
the ha-BI-ru ranks. Thus, Abdi-Ashirta
is called the GAZ-
man;138 "the townsmen of
against the king, are said to `have become
H.'"139; and we
read of Amanhatbi that he
"fled to the SA-GAZ men”.140 If
Canaanites
could so readily become ha-BI-ru (or SA-GAZ)
how can ha-BI-ru denote an ethnic status? The texts in
question, however, mean no more than that certain
leaders and
villagers of
loyalists identified themselves with the efforts
of the ha-BI-ru
in
Canaanites
did not actually become SA-GAZ but became, in
respect to their relationship to the Pharaoh
(the recipient of
these letters), "like GAZ men" (i. e., rebels).141
The major considerations bearing on the
possibility of
ha-BI-ru ethnic unity have now been surveyed. The hypoth-
esis which accounts with the
least difficulty for all the facts
137 DeVaux,
ibid., p.
265, compares the similarly widespread Sutu and
Arameans. Cf. also the Terahites who left elements
of the family in
and
general westward movement of the ha-BI-ru from
covery. Even according to
present evidence the ha-BI-ru are found from
138 EA 91:5.
139 So Greenberg, op. cit., p. 75. The text, EA 288:44, reads: ardutuMES
ip-su a-na L[U.M]ES[h]a-[B]I-[r]i.
140 EA 185:63.
141 In following Abdi Ashirta the people of Ammiya are said to have
become "like GAZ men": i-ba-as-su ki-ma LU.MESGAZ (EA 74:28, 29;
cf. 67:16, 17).
176
is that the ha-BI-ru--at least the characteristic core of
them--did represent one ethnic stock.
2. Professional
Fraternity. Ethnic unity need not have
been the only or even the dominant element in the
Gestalt
called ha-BI-ru. Frequently in the extant record of their
exploits it is their professional role which
occupies the fore-
ground and that role is military. In fact, they are
almost
everywhere and always engaged as professional
warriors. They
man the garrisons at
tolia; conduct razzias along the
Canaan;
and endure the fate of captives of war in
Especially
illuminating are the new pages in ha-BI-ru history
from Alalah and Boghazkoi.
At Alalah the term ha-BI-ru (or
SA-GAZ) denotes the
members of a particular military corps. The
available details
concerning the constituency of this ha-BI-ru corps
contradict
all identifications of the ha-BI-ru as a social class such as the
hupsu. The Hurrianized society of Alalah was
divided into
distinct social classes. The maryannu occupied the top rung,
followed by a free class of tradesmen, the ehelena. Next
came
the rural dwellers called sabe name, subdivided into the hupsu
and haniahu. There were also, as always, the poor (muskenu)
and the slaves.142 Now the significant
thing is that the
membership of the ha-Bl-ru corps cut across these classes.143
It
comprised ehelena,144
muskenu,145 slave,146
and even the
maryannu.147
142 See Wiseman, AT, pp. 10 ff.; Speiser,
JAOS 74, 1954, pp. 18 ff.;
Mendelsohn, BASOR
139, 1955, pp. 9 ff. Wiseman equates only the hupsu
with the sabe name,
associating the haniahu
with the ehelena.
143 Cf. supra, WTJ XIX, p. 16 and n. 78. Eissfeldt
recognized this
(Forschungen and Fortschritte
28:3, March 1954, pp. 80 ff.), but Greenberg
blurs the situation when he comments that the SA-GAZ
"are grouped
with a military class composed of ehele and hanyahe" (op. cit., p. 65).
144 AT 182:27; cf. 180:27.
145 AT 180:31; 182:29.
146 AT 182:14.
147 According to the probable
implications of the charioteers in the
ha-BI-ru corps (AT
180:24; 182:19; 183:6; 226:1) and the most probable
interpretation of the list of family
chiefs (AT 198, esp. line 42; cf. supra,
WTJ XIX, p. 16., n. 80). Since the maryannu status was obtainable by
marriage and royal grant as well as by
inheritance and since this class had
HA-BI-RU 177
Alongside the ha-BI-ru as a second military body at Alalah
is the sanannu corps.148 The two groups have much in
com-
mon. The sanannu corps too
is composed of members of the
various social categories. Both groups consist
in part of
charioteers. The members of both come from towns
around
Alalah and farther afield.149 Both are coordinated with towns
in civil administration. Thus in a cattle census
the totals are
given in terms of the sheep, rams, and asses
belonging to
Alalah, Mukish,150
the SA-GAZ, and the sanannu.151
What the distinction was between the ha-BI-ru and
the
sanannu corps is uncertain.
Perhaps it lay in the area of
military specialization.152 Another
possibility, however, in
line with the apparent ethnic unity of the ha-BI-ru would
be
that the distinction was (at least on the ha-BI-ru side)
ethnic,
as in the case of David's Pelethites
and Cherethites.
Once again in the two new documents153
from the Old
Hittite
royal archives at Hattusha154 the SA-GAZ stand forth
as a distinct corps on a level with other regular
branches of
the Hittite military. In one document155
the SA-GAZ troops
no rigid ethnic barrier (cf. R. T. O'Callaghan, Aram Naharaim,
1948,
p. 66) there is no difficulty in the presence of ha-BI-ru regarded as
substantially an ethnic unity among
the maryannu.
148 See AT 183, 226, and 350.
149 See AT 145 and 341.
150 Wiseman suggests that Mu-ki (-is) -he be read for Mu-ki-he.
252
AT 350; cf. 352. The sanannu total is elsewhere (AT 341) itemized in
terms of sixteen towns around Alalah.
152 Albright (apud Wiseman, op. cit., p. 11, n. 4), relating the sanannu
of Alalah to the tnn of the Ugaritic texts, compares Akkadian
sananu and
suggests tnn, "strive", as the common stem; he translates sanannu as
"archers". Gordon (Ugaritic Manual, Rome, 1955, no. 2049) renders the
Ugaritic tnn, "a kind of soldier"; and the plural,
"members of a certain
guild".
153 At the time of this writing
these documents have not yet been pub-
lished and I am greatly
indebted to Prof. H. Otten for his kindness in
making available to me his article Zwei althethitische Belege
zu den Hapiru
(SA-GAZ)
shortly to appear in Zeitschrift fur Assyriologie,
in which he
presents the texts in transliteration and
translation along with an excellent
discussion. Cf.
P. A. Pohl's reference to these texts in Orientalia 25, 1956,
p. 428.
154 The modern Boghazkoi.
155 141/d=KUB XXXVI, 106.
178
are seen joining the troops of Hatti
in a pledge of allegiance
to the city of
of a self-maledictory
oath, the characteristic covenant form
found in the ritual of oath taking for Hittite
soldiers.156 In
the other document,157 it is the rights of the SA-GAZ troops
which are guaranteed, and that by means of a solemn
oath
taken by the sovereign. This disclosure of the
official status of
that ha-BI-ru within the political establishments at Boghazkoi
and Alalah158 suggests that much of the ha-BI-ru
activity
which has appeared to be independent marauding was
directed
from the capital of one of the ambitious empires of
the day.
3. Proposed
Solution. Two elements are integral to the
entity called ha-BI-ru: ethnic unity and military fraternity.
In
the extant records the military connotation is often
dominant.159
Comparable to this dual character of the ha-BI-ru is
that of
the maryannu. Professionally, they were the experts in chari-
otry; ethnically, the
characteristic core and majority of them
belonged to the Indo-Aryan stock which
constituted the ruling
and patrician class in the unusual symbiosis of Mitannian
society. The maryannu and ha-BI-ru categories are not com-
pletely parallel since, as
noted, the ha-BI-ru
corps cut across
the social classes and included maryannu. Nevertheless, the
maryannu do offer a social
phenomenon in the immediate
historical context of the ha-BI-ru analogous to that presented
here as an interpretation of the ha-BI-ru,
particularly with
respect to the essential point of the
correlativity of ethnic and
professional character in one group.160
And
if the ha-BI-ru
156 Cf. KBo VI, 34 and its duplicate KUB VII, 59.
157
298/n+756/f.
158 Cf. also their employment by governments in the early Babylonian
administrative texts and in some of
the Mari and Amarna letters.
159 Locally the name develops an
even more specialized military sig-
nificance in the LUha-BI-ri officer at Alalah
(AT 164); cf. the SA-GAZ
officer at
16o If it be the case that the ha-BI-ru were
not ethnically one but that
there were additions from various ethnic groups to
the original ethnic
stock of the military organization, that too would
find its parallel in the
maryannu who, though they were
predominantly Indo-Aryan, were not
exclusively so (cf. above n. 147).